diff options
Diffstat (limited to 'src')
| -rw-r--r-- | src/CMakeLists.txt | 56 | ||||
| -rw-r--r-- | src/Content.hpp | 27 | ||||
| -rw-r--r-- | src/ContentManager.cpp | 90 | ||||
| -rw-r--r-- | src/ContentManager.hpp | 38 | ||||
| -rw-r--r-- | src/EventManager.cpp | 206 | ||||
| -rw-r--r-- | src/EventManager.hpp | 49 | ||||
| -rw-r--r-- | src/Filesystem.cpp | 125 | ||||
| -rw-r--r-- | src/Filesystem.hpp | 33 | ||||
| -rw-r--r-- | src/RequestResponse.cpp | 47 | ||||
| -rw-r--r-- | src/RequestResponse.hpp | 29 | ||||
| -rw-r--r-- | src/TemplateManager.cpp | 57 | ||||
| -rw-r--r-- | src/TemplateManager.hpp | 28 | ||||
| -rw-r--r-- | src/content/blog/Blog.cpp | 41 | ||||
| -rw-r--r-- | src/content/blog/Blog.hpp | 25 | ||||
| -rw-r--r-- | src/entity.c | 2190 | ||||
| -rw-r--r-- | src/entity.h | 42 | ||||
| -rw-r--r-- | src/main.cpp | 131 | ||||
| -rw-r--r-- | src/md4c-html.c | 573 | ||||
| -rw-r--r-- | src/md4c-html.h | 68 | ||||
| -rw-r--r-- | src/md4c-html.pc.in | 13 | ||||
| -rw-r--r-- | src/md4c.c | 6383 | ||||
| -rw-r--r-- | src/md4c.h | 405 | ||||
| -rw-r--r-- | src/md4c.pc.in | 13 |
23 files changed, 926 insertions, 9743 deletions
diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt deleted file mode 100644 index 66f2f50..0000000 --- a/src/CMakeLists.txt +++ /dev/null @@ -1,56 +0,0 @@ - -set(CMAKE_WINDOWS_EXPORT_ALL_SYMBOLS 1) -set(CMAKE_C_FLAGS_DEBUG "${CMAKE_C_FLAGS_DEBUG} -DDEBUG") - - -# Build rules for MD4C parser library - -configure_file(md4c.pc.in md4c.pc @ONLY) -add_library(md4c md4c.c md4c.h) -if(CMAKE_C_COMPILER_ID MATCHES "Clang|GNU") - target_compile_options(md4c PRIVATE -Wall -Wextra) -endif() -set_target_properties(md4c PROPERTIES - COMPILE_FLAGS "-DMD4C_USE_UTF8" - VERSION ${MD_VERSION} - SOVERSION ${MD_VERSION_MAJOR} - PUBLIC_HEADER md4c.h -) - -# Build rules for HTML renderer library - -configure_file(md4c-html.pc.in md4c-html.pc @ONLY) -add_library(md4c-html md4c-html.c md4c-html.h entity.c entity.h) -set_target_properties(md4c-html PROPERTIES - VERSION ${MD_VERSION} - SOVERSION ${MD_VERSION_MAJOR} - PUBLIC_HEADER md4c-html.h -) -target_link_libraries(md4c-html md4c) - - -# Install rules - -install( - TARGETS md4c - EXPORT md4cConfig - ARCHIVE DESTINATION ${CMAKE_INSTALL_LIBDIR} - LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR} - RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR} - PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR} - INCLUDES DESTINATION ${CMAKE_INSTALL_INCLUDEDIR} -) -install(FILES ${CMAKE_BINARY_DIR}/src/md4c.pc DESTINATION ${CMAKE_INSTALL_LIBDIR}/pkgconfig) - -install( - TARGETS md4c-html - EXPORT md4cConfig - ARCHIVE DESTINATION ${CMAKE_INSTALL_LIBDIR} - LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR} - RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR} - PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR} -) -install(FILES ${CMAKE_BINARY_DIR}/src/md4c-html.pc DESTINATION ${CMAKE_INSTALL_LIBDIR}/pkgconfig) - -install(EXPORT md4cConfig DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/md4c/ NAMESPACE md4c::) - diff --git a/src/Content.hpp b/src/Content.hpp new file mode 100644 index 0000000..82eb35e --- /dev/null +++ b/src/Content.hpp @@ -0,0 +1,27 @@ +#ifndef CONTENT_H +#define CONTENT_H 1 + +#include "RequestResponse.hpp" + +#include <inja/inja.hpp> +#include <string> +#include <tuple> +#include <vector> + +using Redirections = std::vector<std::string>; +using RenderData = std::unordered_map<std::string, std::string>; + +class Content +{ +public: + virtual bool Init() = 0; + virtual void Shutdown() = 0; + + virtual bool Render(RequestResponse & rr, RenderData & rd) = 0; + + virtual std::string const & GetUriBasePath() const = 0; + virtual std::string const & GetMainTemplate() const = 0; + virtual Redirections const & GetRedirections() const = 0; +}; + +#endif diff --git a/src/ContentManager.cpp b/src/ContentManager.cpp new file mode 100644 index 0000000..6beef5a --- /dev/null +++ b/src/ContentManager.cpp @@ -0,0 +1,90 @@ +#include "ContentManager.hpp" + +void ContentManager::SetStaticFilesystem(std::shared_ptr<Filesystem> & static_fs) +{ + m_StaticFilesystem = static_fs; +} + +void ContentManager::SetTemplateSystem(std::shared_ptr<TemplateManager> & tmgr) +{ + m_TemplateManager = tmgr; +} + +bool ContentManager::RegisterModule(std::shared_ptr<Content> ctnt) +{ + std::string const & basePath = ctnt->GetUriBasePath(); + Redirections const & rs = ctnt->GetRedirections(); + + m_ContentModules[basePath] = ctnt; + for (auto & redirect : rs) + { + m_ContentModules[redirect] = ctnt; + } + + return false; +} + +bool ContentManager::InitAll(void) +{ + bool ret = true; + + for (auto & content : m_ContentModules) + { + if (content.second->Init() == false) + { + ret = false; + } + } + + return ret; +} + +void ContentManager::ShutdownAll(void) +{ + std::unordered_map<std::shared_ptr<Content>, bool> shutdownModules; + + for (auto & content : m_ContentModules) + { + auto const & search = shutdownModules.find(content.second); + if (search != shutdownModules.end()) + { + continue; + } + else + { + content.second->Shutdown(); + shutdownModules[content.second] = true; + } + } + + m_ContentModules.clear(); +} + +bool ContentManager::Render(char const * basePath, RequestResponse & rr, std::string & out) +{ + if (m_ContentModules.find(basePath) == m_ContentModules.end()) + { + return false; + } + + RenderData rd; + auto & cntm = m_ContentModules[basePath]; + auto & main = cntm->GetMainTemplate(); + + if (m_ContentModules[basePath]->Render(rr, rd) == false) + { + return false; + } + + if (m_TemplateManager->RenderTemplate(main, rd, out) == false) + { + return false; + } + + return true; +} + +ContentModules const & ContentManager::GetAllModules() const +{ + return m_ContentModules; +} diff --git a/src/ContentManager.hpp b/src/ContentManager.hpp new file mode 100644 index 0000000..83f4bc9 --- /dev/null +++ b/src/ContentManager.hpp @@ -0,0 +1,38 @@ +#ifndef CONTENTMANAGER_H +#define CONTENTMANAGER_H 1 + +#include "Content.hpp" +#include "Filesystem.hpp" +#include "TemplateManager.hpp" + +#include <memory> +#include <unordered_map> + +using ContentModules = std::unordered_map<std::string, std::shared_ptr<Content> >; + +class ContentManager +{ +public: + ContentManager() + { + } + ~ContentManager() + { + ShutdownAll(); + } + + void SetStaticFilesystem(std::shared_ptr<Filesystem> & static_fs); + void SetTemplateSystem(std::shared_ptr<TemplateManager> & tmgr); + bool RegisterModule(std::shared_ptr<Content> ctnt); + bool InitAll(void); + void ShutdownAll(void); + bool Render(char const * basePath, RequestResponse & rr, std::string & out); + ContentModules const & GetAllModules() const; + +private: + std::shared_ptr<Filesystem> m_StaticFilesystem; + std::shared_ptr<TemplateManager> m_TemplateManager; + ContentModules m_ContentModules; +}; + +#endif diff --git a/src/EventManager.cpp b/src/EventManager.cpp new file mode 100644 index 0000000..3270297 --- /dev/null +++ b/src/EventManager.cpp @@ -0,0 +1,206 @@ +#include "EventManager.hpp" + +#include <event2/buffer.h> +#include <event2/event.h> +#include <event2/http.h> +#include <event2/keyvalq_struct.h> +#include <event2/listener.h> +#include <event2/util.h> +#include <signal.h> + +extern "C" +{ + static void GenerateInternalErrorPage(struct evhttp_request * const req, std::string text) + { + evhttp_add_header(evhttp_request_get_output_headers(req), "Content-Type", "text/html"); + + struct evbuffer * const output = evbuffer_new(); + if (output != nullptr) + { + evbuffer_add(output, text.c_str(), text.size()); + evhttp_send_reply(req, 500, "Internal Error", output); + evbuffer_free(output); + } + } + + static void EvGenericInterceptor(struct evhttp_request * const req, void * ev_c_callback) + { + if (ev_c_callback != nullptr) + { + struct ev_callback * const evcb = (struct ev_callback *)ev_c_callback; + evcb->cb(req, evcb->ud); + } + else + { + GenerateInternalErrorPage(req, "EvGenericInterceptor: ev_c_callback == nullptr"); + } + } + + static void EvContentManagerInterceptor(struct evhttp_request * const req, void * ev_c_callback) + { + if (ev_c_callback != nullptr) + { + struct evhttp_uri const * const uri = evhttp_request_get_evhttp_uri(req); + if (uri == nullptr) + { + GenerateInternalErrorPage(req, "EvContentManagerInterceptor: uri == nullptr"); + return; + } + + char const * const path = evhttp_uri_get_path(uri); + if (path == nullptr) + { + GenerateInternalErrorPage(req, "EvContentManagerInterceptor: path == nullptr"); + return; + } + + std::shared_ptr<ContentManager> const cmgr = *(std::shared_ptr<ContentManager> const *)ev_c_callback; + RequestResponse rr(req); + std::string out; + + if (cmgr->Render(path, rr, out) == false) + { + std::string text; + text = "ContentModule(\"" + std::string(path) + "\")->Render() failed.\n"; + GenerateInternalErrorPage(req, text); + } + else + { + evhttp_add_header(evhttp_request_get_output_headers(req), "Content-Type", "text/html"); + + struct evbuffer * const output = evbuffer_new(); + if (output != nullptr) + { + evbuffer_add(output, out.c_str(), out.size()); + evhttp_send_reply(req, HTTP_OK, "OK", output); + evbuffer_free(output); + } + } + } + else + { + GenerateInternalErrorPage(req, "EvContentManagerInterceptor: ev_c_callback == nullptr"); + } + } + + static void do_term(int sig, short events, void * arg) + { + (void)events; + struct event_base * base = (struct event_base *)arg; + event_base_loopbreak(base); + fprintf(stderr, "Got %i, Terminating\n", sig); + } +} + +static inline void default_evhttp_callback(struct evhttp_request * const req, EvUserData ud) +{ + (void)ud; + + evhttp_add_header(evhttp_request_get_output_headers(req), "Content-Type", "text/html"); + + struct evbuffer * const output = evbuffer_new(); + if (output != nullptr) + { + evbuffer_add_printf(output, "%s\n", "<html><body><b>default page</b></body></html>"); + evhttp_send_reply(req, 200, "OK", output); + evbuffer_free(output); + } +} + +EventManager::EventManager(std::shared_ptr<ContentManager> & cmgr) + : m_ContentManager(cmgr), m_DefaultCallback({default_evhttp_callback, nullptr}) +{ +} + +EventManager::~EventManager() +{ + if (m_EvConfig != nullptr) + event_config_free(m_EvConfig); + if (m_EvHttp != nullptr) + evhttp_free(m_EvHttp); + if (m_EvTermEvent != nullptr) + event_free(m_EvTermEvent); + if (m_EvBase != nullptr) + event_base_free(m_EvBase); +} + +bool EventManager::Init(std::string host, uint16_t port) +{ + struct event_config * cfg = nullptr; + + if (signal(SIGPIPE, SIG_IGN) == SIG_ERR) + { + return false; + } + + event_enable_debug_logging(EVENT_DBG_ALL); + + cfg = event_config_new(); + event_config_set_flag(cfg, EVENT_BASE_FLAG_STARTUP_IOCP); + + m_EvBase = event_base_new_with_config(cfg); + if (m_EvBase == nullptr) + { + fprintf(stderr, "Couldn't create an event_base: exiting\n"); + return false; + } + event_config_free(cfg); + cfg = nullptr; + + /* Create a new evhttp object to handle requests. */ + m_EvHttp = evhttp_new(m_EvBase); + if (m_EvHttp == nullptr) + { + fprintf(stderr, "couldn't create evhttp. Exiting.\n"); + return false; + } + + for (auto & uc : m_UrlCallbacks) + { + struct ev_callback * const evcb = &std::get<1>(uc); + if (evhttp_set_cb(m_EvHttp, std::get<0>(uc).c_str(), EvGenericInterceptor, evcb) != 0) + { + return false; + } + } + for (auto & cm : m_ContentManager->GetAllModules()) + { + if (evhttp_set_cb(m_EvHttp, cm.first.c_str(), EvContentManagerInterceptor, &m_ContentManager) != 0) + { + return false; + } + } + evhttp_set_gencb(m_EvHttp, EvGenericInterceptor, &m_DefaultCallback); + + m_EvSocket = evhttp_bind_socket_with_handle(m_EvHttp, host.c_str(), port); + if (m_EvSocket == nullptr) + { + fprintf(stderr, "couldn't bind to %s:%d. Exiting.\n", host.c_str(), port); + return false; + } + + m_EvTermEvent = evsignal_new(m_EvBase, SIGINT, do_term, m_EvBase); + if (m_EvTermEvent == nullptr) + { + return false; + } + if (event_add(m_EvTermEvent, NULL) != 0) + { + return false; + } + + event_base_dispatch(m_EvBase); + + return true; +} + +void EventManager::SetDefaultCallback(EvFunction fn, EvUserData dat) +{ + m_DefaultCallback.cb = fn; + m_DefaultCallback.ud = dat; +} + +void EventManager::AddCallback(std::string url, EvFunction fn, EvUserData dat) +{ + m_UrlCallbacks.push_back(EvUrlCallback(url, {fn, dat})); +} diff --git a/src/EventManager.hpp b/src/EventManager.hpp new file mode 100644 index 0000000..cf09e14 --- /dev/null +++ b/src/EventManager.hpp @@ -0,0 +1,49 @@ +#ifndef EVENT_MANAGER_H +#define EVENT_MANAGER_H 1 + +#include "ContentManager.hpp" +#include "Filesystem.hpp" + +#include <event2/buffer.h> +#include <event2/event.h> +#include <event2/http.h> + +#include <functional> +#include <string> +#include <vector> + +using EvUserData = void *; +using EvFunction = std::function<void(struct evhttp_request *, EvUserData)>; + +struct ev_callback +{ + EvFunction cb; + EvUserData ud; +}; + +using EvUrlCallback = std::tuple<std::string, struct ev_callback>; + +class EventManager +{ +public: + EventManager(std::shared_ptr<ContentManager> & cmgr); + ~EventManager(); + + bool Init(std::string = "127.0.0.1", uint16_t port = 9000); + void SetDefaultCallback(EvFunction fn, EvUserData dat); + void AddCallback(std::string url, EvFunction fn, EvUserData dat); + +private: + std::shared_ptr<ContentManager> m_ContentManager; + struct ev_callback m_DefaultCallback; + std::vector<EvUrlCallback> m_UrlCallbacks; + + struct event_config * m_EvConfig = nullptr; + struct event_base * m_EvBase = nullptr; + struct evhttp * m_EvHttp = nullptr; + struct evhttp_bound_socket * m_EvSocket = nullptr; + struct evconnlistener * m_EvListener = nullptr; + struct event * m_EvTermEvent = nullptr; +}; + +#endif diff --git a/src/Filesystem.cpp b/src/Filesystem.cpp new file mode 100644 index 0000000..09e5146 --- /dev/null +++ b/src/Filesystem.cpp @@ -0,0 +1,125 @@ +#include "Filesystem.hpp" + +#include <cstring> +#include <filesystem> +#include <fstream> +#include <iostream> + +static std::string make_path_relative(std::string & path, std::string & root) +{ + return std::filesystem::relative(path, root); +} + +bool Filesystem::AddSingleFile(std::string path, std::string root) +{ + std::ifstream ifs(path, std::ios::binary | std::ios::ate); + + if (!ifs) + { + return false; + } + + auto end = ifs.tellg(); + if (end <= 0) + { + return false; + } + if (!ifs.seekg(0, std::ios::beg)) + { + return false; + } + + auto size = std::size_t(end - ifs.tellg()); + + if (size == 0) + { + return false; + } + + struct file_data fd = {}; + try + { + fd.data.resize(size); + } + catch (const std::exception & e) + { + return false; + } + + if (!ifs.read((char *)fd.data.data(), fd.data.size())) + { + return false; + } + + std::string relpath = make_path_relative(path, root); + if (m_Files.count(relpath) > 0) + { + std::cout << "Adding file: " << path << " (" << size << " bytes) and overwriting " << relpath << " to " << this + << std::endl; + } + else + { + std::cout << "Adding file: " << path << " (" << size << " bytes) as " << relpath << " to " << this << std::endl; + } + + m_Files[relpath] = fd; + + return true; +} + +bool Filesystem::Scan(std::string root) +{ + bool retval = true; + + for (const auto & entry : std::filesystem::directory_iterator(root)) + { + if (AddSingleFile(entry.path(), root) == false) + { + retval = true; + } + } + + return retval; +} + +bool Filesystem::Scan(std::string root, std::vector<std::string> extensions, bool exclude_extensions) +{ + bool retval = true; + + for (const auto & entry : std::filesystem::directory_iterator(root)) + { + std::string ext = std::filesystem::path(entry).extension(); + bool found_extension = false; + + for (const auto & extension : extensions) + { + if (ext == "." + extension) + { + found_extension = true; + break; + } + } + + if (found_extension == true && exclude_extensions == false) + { + if (AddSingleFile(entry.path(), root) == false) + { + retval = false; + } + } + if (found_extension == false && exclude_extensions == true) + { + if (AddSingleFile(entry.path(), root) == false) + { + retval = false; + } + } + } + + return retval; +} + +const std::unordered_map<std::string, struct file_data> & Filesystem::GetFiles() const +{ + return m_Files; +} diff --git a/src/Filesystem.hpp b/src/Filesystem.hpp new file mode 100644 index 0000000..13733c9 --- /dev/null +++ b/src/Filesystem.hpp @@ -0,0 +1,33 @@ +#ifndef FILESYSTEM_H +#define FILESYSTEM_H 1 + +#include <inja/inja.hpp> +#include <string> +#include <unordered_map> +#include <vector> + +struct file_data +{ + std::vector<unsigned char> data; +}; + +class Filesystem +{ +public: + Filesystem() + { + } + ~Filesystem() + { + } + + bool AddSingleFile(std::string path, std::string root); + bool Scan(std::string root = "./wwwroot"); + bool Scan(std::string root, std::vector<std::string> extensions, bool exclude_extensions = false); + const std::unordered_map<std::string, struct file_data> & GetFiles() const; + +private: + std::unordered_map<std::string, struct file_data> m_Files; +}; + +#endif diff --git a/src/RequestResponse.cpp b/src/RequestResponse.cpp new file mode 100644 index 0000000..2567253 --- /dev/null +++ b/src/RequestResponse.cpp @@ -0,0 +1,47 @@ +#include "RequestResponse.hpp" + +RequestResponse::RequestResponse(struct evhttp_request * const req) : m_Request(req) +{ +} + +RequestResponse::~RequestResponse() +{ +} + +void RequestResponse::UseInputHeader() +{ + m_InputHeader = evhttp_request_get_input_headers(m_Request); +} + +void RequestResponse::UseOutputHeader() +{ + m_OutputHeader = evhttp_request_get_output_headers(m_Request); +} + +bool RequestResponse::AddOutputHeaderByRef(std::string & key, std::string & value) +{ + return evhttp_add_header(m_OutputHeader, key.c_str(), value.c_str()); +} + +bool RequestResponse::AddOutputHeader(std::string key, std::string value) +{ + return AddOutputHeaderByRef(key, value); +} + +bool RequestResponse::GetInputHeaderByRef(std::string & key, std::string & value) +{ + char const * const v = evhttp_find_header(m_InputHeader, key.c_str()); + + if (v == nullptr) + { + return false; + } + + value = v; + return true; +} + +bool RequestResponse::GetInputHeader(std::string key, std::string value) +{ + return GetInputHeader(key, value); +} diff --git a/src/RequestResponse.hpp b/src/RequestResponse.hpp new file mode 100644 index 0000000..8a7e1d3 --- /dev/null +++ b/src/RequestResponse.hpp @@ -0,0 +1,29 @@ +#ifndef REQUEST_RESPONSE_H +#define REQUEST_RESPONSE_H 1 + +#include <event2/http.h> + +#include <string> + +class RequestResponse +{ +public: + RequestResponse(struct evhttp_request * const req); + ~RequestResponse(); + + void UseInputHeader(); + void UseOutputHeader(); + + bool AddOutputHeaderByRef(std::string & key, std::string & value); + bool AddOutputHeader(std::string key, std::string value); + + bool GetInputHeaderByRef(std::string & key, std::string & value); + bool GetInputHeader(std::string key, std::string value); + +private: + struct evhttp_request * const m_Request; + struct evkeyvalq * m_InputHeader; + struct evkeyvalq * m_OutputHeader; +}; + +#endif diff --git a/src/TemplateManager.cpp b/src/TemplateManager.cpp new file mode 100644 index 0000000..06771ea --- /dev/null +++ b/src/TemplateManager.cpp @@ -0,0 +1,57 @@ +#include "TemplateManager.hpp" + +#include <filesystem> + +TemplateManager::TemplateManager() +{ + AddInjaCallback("test_fn", 0, [](inja::Arguments & args) { + (void)args; + return "Just a test fn."; + }); + AddInjaCallback("test_return_true", 0, [](inja::Arguments & args) { + (void)args; + return true; + }); +} + +void TemplateManager::ParseTemplates(Filesystem const & fs) +{ + for (auto & tpl : fs.GetFiles()) + { + std::string tmpl(tpl.second.data.data(), tpl.second.data.data() + tpl.second.data.size()); + m_Templates[tpl.first] = m_Inja.parse(tmpl); + std::cout << "File: " << tpl.first << " may contain a renderable template." << std::endl; + } +} + +void TemplateManager::AddInjaCallback(std::string functionName, + std::size_t numberOfArgs, + inja::CallbackFunction function) +{ + m_Inja.add_callback(functionName, numberOfArgs, function); +} + +void TemplateManager::AddVoidInjaCallback(std::string functionName, + std::size_t numberOfArgs, + inja::VoidCallbackFunction function) +{ + m_Inja.add_void_callback(functionName, numberOfArgs, function); +} + +bool TemplateManager::TemplateExists(std::string const & templatePath) +{ + return m_Templates.find(templatePath) != m_Templates.end(); +} + +bool TemplateManager::RenderTemplate(std::string const & templatePath, RenderData const & rd, std::string & out) +{ + if (TemplateExists(templatePath) == false) + { + return false; + } + + inja::json ij(rd); + out = m_Inja.render(m_Templates[templatePath].content, ij); + + return true; +} diff --git a/src/TemplateManager.hpp b/src/TemplateManager.hpp new file mode 100644 index 0000000..7b89d71 --- /dev/null +++ b/src/TemplateManager.hpp @@ -0,0 +1,28 @@ +#ifndef TEMPLATEMANAGER_H +#define TEMPLATEMANAGER_H 1 + +#include "Content.hpp" +#include "Filesystem.hpp" + +#include <inja/inja.hpp> + +class TemplateManager +{ +public: + TemplateManager(); + ~TemplateManager() + { + } + + void ParseTemplates(Filesystem const & fs); + void AddInjaCallback(std::string functionName, std::size_t numberOfArgs, inja::CallbackFunction function); + void AddVoidInjaCallback(std::string functionName, std::size_t numberOfArgs, inja::VoidCallbackFunction function); + bool TemplateExists(std::string const & templatePath); + bool RenderTemplate(std::string const & templatePath, RenderData const & rd, std::string & out); + +private: + inja::TemplateStorage m_Templates; + inja::Environment m_Inja; +}; + +#endif diff --git a/src/content/blog/Blog.cpp b/src/content/blog/Blog.cpp new file mode 100644 index 0000000..56a23c1 --- /dev/null +++ b/src/content/blog/Blog.cpp @@ -0,0 +1,41 @@ +#include "Blog.hpp" + +Blog::Blog(std::string uriBasePath, std::string mainTemplatePath) + : Content(), m_UriBasePath(uriBasePath), m_MainTemplatePath(mainTemplatePath) +{ + m_Redirections.push_back(uriBasePath + "/"); + m_Redirections.push_back(uriBasePath + "/index.html"); +} + +bool Blog::Init() +{ + return true; +} + +void Blog::Shutdown() +{ +} + +bool Blog::Render(RequestResponse & rr, RenderData & rd) +{ + rd["blah"] = "Yoh!"; + rr.UseOutputHeader(); + rr.AddOutputHeader("bla", "blubb"); + + return true; +} + +std::string const & Blog::GetUriBasePath() const +{ + return m_UriBasePath; +} + +std::string const & Blog::GetMainTemplate() const +{ + return m_MainTemplatePath; +} + +Redirections const & Blog::GetRedirections() const +{ + return m_Redirections; +} diff --git a/src/content/blog/Blog.hpp b/src/content/blog/Blog.hpp new file mode 100644 index 0000000..086acc7 --- /dev/null +++ b/src/content/blog/Blog.hpp @@ -0,0 +1,25 @@ +#ifndef BLOG_H +#define BLOG_H 1 + +#include "../../Content.hpp" + +class Blog : public Content +{ +public: + explicit Blog(std::string uriBasePath, std::string mainTemplatePath); + + bool Init(); + void Shutdown(); + bool Render(RequestResponse & rr, RenderData & rd); + + std::string const & GetUriBasePath() const; + std::string const & GetMainTemplate() const; + Redirections const & GetRedirections() const; + +private: + std::string m_UriBasePath; + std::string m_MainTemplatePath; + Redirections m_Redirections; +}; + +#endif diff --git a/src/entity.c b/src/entity.c deleted file mode 100644 index 9991ca1..0000000 --- a/src/entity.c +++ /dev/null @@ -1,2190 +0,0 @@ -/* - * MD4C: Markdown parser for C - * (http://github.com/mity/md4c) - * - * Copyright (c) 2016-2017 Martin Mitas - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS - * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ - -#include "entity.h" -#include <string.h> - - -/* The table is generated from https://html.spec.whatwg.org/entities.json */ -static const struct entity entity_table[] = { - { "Æ", { 198, 0 } }, - { "&", { 38, 0 } }, - { "Á", { 193, 0 } }, - { "Ă", { 258, 0 } }, - { "Â", { 194, 0 } }, - { "А", { 1040, 0 } }, - { "𝔄", { 120068, 0 } }, - { "À", { 192, 0 } }, - { "Α", { 913, 0 } }, - { "Ā", { 256, 0 } }, - { "⩓", { 10835, 0 } }, - { "Ą", { 260, 0 } }, - { "𝔸", { 120120, 0 } }, - { "⁡", { 8289, 0 } }, - { "Å", { 197, 0 } }, - { "𝒜", { 119964, 0 } }, - { "≔", { 8788, 0 } }, - { "Ã", { 195, 0 } }, - { "Ä", { 196, 0 } }, - { "∖", { 8726, 0 } }, - { "⫧", { 10983, 0 } }, - { "⌆", { 8966, 0 } }, - { "Б", { 1041, 0 } }, - { "∵", { 8757, 0 } }, - { "ℬ", { 8492, 0 } }, - { "Β", { 914, 0 } }, - { "𝔅", { 120069, 0 } }, - { "𝔹", { 120121, 0 } }, - { "˘", { 728, 0 } }, - { "ℬ", { 8492, 0 } }, - { "≎", { 8782, 0 } }, - { "Ч", { 1063, 0 } }, - { "©", { 169, 0 } }, - { "Ć", { 262, 0 } }, - { "⋒", { 8914, 0 } }, - { "ⅅ", { 8517, 0 } }, - { "ℭ", { 8493, 0 } }, - { "Č", { 268, 0 } }, - { "Ç", { 199, 0 } }, - { "Ĉ", { 264, 0 } }, - { "∰", { 8752, 0 } }, - { "Ċ", { 266, 0 } }, - { "¸", { 184, 0 } }, - { "·", { 183, 0 } }, - { "ℭ", { 8493, 0 } }, - { "Χ", { 935, 0 } }, - { "⊙", { 8857, 0 } }, - { "⊖", { 8854, 0 } }, - { "⊕", { 8853, 0 } }, - { "⊗", { 8855, 0 } }, - { "∲", { 8754, 0 } }, - { "”", { 8221, 0 } }, - { "’", { 8217, 0 } }, - { "∷", { 8759, 0 } }, - { "⩴", { 10868, 0 } }, - { "≡", { 8801, 0 } }, - { "∯", { 8751, 0 } }, - { "∮", { 8750, 0 } }, - { "ℂ", { 8450, 0 } }, - { "∐", { 8720, 0 } }, - { "∳", { 8755, 0 } }, - { "⨯", { 10799, 0 } }, - { "𝒞", { 119966, 0 } }, - { "⋓", { 8915, 0 } }, - { "≍", { 8781, 0 } }, - { "ⅅ", { 8517, 0 } }, - { "⤑", { 10513, 0 } }, - { "Ђ", { 1026, 0 } }, - { "Ѕ", { 1029, 0 } }, - { "Џ", { 1039, 0 } }, - { "‡", { 8225, 0 } }, - { "↡", { 8609, 0 } }, - { "⫤", { 10980, 0 } }, - { "Ď", { 270, 0 } }, - { "Д", { 1044, 0 } }, - { "∇", { 8711, 0 } }, - { "Δ", { 916, 0 } }, - { "𝔇", { 120071, 0 } }, - { "´", { 180, 0 } }, - { "˙", { 729, 0 } }, - { "˝", { 733, 0 } }, - { "`", { 96, 0 } }, - { "˜", { 732, 0 } }, - { "⋄", { 8900, 0 } }, - { "ⅆ", { 8518, 0 } }, - { "𝔻", { 120123, 0 } }, - { "¨", { 168, 0 } }, - { "⃜", { 8412, 0 } }, - { "≐", { 8784, 0 } }, - { "∯", { 8751, 0 } }, - { "¨", { 168, 0 } }, - { "⇓", { 8659, 0 } }, - { "⇐", { 8656, 0 } }, - { "⇔", { 8660, 0 } }, - { "⫤", { 10980, 0 } }, - { "⟸", { 10232, 0 } }, - { "⟺", { 10234, 0 } }, - { "⟹", { 10233, 0 } }, - { "⇒", { 8658, 0 } }, - { "⊨", { 8872, 0 } }, - { "⇑", { 8657, 0 } }, - { "⇕", { 8661, 0 } }, - { "∥", { 8741, 0 } }, - { "↓", { 8595, 0 } }, - { "⤓", { 10515, 0 } }, - { "⇵", { 8693, 0 } }, - { "̑", { 785, 0 } }, - { "⥐", { 10576, 0 } }, - { "⥞", { 10590, 0 } }, - { "↽", { 8637, 0 } }, - { "⥖", { 10582, 0 } }, - { "⥟", { 10591, 0 } }, - { "⇁", { 8641, 0 } }, - { "⥗", { 10583, 0 } }, - { "⊤", { 8868, 0 } }, - { "↧", { 8615, 0 } }, - { "⇓", { 8659, 0 } }, - { "𝒟", { 119967, 0 } }, - { "Đ", { 272, 0 } }, - { "Ŋ", { 330, 0 } }, - { "Ð", { 208, 0 } }, - { "É", { 201, 0 } }, - { "Ě", { 282, 0 } }, - { "Ê", { 202, 0 } }, - { "Э", { 1069, 0 } }, - { "Ė", { 278, 0 } }, - { "𝔈", { 120072, 0 } }, - { "È", { 200, 0 } }, - { "∈", { 8712, 0 } }, - { "Ē", { 274, 0 } }, - { "◻", { 9723, 0 } }, - { "▫", { 9643, 0 } }, - { "Ę", { 280, 0 } }, - { "𝔼", { 120124, 0 } }, - { "Ε", { 917, 0 } }, - { "⩵", { 10869, 0 } }, - { "≂", { 8770, 0 } }, - { "⇌", { 8652, 0 } }, - { "ℰ", { 8496, 0 } }, - { "⩳", { 10867, 0 } }, - { "Η", { 919, 0 } }, - { "Ë", { 203, 0 } }, - { "∃", { 8707, 0 } }, - { "ⅇ", { 8519, 0 } }, - { "Ф", { 1060, 0 } }, - { "𝔉", { 120073, 0 } }, - { "◼", { 9724, 0 } }, - { "▪", { 9642, 0 } }, - { "𝔽", { 120125, 0 } }, - { "∀", { 8704, 0 } }, - { "ℱ", { 8497, 0 } }, - { "ℱ", { 8497, 0 } }, - { "Ѓ", { 1027, 0 } }, - { ">", { 62, 0 } }, - { "Γ", { 915, 0 } }, - { "Ϝ", { 988, 0 } }, - { "Ğ", { 286, 0 } }, - { "Ģ", { 290, 0 } }, - { "Ĝ", { 284, 0 } }, - { "Г", { 1043, 0 } }, - { "Ġ", { 288, 0 } }, - { "𝔊", { 120074, 0 } }, - { "⋙", { 8921, 0 } }, - { "𝔾", { 120126, 0 } }, - { "≥", { 8805, 0 } }, - { "⋛", { 8923, 0 } }, - { "≧", { 8807, 0 } }, - { "⪢", { 10914, 0 } }, - { "≷", { 8823, 0 } }, - { "⩾", { 10878, 0 } }, - { "≳", { 8819, 0 } }, - { "𝒢", { 119970, 0 } }, - { "≫", { 8811, 0 } }, - { "Ъ", { 1066, 0 } }, - { "ˇ", { 711, 0 } }, - { "^", { 94, 0 } }, - { "Ĥ", { 292, 0 } }, - { "ℌ", { 8460, 0 } }, - { "ℋ", { 8459, 0 } }, - { "ℍ", { 8461, 0 } }, - { "─", { 9472, 0 } }, - { "ℋ", { 8459, 0 } }, - { "Ħ", { 294, 0 } }, - { "≎", { 8782, 0 } }, - { "≏", { 8783, 0 } }, - { "Е", { 1045, 0 } }, - { "IJ", { 306, 0 } }, - { "Ё", { 1025, 0 } }, - { "Í", { 205, 0 } }, - { "Î", { 206, 0 } }, - { "И", { 1048, 0 } }, - { "İ", { 304, 0 } }, - { "ℑ", { 8465, 0 } }, - { "Ì", { 204, 0 } }, - { "ℑ", { 8465, 0 } }, - { "Ī", { 298, 0 } }, - { "ⅈ", { 8520, 0 } }, - { "⇒", { 8658, 0 } }, - { "∬", { 8748, 0 } }, - { "∫", { 8747, 0 } }, - { "⋂", { 8898, 0 } }, - { "⁣", { 8291, 0 } }, - { "⁢", { 8290, 0 } }, - { "Į", { 302, 0 } }, - { "𝕀", { 120128, 0 } }, - { "Ι", { 921, 0 } }, - { "ℐ", { 8464, 0 } }, - { "Ĩ", { 296, 0 } }, - { "І", { 1030, 0 } }, - { "Ï", { 207, 0 } }, - { "Ĵ", { 308, 0 } }, - { "Й", { 1049, 0 } }, - { "𝔍", { 120077, 0 } }, - { "𝕁", { 120129, 0 } }, - { "𝒥", { 119973, 0 } }, - { "Ј", { 1032, 0 } }, - { "Є", { 1028, 0 } }, - { "Х", { 1061, 0 } }, - { "Ќ", { 1036, 0 } }, - { "Κ", { 922, 0 } }, - { "Ķ", { 310, 0 } }, - { "К", { 1050, 0 } }, - { "𝔎", { 120078, 0 } }, - { "𝕂", { 120130, 0 } }, - { "𝒦", { 119974, 0 } }, - { "Љ", { 1033, 0 } }, - { "<", { 60, 0 } }, - { "Ĺ", { 313, 0 } }, - { "Λ", { 923, 0 } }, - { "⟪", { 10218, 0 } }, - { "ℒ", { 8466, 0 } }, - { "↞", { 8606, 0 } }, - { "Ľ", { 317, 0 } }, - { "Ļ", { 315, 0 } }, - { "Л", { 1051, 0 } }, - { "⟨", { 10216, 0 } }, - { "←", { 8592, 0 } }, - { "⇤", { 8676, 0 } }, - { "⇆", { 8646, 0 } }, - { "⌈", { 8968, 0 } }, - { "⟦", { 10214, 0 } }, - { "⥡", { 10593, 0 } }, - { "⇃", { 8643, 0 } }, - { "⥙", { 10585, 0 } }, - { "⌊", { 8970, 0 } }, - { "↔", { 8596, 0 } }, - { "⥎", { 10574, 0 } }, - { "⊣", { 8867, 0 } }, - { "↤", { 8612, 0 } }, - { "⥚", { 10586, 0 } }, - { "⊲", { 8882, 0 } }, - { "⧏", { 10703, 0 } }, - { "⊴", { 8884, 0 } }, - { "⥑", { 10577, 0 } }, - { "⥠", { 10592, 0 } }, - { "↿", { 8639, 0 } }, - { "⥘", { 10584, 0 } }, - { "↼", { 8636, 0 } }, - { "⥒", { 10578, 0 } }, - { "⇐", { 8656, 0 } }, - { "⇔", { 8660, 0 } }, - { "⋚", { 8922, 0 } }, - { "≦", { 8806, 0 } }, - { "≶", { 8822, 0 } }, - { "⪡", { 10913, 0 } }, - { "⩽", { 10877, 0 } }, - { "≲", { 8818, 0 } }, - { "𝔏", { 120079, 0 } }, - { "⋘", { 8920, 0 } }, - { "⇚", { 8666, 0 } }, - { "Ŀ", { 319, 0 } }, - { "⟵", { 10229, 0 } }, - { "⟷", { 10231, 0 } }, - { "⟶", { 10230, 0 } }, - { "⟸", { 10232, 0 } }, - { "⟺", { 10234, 0 } }, - { "⟹", { 10233, 0 } }, - { "𝕃", { 120131, 0 } }, - { "↙", { 8601, 0 } }, - { "↘", { 8600, 0 } }, - { "ℒ", { 8466, 0 } }, - { "↰", { 8624, 0 } }, - { "Ł", { 321, 0 } }, - { "≪", { 8810, 0 } }, - { "⤅", { 10501, 0 } }, - { "М", { 1052, 0 } }, - { " ", { 8287, 0 } }, - { "ℳ", { 8499, 0 } }, - { "𝔐", { 120080, 0 } }, - { "∓", { 8723, 0 } }, - { "𝕄", { 120132, 0 } }, - { "ℳ", { 8499, 0 } }, - { "Μ", { 924, 0 } }, - { "Њ", { 1034, 0 } }, - { "Ń", { 323, 0 } }, - { "Ň", { 327, 0 } }, - { "Ņ", { 325, 0 } }, - { "Н", { 1053, 0 } }, - { "​", { 8203, 0 } }, - { "​", { 8203, 0 } }, - { "​", { 8203, 0 } }, - { "​", { 8203, 0 } }, - { "≫", { 8811, 0 } }, - { "≪", { 8810, 0 } }, - { "
", { 10, 0 } }, - { "𝔑", { 120081, 0 } }, - { "⁠", { 8288, 0 } }, - { " ", { 160, 0 } }, - { "ℕ", { 8469, 0 } }, - { "⫬", { 10988, 0 } }, - { "≢", { 8802, 0 } }, - { "≭", { 8813, 0 } }, - { "∦", { 8742, 0 } }, - { "∉", { 8713, 0 } }, - { "≠", { 8800, 0 } }, - { "≂̸", { 8770, 824 } }, - { "∄", { 8708, 0 } }, - { "≯", { 8815, 0 } }, - { "≱", { 8817, 0 } }, - { "≧̸", { 8807, 824 } }, - { "≫̸", { 8811, 824 } }, - { "≹", { 8825, 0 } }, - { "⩾̸", { 10878, 824 } }, - { "≵", { 8821, 0 } }, - { "≎̸", { 8782, 824 } }, - { "≏̸", { 8783, 824 } }, - { "⋪", { 8938, 0 } }, - { "⧏̸", { 10703, 824 } }, - { "⋬", { 8940, 0 } }, - { "≮", { 8814, 0 } }, - { "≰", { 8816, 0 } }, - { "≸", { 8824, 0 } }, - { "≪̸", { 8810, 824 } }, - { "⩽̸", { 10877, 824 } }, - { "≴", { 8820, 0 } }, - { "⪢̸", { 10914, 824 } }, - { "⪡̸", { 10913, 824 } }, - { "⊀", { 8832, 0 } }, - { "⪯̸", { 10927, 824 } }, - { "⋠", { 8928, 0 } }, - { "∌", { 8716, 0 } }, - { "⋫", { 8939, 0 } }, - { "⧐̸", { 10704, 824 } }, - { "⋭", { 8941, 0 } }, - { "⊏̸", { 8847, 824 } }, - { "⋢", { 8930, 0 } }, - { "⊐̸", { 8848, 824 } }, - { "⋣", { 8931, 0 } }, - { "⊂⃒", { 8834, 8402 } }, - { "⊈", { 8840, 0 } }, - { "⊁", { 8833, 0 } }, - { "⪰̸", { 10928, 824 } }, - { "⋡", { 8929, 0 } }, - { "≿̸", { 8831, 824 } }, - { "⊃⃒", { 8835, 8402 } }, - { "⊉", { 8841, 0 } }, - { "≁", { 8769, 0 } }, - { "≄", { 8772, 0 } }, - { "≇", { 8775, 0 } }, - { "≉", { 8777, 0 } }, - { "∤", { 8740, 0 } }, - { "𝒩", { 119977, 0 } }, - { "Ñ", { 209, 0 } }, - { "Ν", { 925, 0 } }, - { "Œ", { 338, 0 } }, - { "Ó", { 211, 0 } }, - { "Ô", { 212, 0 } }, - { "О", { 1054, 0 } }, - { "Ő", { 336, 0 } }, - { "𝔒", { 120082, 0 } }, - { "Ò", { 210, 0 } }, - { "Ō", { 332, 0 } }, - { "Ω", { 937, 0 } }, - { "Ο", { 927, 0 } }, - { "𝕆", { 120134, 0 } }, - { "“", { 8220, 0 } }, - { "‘", { 8216, 0 } }, - { "⩔", { 10836, 0 } }, - { "𝒪", { 119978, 0 } }, - { "Ø", { 216, 0 } }, - { "Õ", { 213, 0 } }, - { "⨷", { 10807, 0 } }, - { "Ö", { 214, 0 } }, - { "‾", { 8254, 0 } }, - { "⏞", { 9182, 0 } }, - { "⎴", { 9140, 0 } }, - { "⏜", { 9180, 0 } }, - { "∂", { 8706, 0 } }, - { "П", { 1055, 0 } }, - { "𝔓", { 120083, 0 } }, - { "Φ", { 934, 0 } }, - { "Π", { 928, 0 } }, - { "±", { 177, 0 } }, - { "ℌ", { 8460, 0 } }, - { "ℙ", { 8473, 0 } }, - { "⪻", { 10939, 0 } }, - { "≺", { 8826, 0 } }, - { "⪯", { 10927, 0 } }, - { "≼", { 8828, 0 } }, - { "≾", { 8830, 0 } }, - { "″", { 8243, 0 } }, - { "∏", { 8719, 0 } }, - { "∷", { 8759, 0 } }, - { "∝", { 8733, 0 } }, - { "𝒫", { 119979, 0 } }, - { "Ψ", { 936, 0 } }, - { """, { 34, 0 } }, - { "𝔔", { 120084, 0 } }, - { "ℚ", { 8474, 0 } }, - { "𝒬", { 119980, 0 } }, - { "⤐", { 10512, 0 } }, - { "®", { 174, 0 } }, - { "Ŕ", { 340, 0 } }, - { "⟫", { 10219, 0 } }, - { "↠", { 8608, 0 } }, - { "⤖", { 10518, 0 } }, - { "Ř", { 344, 0 } }, - { "Ŗ", { 342, 0 } }, - { "Р", { 1056, 0 } }, - { "ℜ", { 8476, 0 } }, - { "∋", { 8715, 0 } }, - { "⇋", { 8651, 0 } }, - { "⥯", { 10607, 0 } }, - { "ℜ", { 8476, 0 } }, - { "Ρ", { 929, 0 } }, - { "⟩", { 10217, 0 } }, - { "→", { 8594, 0 } }, - { "⇥", { 8677, 0 } }, - { "⇄", { 8644, 0 } }, - { "⌉", { 8969, 0 } }, - { "⟧", { 10215, 0 } }, - { "⥝", { 10589, 0 } }, - { "⇂", { 8642, 0 } }, - { "⥕", { 10581, 0 } }, - { "⌋", { 8971, 0 } }, - { "⊢", { 8866, 0 } }, - { "↦", { 8614, 0 } }, - { "⥛", { 10587, 0 } }, - { "⊳", { 8883, 0 } }, - { "⧐", { 10704, 0 } }, - { "⊵", { 8885, 0 } }, - { "⥏", { 10575, 0 } }, - { "⥜", { 10588, 0 } }, - { "↾", { 8638, 0 } }, - { "⥔", { 10580, 0 } }, - { "⇀", { 8640, 0 } }, - { "⥓", { 10579, 0 } }, - { "⇒", { 8658, 0 } }, - { "ℝ", { 8477, 0 } }, - { "⥰", { 10608, 0 } }, - { "⇛", { 8667, 0 } }, - { "ℛ", { 8475, 0 } }, - { "↱", { 8625, 0 } }, - { "⧴", { 10740, 0 } }, - { "Щ", { 1065, 0 } }, - { "Ш", { 1064, 0 } }, - { "Ь", { 1068, 0 } }, - { "Ś", { 346, 0 } }, - { "⪼", { 10940, 0 } }, - { "Š", { 352, 0 } }, - { "Ş", { 350, 0 } }, - { "Ŝ", { 348, 0 } }, - { "С", { 1057, 0 } }, - { "𝔖", { 120086, 0 } }, - { "↓", { 8595, 0 } }, - { "←", { 8592, 0 } }, - { "→", { 8594, 0 } }, - { "↑", { 8593, 0 } }, - { "Σ", { 931, 0 } }, - { "∘", { 8728, 0 } }, - { "𝕊", { 120138, 0 } }, - { "√", { 8730, 0 } }, - { "□", { 9633, 0 } }, - { "⊓", { 8851, 0 } }, - { "⊏", { 8847, 0 } }, - { "⊑", { 8849, 0 } }, - { "⊐", { 8848, 0 } }, - { "⊒", { 8850, 0 } }, - { "⊔", { 8852, 0 } }, - { "𝒮", { 119982, 0 } }, - { "⋆", { 8902, 0 } }, - { "⋐", { 8912, 0 } }, - { "⋐", { 8912, 0 } }, - { "⊆", { 8838, 0 } }, - { "≻", { 8827, 0 } }, - { "⪰", { 10928, 0 } }, - { "≽", { 8829, 0 } }, - { "≿", { 8831, 0 } }, - { "∋", { 8715, 0 } }, - { "∑", { 8721, 0 } }, - { "⋑", { 8913, 0 } }, - { "⊃", { 8835, 0 } }, - { "⊇", { 8839, 0 } }, - { "⋑", { 8913, 0 } }, - { "Þ", { 222, 0 } }, - { "™", { 8482, 0 } }, - { "Ћ", { 1035, 0 } }, - { "Ц", { 1062, 0 } }, - { "	", { 9, 0 } }, - { "Τ", { 932, 0 } }, - { "Ť", { 356, 0 } }, - { "Ţ", { 354, 0 } }, - { "Т", { 1058, 0 } }, - { "𝔗", { 120087, 0 } }, - { "∴", { 8756, 0 } }, - { "Θ", { 920, 0 } }, - { "  ", { 8287, 8202 } }, - { " ", { 8201, 0 } }, - { "∼", { 8764, 0 } }, - { "≃", { 8771, 0 } }, - { "≅", { 8773, 0 } }, - { "≈", { 8776, 0 } }, - { "𝕋", { 120139, 0 } }, - { "⃛", { 8411, 0 } }, - { "𝒯", { 119983, 0 } }, - { "Ŧ", { 358, 0 } }, - { "Ú", { 218, 0 } }, - { "↟", { 8607, 0 } }, - { "⥉", { 10569, 0 } }, - { "Ў", { 1038, 0 } }, - { "Ŭ", { 364, 0 } }, - { "Û", { 219, 0 } }, - { "У", { 1059, 0 } }, - { "Ű", { 368, 0 } }, - { "𝔘", { 120088, 0 } }, - { "Ù", { 217, 0 } }, - { "Ū", { 362, 0 } }, - { "_", { 95, 0 } }, - { "⏟", { 9183, 0 } }, - { "⎵", { 9141, 0 } }, - { "⏝", { 9181, 0 } }, - { "⋃", { 8899, 0 } }, - { "⊎", { 8846, 0 } }, - { "Ų", { 370, 0 } }, - { "𝕌", { 120140, 0 } }, - { "↑", { 8593, 0 } }, - { "⤒", { 10514, 0 } }, - { "⇅", { 8645, 0 } }, - { "↕", { 8597, 0 } }, - { "⥮", { 10606, 0 } }, - { "⊥", { 8869, 0 } }, - { "↥", { 8613, 0 } }, - { "⇑", { 8657, 0 } }, - { "⇕", { 8661, 0 } }, - { "↖", { 8598, 0 } }, - { "↗", { 8599, 0 } }, - { "ϒ", { 978, 0 } }, - { "Υ", { 933, 0 } }, - { "Ů", { 366, 0 } }, - { "𝒰", { 119984, 0 } }, - { "Ũ", { 360, 0 } }, - { "Ü", { 220, 0 } }, - { "⊫", { 8875, 0 } }, - { "⫫", { 10987, 0 } }, - { "В", { 1042, 0 } }, - { "⊩", { 8873, 0 } }, - { "⫦", { 10982, 0 } }, - { "⋁", { 8897, 0 } }, - { "‖", { 8214, 0 } }, - { "‖", { 8214, 0 } }, - { "∣", { 8739, 0 } }, - { "|", { 124, 0 } }, - { "❘", { 10072, 0 } }, - { "≀", { 8768, 0 } }, - { " ", { 8202, 0 } }, - { "𝔙", { 120089, 0 } }, - { "𝕍", { 120141, 0 } }, - { "𝒱", { 119985, 0 } }, - { "⊪", { 8874, 0 } }, - { "Ŵ", { 372, 0 } }, - { "⋀", { 8896, 0 } }, - { "𝔚", { 120090, 0 } }, - { "𝕎", { 120142, 0 } }, - { "𝒲", { 119986, 0 } }, - { "𝔛", { 120091, 0 } }, - { "Ξ", { 926, 0 } }, - { "𝕏", { 120143, 0 } }, - { "𝒳", { 119987, 0 } }, - { "Я", { 1071, 0 } }, - { "Ї", { 1031, 0 } }, - { "Ю", { 1070, 0 } }, - { "Ý", { 221, 0 } }, - { "Ŷ", { 374, 0 } }, - { "Ы", { 1067, 0 } }, - { "𝔜", { 120092, 0 } }, - { "𝕐", { 120144, 0 } }, - { "𝒴", { 119988, 0 } }, - { "Ÿ", { 376, 0 } }, - { "Ж", { 1046, 0 } }, - { "Ź", { 377, 0 } }, - { "Ž", { 381, 0 } }, - { "З", { 1047, 0 } }, - { "Ż", { 379, 0 } }, - { "​", { 8203, 0 } }, - { "Ζ", { 918, 0 } }, - { "ℨ", { 8488, 0 } }, - { "ℤ", { 8484, 0 } }, - { "𝒵", { 119989, 0 } }, - { "á", { 225, 0 } }, - { "ă", { 259, 0 } }, - { "∾", { 8766, 0 } }, - { "∾̳", { 8766, 819 } }, - { "∿", { 8767, 0 } }, - { "â", { 226, 0 } }, - { "´", { 180, 0 } }, - { "а", { 1072, 0 } }, - { "æ", { 230, 0 } }, - { "⁡", { 8289, 0 } }, - { "𝔞", { 120094, 0 } }, - { "à", { 224, 0 } }, - { "ℵ", { 8501, 0 } }, - { "ℵ", { 8501, 0 } }, - { "α", { 945, 0 } }, - { "ā", { 257, 0 } }, - { "⨿", { 10815, 0 } }, - { "&", { 38, 0 } }, - { "∧", { 8743, 0 } }, - { "⩕", { 10837, 0 } }, - { "⩜", { 10844, 0 } }, - { "⩘", { 10840, 0 } }, - { "⩚", { 10842, 0 } }, - { "∠", { 8736, 0 } }, - { "⦤", { 10660, 0 } }, - { "∠", { 8736, 0 } }, - { "∡", { 8737, 0 } }, - { "⦨", { 10664, 0 } }, - { "⦩", { 10665, 0 } }, - { "⦪", { 10666, 0 } }, - { "⦫", { 10667, 0 } }, - { "⦬", { 10668, 0 } }, - { "⦭", { 10669, 0 } }, - { "⦮", { 10670, 0 } }, - { "⦯", { 10671, 0 } }, - { "∟", { 8735, 0 } }, - { "⊾", { 8894, 0 } }, - { "⦝", { 10653, 0 } }, - { "∢", { 8738, 0 } }, - { "Å", { 197, 0 } }, - { "⍼", { 9084, 0 } }, - { "ą", { 261, 0 } }, - { "𝕒", { 120146, 0 } }, - { "≈", { 8776, 0 } }, - { "⩰", { 10864, 0 } }, - { "⩯", { 10863, 0 } }, - { "≊", { 8778, 0 } }, - { "≋", { 8779, 0 } }, - { "'", { 39, 0 } }, - { "≈", { 8776, 0 } }, - { "≊", { 8778, 0 } }, - { "å", { 229, 0 } }, - { "𝒶", { 119990, 0 } }, - { "*", { 42, 0 } }, - { "≈", { 8776, 0 } }, - { "≍", { 8781, 0 } }, - { "ã", { 227, 0 } }, - { "ä", { 228, 0 } }, - { "∳", { 8755, 0 } }, - { "⨑", { 10769, 0 } }, - { "⫭", { 10989, 0 } }, - { "≌", { 8780, 0 } }, - { "϶", { 1014, 0 } }, - { "‵", { 8245, 0 } }, - { "∽", { 8765, 0 } }, - { "⋍", { 8909, 0 } }, - { "⊽", { 8893, 0 } }, - { "⌅", { 8965, 0 } }, - { "⌅", { 8965, 0 } }, - { "⎵", { 9141, 0 } }, - { "⎶", { 9142, 0 } }, - { "≌", { 8780, 0 } }, - { "б", { 1073, 0 } }, - { "„", { 8222, 0 } }, - { "∵", { 8757, 0 } }, - { "∵", { 8757, 0 } }, - { "⦰", { 10672, 0 } }, - { "϶", { 1014, 0 } }, - { "ℬ", { 8492, 0 } }, - { "β", { 946, 0 } }, - { "ℶ", { 8502, 0 } }, - { "≬", { 8812, 0 } }, - { "𝔟", { 120095, 0 } }, - { "⋂", { 8898, 0 } }, - { "◯", { 9711, 0 } }, - { "⋃", { 8899, 0 } }, - { "⨀", { 10752, 0 } }, - { "⨁", { 10753, 0 } }, - { "⨂", { 10754, 0 } }, - { "⨆", { 10758, 0 } }, - { "★", { 9733, 0 } }, - { "▽", { 9661, 0 } }, - { "△", { 9651, 0 } }, - { "⨄", { 10756, 0 } }, - { "⋁", { 8897, 0 } }, - { "⋀", { 8896, 0 } }, - { "⤍", { 10509, 0 } }, - { "⧫", { 10731, 0 } }, - { "▪", { 9642, 0 } }, - { "▴", { 9652, 0 } }, - { "▾", { 9662, 0 } }, - { "◂", { 9666, 0 } }, - { "▸", { 9656, 0 } }, - { "␣", { 9251, 0 } }, - { "▒", { 9618, 0 } }, - { "░", { 9617, 0 } }, - { "▓", { 9619, 0 } }, - { "█", { 9608, 0 } }, - { "=⃥", { 61, 8421 } }, - { "≡⃥", { 8801, 8421 } }, - { "⌐", { 8976, 0 } }, - { "𝕓", { 120147, 0 } }, - { "⊥", { 8869, 0 } }, - { "⊥", { 8869, 0 } }, - { "⋈", { 8904, 0 } }, - { "╗", { 9559, 0 } }, - { "╔", { 9556, 0 } }, - { "╖", { 9558, 0 } }, - { "╓", { 9555, 0 } }, - { "═", { 9552, 0 } }, - { "╦", { 9574, 0 } }, - { "╩", { 9577, 0 } }, - { "╤", { 9572, 0 } }, - { "╧", { 9575, 0 } }, - { "╝", { 9565, 0 } }, - { "╚", { 9562, 0 } }, - { "╜", { 9564, 0 } }, - { "╙", { 9561, 0 } }, - { "║", { 9553, 0 } }, - { "╬", { 9580, 0 } }, - { "╣", { 9571, 0 } }, - { "╠", { 9568, 0 } }, - { "╫", { 9579, 0 } }, - { "╢", { 9570, 0 } }, - { "╟", { 9567, 0 } }, - { "⧉", { 10697, 0 } }, - { "╕", { 9557, 0 } }, - { "╒", { 9554, 0 } }, - { "┐", { 9488, 0 } }, - { "┌", { 9484, 0 } }, - { "─", { 9472, 0 } }, - { "╥", { 9573, 0 } }, - { "╨", { 9576, 0 } }, - { "┬", { 9516, 0 } }, - { "┴", { 9524, 0 } }, - { "⊟", { 8863, 0 } }, - { "⊞", { 8862, 0 } }, - { "⊠", { 8864, 0 } }, - { "╛", { 9563, 0 } }, - { "╘", { 9560, 0 } }, - { "┘", { 9496, 0 } }, - { "└", { 9492, 0 } }, - { "│", { 9474, 0 } }, - { "╪", { 9578, 0 } }, - { "╡", { 9569, 0 } }, - { "╞", { 9566, 0 } }, - { "┼", { 9532, 0 } }, - { "┤", { 9508, 0 } }, - { "├", { 9500, 0 } }, - { "‵", { 8245, 0 } }, - { "˘", { 728, 0 } }, - { "¦", { 166, 0 } }, - { "𝒷", { 119991, 0 } }, - { "⁏", { 8271, 0 } }, - { "∽", { 8765, 0 } }, - { "⋍", { 8909, 0 } }, - { "\", { 92, 0 } }, - { "⧅", { 10693, 0 } }, - { "⟈", { 10184, 0 } }, - { "•", { 8226, 0 } }, - { "•", { 8226, 0 } }, - { "≎", { 8782, 0 } }, - { "⪮", { 10926, 0 } }, - { "≏", { 8783, 0 } }, - { "≏", { 8783, 0 } }, - { "ć", { 263, 0 } }, - { "∩", { 8745, 0 } }, - { "⩄", { 10820, 0 } }, - { "⩉", { 10825, 0 } }, - { "⩋", { 10827, 0 } }, - { "⩇", { 10823, 0 } }, - { "⩀", { 10816, 0 } }, - { "∩︀", { 8745, 65024 } }, - { "⁁", { 8257, 0 } }, - { "ˇ", { 711, 0 } }, - { "⩍", { 10829, 0 } }, - { "č", { 269, 0 } }, - { "ç", { 231, 0 } }, - { "ĉ", { 265, 0 } }, - { "⩌", { 10828, 0 } }, - { "⩐", { 10832, 0 } }, - { "ċ", { 267, 0 } }, - { "¸", { 184, 0 } }, - { "⦲", { 10674, 0 } }, - { "¢", { 162, 0 } }, - { "·", { 183, 0 } }, - { "𝔠", { 120096, 0 } }, - { "ч", { 1095, 0 } }, - { "✓", { 10003, 0 } }, - { "✓", { 10003, 0 } }, - { "χ", { 967, 0 } }, - { "○", { 9675, 0 } }, - { "⧃", { 10691, 0 } }, - { "ˆ", { 710, 0 } }, - { "≗", { 8791, 0 } }, - { "↺", { 8634, 0 } }, - { "↻", { 8635, 0 } }, - { "®", { 174, 0 } }, - { "Ⓢ", { 9416, 0 } }, - { "⊛", { 8859, 0 } }, - { "⊚", { 8858, 0 } }, - { "⊝", { 8861, 0 } }, - { "≗", { 8791, 0 } }, - { "⨐", { 10768, 0 } }, - { "⫯", { 10991, 0 } }, - { "⧂", { 10690, 0 } }, - { "♣", { 9827, 0 } }, - { "♣", { 9827, 0 } }, - { ":", { 58, 0 } }, - { "≔", { 8788, 0 } }, - { "≔", { 8788, 0 } }, - { ",", { 44, 0 } }, - { "@", { 64, 0 } }, - { "∁", { 8705, 0 } }, - { "∘", { 8728, 0 } }, - { "∁", { 8705, 0 } }, - { "ℂ", { 8450, 0 } }, - { "≅", { 8773, 0 } }, - { "⩭", { 10861, 0 } }, - { "∮", { 8750, 0 } }, - { "𝕔", { 120148, 0 } }, - { "∐", { 8720, 0 } }, - { "©", { 169, 0 } }, - { "℗", { 8471, 0 } }, - { "↵", { 8629, 0 } }, - { "✗", { 10007, 0 } }, - { "𝒸", { 119992, 0 } }, - { "⫏", { 10959, 0 } }, - { "⫑", { 10961, 0 } }, - { "⫐", { 10960, 0 } }, - { "⫒", { 10962, 0 } }, - { "⋯", { 8943, 0 } }, - { "⤸", { 10552, 0 } }, - { "⤵", { 10549, 0 } }, - { "⋞", { 8926, 0 } }, - { "⋟", { 8927, 0 } }, - { "↶", { 8630, 0 } }, - { "⤽", { 10557, 0 } }, - { "∪", { 8746, 0 } }, - { "⩈", { 10824, 0 } }, - { "⩆", { 10822, 0 } }, - { "⩊", { 10826, 0 } }, - { "⊍", { 8845, 0 } }, - { "⩅", { 10821, 0 } }, - { "∪︀", { 8746, 65024 } }, - { "↷", { 8631, 0 } }, - { "⤼", { 10556, 0 } }, - { "⋞", { 8926, 0 } }, - { "⋟", { 8927, 0 } }, - { "⋎", { 8910, 0 } }, - { "⋏", { 8911, 0 } }, - { "¤", { 164, 0 } }, - { "↶", { 8630, 0 } }, - { "↷", { 8631, 0 } }, - { "⋎", { 8910, 0 } }, - { "⋏", { 8911, 0 } }, - { "∲", { 8754, 0 } }, - { "∱", { 8753, 0 } }, - { "⌭", { 9005, 0 } }, - { "⇓", { 8659, 0 } }, - { "⥥", { 10597, 0 } }, - { "†", { 8224, 0 } }, - { "ℸ", { 8504, 0 } }, - { "↓", { 8595, 0 } }, - { "‐", { 8208, 0 } }, - { "⊣", { 8867, 0 } }, - { "⤏", { 10511, 0 } }, - { "˝", { 733, 0 } }, - { "ď", { 271, 0 } }, - { "д", { 1076, 0 } }, - { "ⅆ", { 8518, 0 } }, - { "‡", { 8225, 0 } }, - { "⇊", { 8650, 0 } }, - { "⩷", { 10871, 0 } }, - { "°", { 176, 0 } }, - { "δ", { 948, 0 } }, - { "⦱", { 10673, 0 } }, - { "⥿", { 10623, 0 } }, - { "𝔡", { 120097, 0 } }, - { "⇃", { 8643, 0 } }, - { "⇂", { 8642, 0 } }, - { "⋄", { 8900, 0 } }, - { "⋄", { 8900, 0 } }, - { "♦", { 9830, 0 } }, - { "♦", { 9830, 0 } }, - { "¨", { 168, 0 } }, - { "ϝ", { 989, 0 } }, - { "⋲", { 8946, 0 } }, - { "÷", { 247, 0 } }, - { "÷", { 247, 0 } }, - { "⋇", { 8903, 0 } }, - { "⋇", { 8903, 0 } }, - { "ђ", { 1106, 0 } }, - { "⌞", { 8990, 0 } }, - { "⌍", { 8973, 0 } }, - { "$", { 36, 0 } }, - { "𝕕", { 120149, 0 } }, - { "˙", { 729, 0 } }, - { "≐", { 8784, 0 } }, - { "≑", { 8785, 0 } }, - { "∸", { 8760, 0 } }, - { "∔", { 8724, 0 } }, - { "⊡", { 8865, 0 } }, - { "⌆", { 8966, 0 } }, - { "↓", { 8595, 0 } }, - { "⇊", { 8650, 0 } }, - { "⇃", { 8643, 0 } }, - { "⇂", { 8642, 0 } }, - { "⤐", { 10512, 0 } }, - { "⌟", { 8991, 0 } }, - { "⌌", { 8972, 0 } }, - { "𝒹", { 119993, 0 } }, - { "ѕ", { 1109, 0 } }, - { "⧶", { 10742, 0 } }, - { "đ", { 273, 0 } }, - { "⋱", { 8945, 0 } }, - { "▿", { 9663, 0 } }, - { "▾", { 9662, 0 } }, - { "⇵", { 8693, 0 } }, - { "⥯", { 10607, 0 } }, - { "⦦", { 10662, 0 } }, - { "џ", { 1119, 0 } }, - { "⟿", { 10239, 0 } }, - { "⩷", { 10871, 0 } }, - { "≑", { 8785, 0 } }, - { "é", { 233, 0 } }, - { "⩮", { 10862, 0 } }, - { "ě", { 283, 0 } }, - { "≖", { 8790, 0 } }, - { "ê", { 234, 0 } }, - { "≕", { 8789, 0 } }, - { "э", { 1101, 0 } }, - { "ė", { 279, 0 } }, - { "ⅇ", { 8519, 0 } }, - { "≒", { 8786, 0 } }, - { "𝔢", { 120098, 0 } }, - { "⪚", { 10906, 0 } }, - { "è", { 232, 0 } }, - { "⪖", { 10902, 0 } }, - { "⪘", { 10904, 0 } }, - { "⪙", { 10905, 0 } }, - { "⏧", { 9191, 0 } }, - { "ℓ", { 8467, 0 } }, - { "⪕", { 10901, 0 } }, - { "⪗", { 10903, 0 } }, - { "ē", { 275, 0 } }, - { "∅", { 8709, 0 } }, - { "∅", { 8709, 0 } }, - { "∅", { 8709, 0 } }, - { " ", { 8196, 0 } }, - { " ", { 8197, 0 } }, - { " ", { 8195, 0 } }, - { "ŋ", { 331, 0 } }, - { " ", { 8194, 0 } }, - { "ę", { 281, 0 } }, - { "𝕖", { 120150, 0 } }, - { "⋕", { 8917, 0 } }, - { "⧣", { 10723, 0 } }, - { "⩱", { 10865, 0 } }, - { "ε", { 949, 0 } }, - { "ε", { 949, 0 } }, - { "ϵ", { 1013, 0 } }, - { "≖", { 8790, 0 } }, - { "≕", { 8789, 0 } }, - { "≂", { 8770, 0 } }, - { "⪖", { 10902, 0 } }, - { "⪕", { 10901, 0 } }, - { "=", { 61, 0 } }, - { "≟", { 8799, 0 } }, - { "≡", { 8801, 0 } }, - { "⩸", { 10872, 0 } }, - { "⧥", { 10725, 0 } }, - { "≓", { 8787, 0 } }, - { "⥱", { 10609, 0 } }, - { "ℯ", { 8495, 0 } }, - { "≐", { 8784, 0 } }, - { "≂", { 8770, 0 } }, - { "η", { 951, 0 } }, - { "ð", { 240, 0 } }, - { "ë", { 235, 0 } }, - { "€", { 8364, 0 } }, - { "!", { 33, 0 } }, - { "∃", { 8707, 0 } }, - { "ℰ", { 8496, 0 } }, - { "ⅇ", { 8519, 0 } }, - { "≒", { 8786, 0 } }, - { "ф", { 1092, 0 } }, - { "♀", { 9792, 0 } }, - { "ffi", { 64259, 0 } }, - { "ff", { 64256, 0 } }, - { "ffl", { 64260, 0 } }, - { "𝔣", { 120099, 0 } }, - { "fi", { 64257, 0 } }, - { "fj", { 102, 106 } }, - { "♭", { 9837, 0 } }, - { "fl", { 64258, 0 } }, - { "▱", { 9649, 0 } }, - { "ƒ", { 402, 0 } }, - { "𝕗", { 120151, 0 } }, - { "∀", { 8704, 0 } }, - { "⋔", { 8916, 0 } }, - { "⫙", { 10969, 0 } }, - { "⨍", { 10765, 0 } }, - { "½", { 189, 0 } }, - { "½", { 189, 0 } }, - { "⅓", { 8531, 0 } }, - { "¼", { 188, 0 } }, - { "¼", { 188, 0 } }, - { "⅕", { 8533, 0 } }, - { "⅙", { 8537, 0 } }, - { "⅛", { 8539, 0 } }, - { "⅔", { 8532, 0 } }, - { "⅖", { 8534, 0 } }, - { "¾", { 190, 0 } }, - { "¾", { 190, 0 } }, - { "⅗", { 8535, 0 } }, - { "⅜", { 8540, 0 } }, - { "⅘", { 8536, 0 } }, - { "⅚", { 8538, 0 } }, - { "⅝", { 8541, 0 } }, - { "⅞", { 8542, 0 } }, - { "⁄", { 8260, 0 } }, - { "⌢", { 8994, 0 } }, - { "𝒻", { 119995, 0 } }, - { "≧", { 8807, 0 } }, - { "⪌", { 10892, 0 } }, - { "ǵ", { 501, 0 } }, - { "γ", { 947, 0 } }, - { "ϝ", { 989, 0 } }, - { "⪆", { 10886, 0 } }, - { "ğ", { 287, 0 } }, - { "ĝ", { 285, 0 } }, - { "г", { 1075, 0 } }, - { "ġ", { 289, 0 } }, - { "≥", { 8805, 0 } }, - { "⋛", { 8923, 0 } }, - { "≥", { 8805, 0 } }, - { "≧", { 8807, 0 } }, - { "⩾", { 10878, 0 } }, - { "⩾", { 10878, 0 } }, - { "⪩", { 10921, 0 } }, - { "⪀", { 10880, 0 } }, - { "⪂", { 10882, 0 } }, - { "⪄", { 10884, 0 } }, - { "⋛︀", { 8923, 65024 } }, - { "⪔", { 10900, 0 } }, - { "𝔤", { 120100, 0 } }, - { "≫", { 8811, 0 } }, - { "⋙", { 8921, 0 } }, - { "ℷ", { 8503, 0 } }, - { "ѓ", { 1107, 0 } }, - { "≷", { 8823, 0 } }, - { "⪒", { 10898, 0 } }, - { "⪥", { 10917, 0 } }, - { "⪤", { 10916, 0 } }, - { "≩", { 8809, 0 } }, - { "⪊", { 10890, 0 } }, - { "⪊", { 10890, 0 } }, - { "⪈", { 10888, 0 } }, - { "⪈", { 10888, 0 } }, - { "≩", { 8809, 0 } }, - { "⋧", { 8935, 0 } }, - { "𝕘", { 120152, 0 } }, - { "`", { 96, 0 } }, - { "ℊ", { 8458, 0 } }, - { "≳", { 8819, 0 } }, - { "⪎", { 10894, 0 } }, - { "⪐", { 10896, 0 } }, - { ">", { 62, 0 } }, - { "⪧", { 10919, 0 } }, - { "⩺", { 10874, 0 } }, - { "⋗", { 8919, 0 } }, - { "⦕", { 10645, 0 } }, - { "⩼", { 10876, 0 } }, - { "⪆", { 10886, 0 } }, - { "⥸", { 10616, 0 } }, - { "⋗", { 8919, 0 } }, - { "⋛", { 8923, 0 } }, - { "⪌", { 10892, 0 } }, - { "≷", { 8823, 0 } }, - { "≳", { 8819, 0 } }, - { "≩︀", { 8809, 65024 } }, - { "≩︀", { 8809, 65024 } }, - { "⇔", { 8660, 0 } }, - { " ", { 8202, 0 } }, - { "½", { 189, 0 } }, - { "ℋ", { 8459, 0 } }, - { "ъ", { 1098, 0 } }, - { "↔", { 8596, 0 } }, - { "⥈", { 10568, 0 } }, - { "↭", { 8621, 0 } }, - { "ℏ", { 8463, 0 } }, - { "ĥ", { 293, 0 } }, - { "♥", { 9829, 0 } }, - { "♥", { 9829, 0 } }, - { "…", { 8230, 0 } }, - { "⊹", { 8889, 0 } }, - { "𝔥", { 120101, 0 } }, - { "⤥", { 10533, 0 } }, - { "⤦", { 10534, 0 } }, - { "⇿", { 8703, 0 } }, - { "∻", { 8763, 0 } }, - { "↩", { 8617, 0 } }, - { "↪", { 8618, 0 } }, - { "𝕙", { 120153, 0 } }, - { "―", { 8213, 0 } }, - { "𝒽", { 119997, 0 } }, - { "ℏ", { 8463, 0 } }, - { "ħ", { 295, 0 } }, - { "⁃", { 8259, 0 } }, - { "‐", { 8208, 0 } }, - { "í", { 237, 0 } }, - { "⁣", { 8291, 0 } }, - { "î", { 238, 0 } }, - { "и", { 1080, 0 } }, - { "е", { 1077, 0 } }, - { "¡", { 161, 0 } }, - { "⇔", { 8660, 0 } }, - { "𝔦", { 120102, 0 } }, - { "ì", { 236, 0 } }, - { "ⅈ", { 8520, 0 } }, - { "⨌", { 10764, 0 } }, - { "∭", { 8749, 0 } }, - { "⧜", { 10716, 0 } }, - { "℩", { 8489, 0 } }, - { "ij", { 307, 0 } }, - { "ī", { 299, 0 } }, - { "ℑ", { 8465, 0 } }, - { "ℐ", { 8464, 0 } }, - { "ℑ", { 8465, 0 } }, - { "ı", { 305, 0 } }, - { "⊷", { 8887, 0 } }, - { "Ƶ", { 437, 0 } }, - { "∈", { 8712, 0 } }, - { "℅", { 8453, 0 } }, - { "∞", { 8734, 0 } }, - { "⧝", { 10717, 0 } }, - { "ı", { 305, 0 } }, - { "∫", { 8747, 0 } }, - { "⊺", { 8890, 0 } }, - { "ℤ", { 8484, 0 } }, - { "⊺", { 8890, 0 } }, - { "⨗", { 10775, 0 } }, - { "⨼", { 10812, 0 } }, - { "ё", { 1105, 0 } }, - { "į", { 303, 0 } }, - { "𝕚", { 120154, 0 } }, - { "ι", { 953, 0 } }, - { "⨼", { 10812, 0 } }, - { "¿", { 191, 0 } }, - { "𝒾", { 119998, 0 } }, - { "∈", { 8712, 0 } }, - { "⋹", { 8953, 0 } }, - { "⋵", { 8949, 0 } }, - { "⋴", { 8948, 0 } }, - { "⋳", { 8947, 0 } }, - { "∈", { 8712, 0 } }, - { "⁢", { 8290, 0 } }, - { "ĩ", { 297, 0 } }, - { "і", { 1110, 0 } }, - { "ï", { 239, 0 } }, - { "ĵ", { 309, 0 } }, - { "й", { 1081, 0 } }, - { "𝔧", { 120103, 0 } }, - { "ȷ", { 567, 0 } }, - { "𝕛", { 120155, 0 } }, - { "𝒿", { 119999, 0 } }, - { "ј", { 1112, 0 } }, - { "є", { 1108, 0 } }, - { "κ", { 954, 0 } }, - { "ϰ", { 1008, 0 } }, - { "ķ", { 311, 0 } }, - { "к", { 1082, 0 } }, - { "𝔨", { 120104, 0 } }, - { "ĸ", { 312, 0 } }, - { "х", { 1093, 0 } }, - { "ќ", { 1116, 0 } }, - { "𝕜", { 120156, 0 } }, - { "𝓀", { 120000, 0 } }, - { "⇚", { 8666, 0 } }, - { "⇐", { 8656, 0 } }, - { "⤛", { 10523, 0 } }, - { "⤎", { 10510, 0 } }, - { "≦", { 8806, 0 } }, - { "⪋", { 10891, 0 } }, - { "⥢", { 10594, 0 } }, - { "ĺ", { 314, 0 } }, - { "⦴", { 10676, 0 } }, - { "ℒ", { 8466, 0 } }, - { "λ", { 955, 0 } }, - { "⟨", { 10216, 0 } }, - { "⦑", { 10641, 0 } }, - { "⟨", { 10216, 0 } }, - { "⪅", { 10885, 0 } }, - { "«", { 171, 0 } }, - { "←", { 8592, 0 } }, - { "⇤", { 8676, 0 } }, - { "⤟", { 10527, 0 } }, - { "⤝", { 10525, 0 } }, - { "↩", { 8617, 0 } }, - { "↫", { 8619, 0 } }, - { "⤹", { 10553, 0 } }, - { "⥳", { 10611, 0 } }, - { "↢", { 8610, 0 } }, - { "⪫", { 10923, 0 } }, - { "⤙", { 10521, 0 } }, - { "⪭", { 10925, 0 } }, - { "⪭︀", { 10925, 65024 } }, - { "⤌", { 10508, 0 } }, - { "❲", { 10098, 0 } }, - { "{", { 123, 0 } }, - { "[", { 91, 0 } }, - { "⦋", { 10635, 0 } }, - { "⦏", { 10639, 0 } }, - { "⦍", { 10637, 0 } }, - { "ľ", { 318, 0 } }, - { "ļ", { 316, 0 } }, - { "⌈", { 8968, 0 } }, - { "{", { 123, 0 } }, - { "л", { 1083, 0 } }, - { "⤶", { 10550, 0 } }, - { "“", { 8220, 0 } }, - { "„", { 8222, 0 } }, - { "⥧", { 10599, 0 } }, - { "⥋", { 10571, 0 } }, - { "↲", { 8626, 0 } }, - { "≤", { 8804, 0 } }, - { "←", { 8592, 0 } }, - { "↢", { 8610, 0 } }, - { "↽", { 8637, 0 } }, - { "↼", { 8636, 0 } }, - { "⇇", { 8647, 0 } }, - { "↔", { 8596, 0 } }, - { "⇆", { 8646, 0 } }, - { "⇋", { 8651, 0 } }, - { "↭", { 8621, 0 } }, - { "⋋", { 8907, 0 } }, - { "⋚", { 8922, 0 } }, - { "≤", { 8804, 0 } }, - { "≦", { 8806, 0 } }, - { "⩽", { 10877, 0 } }, - { "⩽", { 10877, 0 } }, - { "⪨", { 10920, 0 } }, - { "⩿", { 10879, 0 } }, - { "⪁", { 10881, 0 } }, - { "⪃", { 10883, 0 } }, - { "⋚︀", { 8922, 65024 } }, - { "⪓", { 10899, 0 } }, - { "⪅", { 10885, 0 } }, - { "⋖", { 8918, 0 } }, - { "⋚", { 8922, 0 } }, - { "⪋", { 10891, 0 } }, - { "≶", { 8822, 0 } }, - { "≲", { 8818, 0 } }, - { "⥼", { 10620, 0 } }, - { "⌊", { 8970, 0 } }, - { "𝔩", { 120105, 0 } }, - { "≶", { 8822, 0 } }, - { "⪑", { 10897, 0 } }, - { "↽", { 8637, 0 } }, - { "↼", { 8636, 0 } }, - { "⥪", { 10602, 0 } }, - { "▄", { 9604, 0 } }, - { "љ", { 1113, 0 } }, - { "≪", { 8810, 0 } }, - { "⇇", { 8647, 0 } }, - { "⌞", { 8990, 0 } }, - { "⥫", { 10603, 0 } }, - { "◺", { 9722, 0 } }, - { "ŀ", { 320, 0 } }, - { "⎰", { 9136, 0 } }, - { "⎰", { 9136, 0 } }, - { "≨", { 8808, 0 } }, - { "⪉", { 10889, 0 } }, - { "⪉", { 10889, 0 } }, - { "⪇", { 10887, 0 } }, - { "⪇", { 10887, 0 } }, - { "≨", { 8808, 0 } }, - { "⋦", { 8934, 0 } }, - { "⟬", { 10220, 0 } }, - { "⇽", { 8701, 0 } }, - { "⟦", { 10214, 0 } }, - { "⟵", { 10229, 0 } }, - { "⟷", { 10231, 0 } }, - { "⟼", { 10236, 0 } }, - { "⟶", { 10230, 0 } }, - { "↫", { 8619, 0 } }, - { "↬", { 8620, 0 } }, - { "⦅", { 10629, 0 } }, - { "𝕝", { 120157, 0 } }, - { "⨭", { 10797, 0 } }, - { "⨴", { 10804, 0 } }, - { "∗", { 8727, 0 } }, - { "_", { 95, 0 } }, - { "◊", { 9674, 0 } }, - { "◊", { 9674, 0 } }, - { "⧫", { 10731, 0 } }, - { "(", { 40, 0 } }, - { "⦓", { 10643, 0 } }, - { "⇆", { 8646, 0 } }, - { "⌟", { 8991, 0 } }, - { "⇋", { 8651, 0 } }, - { "⥭", { 10605, 0 } }, - { "‎", { 8206, 0 } }, - { "⊿", { 8895, 0 } }, - { "‹", { 8249, 0 } }, - { "𝓁", { 120001, 0 } }, - { "↰", { 8624, 0 } }, - { "≲", { 8818, 0 } }, - { "⪍", { 10893, 0 } }, - { "⪏", { 10895, 0 } }, - { "[", { 91, 0 } }, - { "‘", { 8216, 0 } }, - { "‚", { 8218, 0 } }, - { "ł", { 322, 0 } }, - { "<", { 60, 0 } }, - { "⪦", { 10918, 0 } }, - { "⩹", { 10873, 0 } }, - { "⋖", { 8918, 0 } }, - { "⋋", { 8907, 0 } }, - { "⋉", { 8905, 0 } }, - { "⥶", { 10614, 0 } }, - { "⩻", { 10875, 0 } }, - { "⦖", { 10646, 0 } }, - { "◃", { 9667, 0 } }, - { "⊴", { 8884, 0 } }, - { "◂", { 9666, 0 } }, - { "⥊", { 10570, 0 } }, - { "⥦", { 10598, 0 } }, - { "≨︀", { 8808, 65024 } }, - { "≨︀", { 8808, 65024 } }, - { "∺", { 8762, 0 } }, - { "¯", { 175, 0 } }, - { "♂", { 9794, 0 } }, - { "✠", { 10016, 0 } }, - { "✠", { 10016, 0 } }, - { "↦", { 8614, 0 } }, - { "↦", { 8614, 0 } }, - { "↧", { 8615, 0 } }, - { "↤", { 8612, 0 } }, - { "↥", { 8613, 0 } }, - { "▮", { 9646, 0 } }, - { "⨩", { 10793, 0 } }, - { "м", { 1084, 0 } }, - { "—", { 8212, 0 } }, - { "∡", { 8737, 0 } }, - { "𝔪", { 120106, 0 } }, - { "℧", { 8487, 0 } }, - { "µ", { 181, 0 } }, - { "∣", { 8739, 0 } }, - { "*", { 42, 0 } }, - { "⫰", { 10992, 0 } }, - { "·", { 183, 0 } }, - { "−", { 8722, 0 } }, - { "⊟", { 8863, 0 } }, - { "∸", { 8760, 0 } }, - { "⨪", { 10794, 0 } }, - { "⫛", { 10971, 0 } }, - { "…", { 8230, 0 } }, - { "∓", { 8723, 0 } }, - { "⊧", { 8871, 0 } }, - { "𝕞", { 120158, 0 } }, - { "∓", { 8723, 0 } }, - { "𝓂", { 120002, 0 } }, - { "∾", { 8766, 0 } }, - { "μ", { 956, 0 } }, - { "⊸", { 8888, 0 } }, - { "⊸", { 8888, 0 } }, - { "⋙̸", { 8921, 824 } }, - { "≫⃒", { 8811, 8402 } }, - { "≫̸", { 8811, 824 } }, - { "⇍", { 8653, 0 } }, - { "⇎", { 8654, 0 } }, - { "⋘̸", { 8920, 824 } }, - { "≪⃒", { 8810, 8402 } }, - { "≪̸", { 8810, 824 } }, - { "⇏", { 8655, 0 } }, - { "⊯", { 8879, 0 } }, - { "⊮", { 8878, 0 } }, - { "∇", { 8711, 0 } }, - { "ń", { 324, 0 } }, - { "∠⃒", { 8736, 8402 } }, - { "≉", { 8777, 0 } }, - { "⩰̸", { 10864, 824 } }, - { "≋̸", { 8779, 824 } }, - { "ʼn", { 329, 0 } }, - { "≉", { 8777, 0 } }, - { "♮", { 9838, 0 } }, - { "♮", { 9838, 0 } }, - { "ℕ", { 8469, 0 } }, - { " ", { 160, 0 } }, - { "≎̸", { 8782, 824 } }, - { "≏̸", { 8783, 824 } }, - { "⩃", { 10819, 0 } }, - { "ň", { 328, 0 } }, - { "ņ", { 326, 0 } }, - { "≇", { 8775, 0 } }, - { "⩭̸", { 10861, 824 } }, - { "⩂", { 10818, 0 } }, - { "н", { 1085, 0 } }, - { "–", { 8211, 0 } }, - { "≠", { 8800, 0 } }, - { "⇗", { 8663, 0 } }, - { "⤤", { 10532, 0 } }, - { "↗", { 8599, 0 } }, - { "↗", { 8599, 0 } }, - { "≐̸", { 8784, 824 } }, - { "≢", { 8802, 0 } }, - { "⤨", { 10536, 0 } }, - { "≂̸", { 8770, 824 } }, - { "∄", { 8708, 0 } }, - { "∄", { 8708, 0 } }, - { "𝔫", { 120107, 0 } }, - { "≧̸", { 8807, 824 } }, - { "≱", { 8817, 0 } }, - { "≱", { 8817, 0 } }, - { "≧̸", { 8807, 824 } }, - { "⩾̸", { 10878, 824 } }, - { "⩾̸", { 10878, 824 } }, - { "≵", { 8821, 0 } }, - { "≯", { 8815, 0 } }, - { "≯", { 8815, 0 } }, - { "⇎", { 8654, 0 } }, - { "↮", { 8622, 0 } }, - { "⫲", { 10994, 0 } }, - { "∋", { 8715, 0 } }, - { "⋼", { 8956, 0 } }, - { "⋺", { 8954, 0 } }, - { "∋", { 8715, 0 } }, - { "њ", { 1114, 0 } }, - { "⇍", { 8653, 0 } }, - { "≦̸", { 8806, 824 } }, - { "↚", { 8602, 0 } }, - { "‥", { 8229, 0 } }, - { "≰", { 8816, 0 } }, - { "↚", { 8602, 0 } }, - { "↮", { 8622, 0 } }, - { "≰", { 8816, 0 } }, - { "≦̸", { 8806, 824 } }, - { "⩽̸", { 10877, 824 } }, - { "⩽̸", { 10877, 824 } }, - { "≮", { 8814, 0 } }, - { "≴", { 8820, 0 } }, - { "≮", { 8814, 0 } }, - { "⋪", { 8938, 0 } }, - { "⋬", { 8940, 0 } }, - { "∤", { 8740, 0 } }, - { "𝕟", { 120159, 0 } }, - { "¬", { 172, 0 } }, - { "∉", { 8713, 0 } }, - { "⋹̸", { 8953, 824 } }, - { "⋵̸", { 8949, 824 } }, - { "∉", { 8713, 0 } }, - { "⋷", { 8951, 0 } }, - { "⋶", { 8950, 0 } }, - { "∌", { 8716, 0 } }, - { "∌", { 8716, 0 } }, - { "⋾", { 8958, 0 } }, - { "⋽", { 8957, 0 } }, - { "∦", { 8742, 0 } }, - { "∦", { 8742, 0 } }, - { "⫽⃥", { 11005, 8421 } }, - { "∂̸", { 8706, 824 } }, - { "⨔", { 10772, 0 } }, - { "⊀", { 8832, 0 } }, - { "⋠", { 8928, 0 } }, - { "⪯̸", { 10927, 824 } }, - { "⊀", { 8832, 0 } }, - { "⪯̸", { 10927, 824 } }, - { "⇏", { 8655, 0 } }, - { "↛", { 8603, 0 } }, - { "⤳̸", { 10547, 824 } }, - { "↝̸", { 8605, 824 } }, - { "↛", { 8603, 0 } }, - { "⋫", { 8939, 0 } }, - { "⋭", { 8941, 0 } }, - { "⊁", { 8833, 0 } }, - { "⋡", { 8929, 0 } }, - { "⪰̸", { 10928, 824 } }, - { "𝓃", { 120003, 0 } }, - { "∤", { 8740, 0 } }, - { "∦", { 8742, 0 } }, - { "≁", { 8769, 0 } }, - { "≄", { 8772, 0 } }, - { "≄", { 8772, 0 } }, - { "∤", { 8740, 0 } }, - { "∦", { 8742, 0 } }, - { "⋢", { 8930, 0 } }, - { "⋣", { 8931, 0 } }, - { "⊄", { 8836, 0 } }, - { "⫅̸", { 10949, 824 } }, - { "⊈", { 8840, 0 } }, - { "⊂⃒", { 8834, 8402 } }, - { "⊈", { 8840, 0 } }, - { "⫅̸", { 10949, 824 } }, - { "⊁", { 8833, 0 } }, - { "⪰̸", { 10928, 824 } }, - { "⊅", { 8837, 0 } }, - { "⫆̸", { 10950, 824 } }, - { "⊉", { 8841, 0 } }, - { "⊃⃒", { 8835, 8402 } }, - { "⊉", { 8841, 0 } }, - { "⫆̸", { 10950, 824 } }, - { "≹", { 8825, 0 } }, - { "ñ", { 241, 0 } }, - { "≸", { 8824, 0 } }, - { "⋪", { 8938, 0 } }, - { "⋬", { 8940, 0 } }, - { "⋫", { 8939, 0 } }, - { "⋭", { 8941, 0 } }, - { "ν", { 957, 0 } }, - { "#", { 35, 0 } }, - { "№", { 8470, 0 } }, - { " ", { 8199, 0 } }, - { "⊭", { 8877, 0 } }, - { "⤄", { 10500, 0 } }, - { "≍⃒", { 8781, 8402 } }, - { "⊬", { 8876, 0 } }, - { "≥⃒", { 8805, 8402 } }, - { ">⃒", { 62, 8402 } }, - { "⧞", { 10718, 0 } }, - { "⤂", { 10498, 0 } }, - { "≤⃒", { 8804, 8402 } }, - { "<⃒", { 60, 8402 } }, - { "⊴⃒", { 8884, 8402 } }, - { "⤃", { 10499, 0 } }, - { "⊵⃒", { 8885, 8402 } }, - { "∼⃒", { 8764, 8402 } }, - { "⇖", { 8662, 0 } }, - { "⤣", { 10531, 0 } }, - { "↖", { 8598, 0 } }, - { "↖", { 8598, 0 } }, - { "⤧", { 10535, 0 } }, - { "Ⓢ", { 9416, 0 } }, - { "ó", { 243, 0 } }, - { "⊛", { 8859, 0 } }, - { "⊚", { 8858, 0 } }, - { "ô", { 244, 0 } }, - { "о", { 1086, 0 } }, - { "⊝", { 8861, 0 } }, - { "ő", { 337, 0 } }, - { "⨸", { 10808, 0 } }, - { "⊙", { 8857, 0 } }, - { "⦼", { 10684, 0 } }, - { "œ", { 339, 0 } }, - { "⦿", { 10687, 0 } }, - { "𝔬", { 120108, 0 } }, - { "˛", { 731, 0 } }, - { "ò", { 242, 0 } }, - { "⧁", { 10689, 0 } }, - { "⦵", { 10677, 0 } }, - { "Ω", { 937, 0 } }, - { "∮", { 8750, 0 } }, - { "↺", { 8634, 0 } }, - { "⦾", { 10686, 0 } }, - { "⦻", { 10683, 0 } }, - { "‾", { 8254, 0 } }, - { "⧀", { 10688, 0 } }, - { "ō", { 333, 0 } }, - { "ω", { 969, 0 } }, - { "ο", { 959, 0 } }, - { "⦶", { 10678, 0 } }, - { "⊖", { 8854, 0 } }, - { "𝕠", { 120160, 0 } }, - { "⦷", { 10679, 0 } }, - { "⦹", { 10681, 0 } }, - { "⊕", { 8853, 0 } }, - { "∨", { 8744, 0 } }, - { "↻", { 8635, 0 } }, - { "⩝", { 10845, 0 } }, - { "ℴ", { 8500, 0 } }, - { "ℴ", { 8500, 0 } }, - { "ª", { 170, 0 } }, - { "º", { 186, 0 } }, - { "⊶", { 8886, 0 } }, - { "⩖", { 10838, 0 } }, - { "⩗", { 10839, 0 } }, - { "⩛", { 10843, 0 } }, - { "ℴ", { 8500, 0 } }, - { "ø", { 248, 0 } }, - { "⊘", { 8856, 0 } }, - { "õ", { 245, 0 } }, - { "⊗", { 8855, 0 } }, - { "⨶", { 10806, 0 } }, - { "ö", { 246, 0 } }, - { "⌽", { 9021, 0 } }, - { "∥", { 8741, 0 } }, - { "¶", { 182, 0 } }, - { "∥", { 8741, 0 } }, - { "⫳", { 10995, 0 } }, - { "⫽", { 11005, 0 } }, - { "∂", { 8706, 0 } }, - { "п", { 1087, 0 } }, - { "%", { 37, 0 } }, - { ".", { 46, 0 } }, - { "‰", { 8240, 0 } }, - { "⊥", { 8869, 0 } }, - { "‱", { 8241, 0 } }, - { "𝔭", { 120109, 0 } }, - { "φ", { 966, 0 } }, - { "ϕ", { 981, 0 } }, - { "ℳ", { 8499, 0 } }, - { "☎", { 9742, 0 } }, - { "π", { 960, 0 } }, - { "⋔", { 8916, 0 } }, - { "ϖ", { 982, 0 } }, - { "ℏ", { 8463, 0 } }, - { "ℎ", { 8462, 0 } }, - { "ℏ", { 8463, 0 } }, - { "+", { 43, 0 } }, - { "⨣", { 10787, 0 } }, - { "⊞", { 8862, 0 } }, - { "⨢", { 10786, 0 } }, - { "∔", { 8724, 0 } }, - { "⨥", { 10789, 0 } }, - { "⩲", { 10866, 0 } }, - { "±", { 177, 0 } }, - { "⨦", { 10790, 0 } }, - { "⨧", { 10791, 0 } }, - { "±", { 177, 0 } }, - { "⨕", { 10773, 0 } }, - { "𝕡", { 120161, 0 } }, - { "£", { 163, 0 } }, - { "≺", { 8826, 0 } }, - { "⪳", { 10931, 0 } }, - { "⪷", { 10935, 0 } }, - { "≼", { 8828, 0 } }, - { "⪯", { 10927, 0 } }, - { "≺", { 8826, 0 } }, - { "⪷", { 10935, 0 } }, - { "≼", { 8828, 0 } }, - { "⪯", { 10927, 0 } }, - { "⪹", { 10937, 0 } }, - { "⪵", { 10933, 0 } }, - { "⋨", { 8936, 0 } }, - { "≾", { 8830, 0 } }, - { "′", { 8242, 0 } }, - { "ℙ", { 8473, 0 } }, - { "⪵", { 10933, 0 } }, - { "⪹", { 10937, 0 } }, - { "⋨", { 8936, 0 } }, - { "∏", { 8719, 0 } }, - { "⌮", { 9006, 0 } }, - { "⌒", { 8978, 0 } }, - { "⌓", { 8979, 0 } }, - { "∝", { 8733, 0 } }, - { "∝", { 8733, 0 } }, - { "≾", { 8830, 0 } }, - { "⊰", { 8880, 0 } }, - { "𝓅", { 120005, 0 } }, - { "ψ", { 968, 0 } }, - { " ", { 8200, 0 } }, - { "𝔮", { 120110, 0 } }, - { "⨌", { 10764, 0 } }, - { "𝕢", { 120162, 0 } }, - { "⁗", { 8279, 0 } }, - { "𝓆", { 120006, 0 } }, - { "ℍ", { 8461, 0 } }, - { "⨖", { 10774, 0 } }, - { "?", { 63, 0 } }, - { "≟", { 8799, 0 } }, - { """, { 34, 0 } }, - { "⇛", { 8667, 0 } }, - { "⇒", { 8658, 0 } }, - { "⤜", { 10524, 0 } }, - { "⤏", { 10511, 0 } }, - { "⥤", { 10596, 0 } }, - { "∽̱", { 8765, 817 } }, - { "ŕ", { 341, 0 } }, - { "√", { 8730, 0 } }, - { "⦳", { 10675, 0 } }, - { "⟩", { 10217, 0 } }, - { "⦒", { 10642, 0 } }, - { "⦥", { 10661, 0 } }, - { "⟩", { 10217, 0 } }, - { "»", { 187, 0 } }, - { "→", { 8594, 0 } }, - { "⥵", { 10613, 0 } }, - { "⇥", { 8677, 0 } }, - { "⤠", { 10528, 0 } }, - { "⤳", { 10547, 0 } }, - { "⤞", { 10526, 0 } }, - { "↪", { 8618, 0 } }, - { "↬", { 8620, 0 } }, - { "⥅", { 10565, 0 } }, - { "⥴", { 10612, 0 } }, - { "↣", { 8611, 0 } }, - { "↝", { 8605, 0 } }, - { "⤚", { 10522, 0 } }, - { "∶", { 8758, 0 } }, - { "ℚ", { 8474, 0 } }, - { "⤍", { 10509, 0 } }, - { "❳", { 10099, 0 } }, - { "}", { 125, 0 } }, - { "]", { 93, 0 } }, - { "⦌", { 10636, 0 } }, - { "⦎", { 10638, 0 } }, - { "⦐", { 10640, 0 } }, - { "ř", { 345, 0 } }, - { "ŗ", { 343, 0 } }, - { "⌉", { 8969, 0 } }, - { "}", { 125, 0 } }, - { "р", { 1088, 0 } }, - { "⤷", { 10551, 0 } }, - { "⥩", { 10601, 0 } }, - { "”", { 8221, 0 } }, - { "”", { 8221, 0 } }, - { "↳", { 8627, 0 } }, - { "ℜ", { 8476, 0 } }, - { "ℛ", { 8475, 0 } }, - { "ℜ", { 8476, 0 } }, - { "ℝ", { 8477, 0 } }, - { "▭", { 9645, 0 } }, - { "®", { 174, 0 } }, - { "⥽", { 10621, 0 } }, - { "⌋", { 8971, 0 } }, - { "𝔯", { 120111, 0 } }, - { "⇁", { 8641, 0 } }, - { "⇀", { 8640, 0 } }, - { "⥬", { 10604, 0 } }, - { "ρ", { 961, 0 } }, - { "ϱ", { 1009, 0 } }, - { "→", { 8594, 0 } }, - { "↣", { 8611, 0 } }, - { "⇁", { 8641, 0 } }, - { "⇀", { 8640, 0 } }, - { "⇄", { 8644, 0 } }, - { "⇌", { 8652, 0 } }, - { "⇉", { 8649, 0 } }, - { "↝", { 8605, 0 } }, - { "⋌", { 8908, 0 } }, - { "˚", { 730, 0 } }, - { "≓", { 8787, 0 } }, - { "⇄", { 8644, 0 } }, - { "⇌", { 8652, 0 } }, - { "‏", { 8207, 0 } }, - { "⎱", { 9137, 0 } }, - { "⎱", { 9137, 0 } }, - { "⫮", { 10990, 0 } }, - { "⟭", { 10221, 0 } }, - { "⇾", { 8702, 0 } }, - { "⟧", { 10215, 0 } }, - { "⦆", { 10630, 0 } }, - { "𝕣", { 120163, 0 } }, - { "⨮", { 10798, 0 } }, - { "⨵", { 10805, 0 } }, - { ")", { 41, 0 } }, - { "⦔", { 10644, 0 } }, - { "⨒", { 10770, 0 } }, - { "⇉", { 8649, 0 } }, - { "›", { 8250, 0 } }, - { "𝓇", { 120007, 0 } }, - { "↱", { 8625, 0 } }, - { "]", { 93, 0 } }, - { "’", { 8217, 0 } }, - { "’", { 8217, 0 } }, - { "⋌", { 8908, 0 } }, - { "⋊", { 8906, 0 } }, - { "▹", { 9657, 0 } }, - { "⊵", { 8885, 0 } }, - { "▸", { 9656, 0 } }, - { "⧎", { 10702, 0 } }, - { "⥨", { 10600, 0 } }, - { "℞", { 8478, 0 } }, - { "ś", { 347, 0 } }, - { "‚", { 8218, 0 } }, - { "≻", { 8827, 0 } }, - { "⪴", { 10932, 0 } }, - { "⪸", { 10936, 0 } }, - { "š", { 353, 0 } }, - { "≽", { 8829, 0 } }, - { "⪰", { 10928, 0 } }, - { "ş", { 351, 0 } }, - { "ŝ", { 349, 0 } }, - { "⪶", { 10934, 0 } }, - { "⪺", { 10938, 0 } }, - { "⋩", { 8937, 0 } }, - { "⨓", { 10771, 0 } }, - { "≿", { 8831, 0 } }, - { "с", { 1089, 0 } }, - { "⋅", { 8901, 0 } }, - { "⊡", { 8865, 0 } }, - { "⩦", { 10854, 0 } }, - { "⇘", { 8664, 0 } }, - { "⤥", { 10533, 0 } }, - { "↘", { 8600, 0 } }, - { "↘", { 8600, 0 } }, - { "§", { 167, 0 } }, - { ";", { 59, 0 } }, - { "⤩", { 10537, 0 } }, - { "∖", { 8726, 0 } }, - { "∖", { 8726, 0 } }, - { "✶", { 10038, 0 } }, - { "𝔰", { 120112, 0 } }, - { "⌢", { 8994, 0 } }, - { "♯", { 9839, 0 } }, - { "щ", { 1097, 0 } }, - { "ш", { 1096, 0 } }, - { "∣", { 8739, 0 } }, - { "∥", { 8741, 0 } }, - { "­", { 173, 0 } }, - { "σ", { 963, 0 } }, - { "ς", { 962, 0 } }, - { "ς", { 962, 0 } }, - { "∼", { 8764, 0 } }, - { "⩪", { 10858, 0 } }, - { "≃", { 8771, 0 } }, - { "≃", { 8771, 0 } }, - { "⪞", { 10910, 0 } }, - { "⪠", { 10912, 0 } }, - { "⪝", { 10909, 0 } }, - { "⪟", { 10911, 0 } }, - { "≆", { 8774, 0 } }, - { "⨤", { 10788, 0 } }, - { "⥲", { 10610, 0 } }, - { "←", { 8592, 0 } }, - { "∖", { 8726, 0 } }, - { "⨳", { 10803, 0 } }, - { "⧤", { 10724, 0 } }, - { "∣", { 8739, 0 } }, - { "⌣", { 8995, 0 } }, - { "⪪", { 10922, 0 } }, - { "⪬", { 10924, 0 } }, - { "⪬︀", { 10924, 65024 } }, - { "ь", { 1100, 0 } }, - { "/", { 47, 0 } }, - { "⧄", { 10692, 0 } }, - { "⌿", { 9023, 0 } }, - { "𝕤", { 120164, 0 } }, - { "♠", { 9824, 0 } }, - { "♠", { 9824, 0 } }, - { "∥", { 8741, 0 } }, - { "⊓", { 8851, 0 } }, - { "⊓︀", { 8851, 65024 } }, - { "⊔", { 8852, 0 } }, - { "⊔︀", { 8852, 65024 } }, - { "⊏", { 8847, 0 } }, - { "⊑", { 8849, 0 } }, - { "⊏", { 8847, 0 } }, - { "⊑", { 8849, 0 } }, - { "⊐", { 8848, 0 } }, - { "⊒", { 8850, 0 } }, - { "⊐", { 8848, 0 } }, - { "⊒", { 8850, 0 } }, - { "□", { 9633, 0 } }, - { "□", { 9633, 0 } }, - { "▪", { 9642, 0 } }, - { "▪", { 9642, 0 } }, - { "→", { 8594, 0 } }, - { "𝓈", { 120008, 0 } }, - { "∖", { 8726, 0 } }, - { "⌣", { 8995, 0 } }, - { "⋆", { 8902, 0 } }, - { "☆", { 9734, 0 } }, - { "★", { 9733, 0 } }, - { "ϵ", { 1013, 0 } }, - { "ϕ", { 981, 0 } }, - { "¯", { 175, 0 } }, - { "⊂", { 8834, 0 } }, - { "⫅", { 10949, 0 } }, - { "⪽", { 10941, 0 } }, - { "⊆", { 8838, 0 } }, - { "⫃", { 10947, 0 } }, - { "⫁", { 10945, 0 } }, - { "⫋", { 10955, 0 } }, - { "⊊", { 8842, 0 } }, - { "⪿", { 10943, 0 } }, - { "⥹", { 10617, 0 } }, - { "⊂", { 8834, 0 } }, - { "⊆", { 8838, 0 } }, - { "⫅", { 10949, 0 } }, - { "⊊", { 8842, 0 } }, - { "⫋", { 10955, 0 } }, - { "⫇", { 10951, 0 } }, - { "⫕", { 10965, 0 } }, - { "⫓", { 10963, 0 } }, - { "≻", { 8827, 0 } }, - { "⪸", { 10936, 0 } }, - { "≽", { 8829, 0 } }, - { "⪰", { 10928, 0 } }, - { "⪺", { 10938, 0 } }, - { "⪶", { 10934, 0 } }, - { "⋩", { 8937, 0 } }, - { "≿", { 8831, 0 } }, - { "∑", { 8721, 0 } }, - { "♪", { 9834, 0 } }, - { "¹", { 185, 0 } }, - { "¹", { 185, 0 } }, - { "²", { 178, 0 } }, - { "²", { 178, 0 } }, - { "³", { 179, 0 } }, - { "³", { 179, 0 } }, - { "⊃", { 8835, 0 } }, - { "⫆", { 10950, 0 } }, - { "⪾", { 10942, 0 } }, - { "⫘", { 10968, 0 } }, - { "⊇", { 8839, 0 } }, - { "⫄", { 10948, 0 } }, - { "⟉", { 10185, 0 } }, - { "⫗", { 10967, 0 } }, - { "⥻", { 10619, 0 } }, - { "⫂", { 10946, 0 } }, - { "⫌", { 10956, 0 } }, - { "⊋", { 8843, 0 } }, - { "⫀", { 10944, 0 } }, - { "⊃", { 8835, 0 } }, - { "⊇", { 8839, 0 } }, - { "⫆", { 10950, 0 } }, - { "⊋", { 8843, 0 } }, - { "⫌", { 10956, 0 } }, - { "⫈", { 10952, 0 } }, - { "⫔", { 10964, 0 } }, - { "⫖", { 10966, 0 } }, - { "⇙", { 8665, 0 } }, - { "⤦", { 10534, 0 } }, - { "↙", { 8601, 0 } }, - { "↙", { 8601, 0 } }, - { "⤪", { 10538, 0 } }, - { "ß", { 223, 0 } }, - { "⌖", { 8982, 0 } }, - { "τ", { 964, 0 } }, - { "⎴", { 9140, 0 } }, - { "ť", { 357, 0 } }, - { "ţ", { 355, 0 } }, - { "т", { 1090, 0 } }, - { "⃛", { 8411, 0 } }, - { "⌕", { 8981, 0 } }, - { "𝔱", { 120113, 0 } }, - { "∴", { 8756, 0 } }, - { "∴", { 8756, 0 } }, - { "θ", { 952, 0 } }, - { "ϑ", { 977, 0 } }, - { "ϑ", { 977, 0 } }, - { "≈", { 8776, 0 } }, - { "∼", { 8764, 0 } }, - { " ", { 8201, 0 } }, - { "≈", { 8776, 0 } }, - { "∼", { 8764, 0 } }, - { "þ", { 254, 0 } }, - { "˜", { 732, 0 } }, - { "×", { 215, 0 } }, - { "⊠", { 8864, 0 } }, - { "⨱", { 10801, 0 } }, - { "⨰", { 10800, 0 } }, - { "∭", { 8749, 0 } }, - { "⤨", { 10536, 0 } }, - { "⊤", { 8868, 0 } }, - { "⌶", { 9014, 0 } }, - { "⫱", { 10993, 0 } }, - { "𝕥", { 120165, 0 } }, - { "⫚", { 10970, 0 } }, - { "⤩", { 10537, 0 } }, - { "‴", { 8244, 0 } }, - { "™", { 8482, 0 } }, - { "▵", { 9653, 0 } }, - { "▿", { 9663, 0 } }, - { "◃", { 9667, 0 } }, - { "⊴", { 8884, 0 } }, - { "≜", { 8796, 0 } }, - { "▹", { 9657, 0 } }, - { "⊵", { 8885, 0 } }, - { "◬", { 9708, 0 } }, - { "≜", { 8796, 0 } }, - { "⨺", { 10810, 0 } }, - { "⨹", { 10809, 0 } }, - { "⧍", { 10701, 0 } }, - { "⨻", { 10811, 0 } }, - { "⏢", { 9186, 0 } }, - { "𝓉", { 120009, 0 } }, - { "ц", { 1094, 0 } }, - { "ћ", { 1115, 0 } }, - { "ŧ", { 359, 0 } }, - { "≬", { 8812, 0 } }, - { "↞", { 8606, 0 } }, - { "↠", { 8608, 0 } }, - { "⇑", { 8657, 0 } }, - { "⥣", { 10595, 0 } }, - { "ú", { 250, 0 } }, - { "↑", { 8593, 0 } }, - { "ў", { 1118, 0 } }, - { "ŭ", { 365, 0 } }, - { "û", { 251, 0 } }, - { "у", { 1091, 0 } }, - { "⇅", { 8645, 0 } }, - { "ű", { 369, 0 } }, - { "⥮", { 10606, 0 } }, - { "⥾", { 10622, 0 } }, - { "𝔲", { 120114, 0 } }, - { "ù", { 249, 0 } }, - { "↿", { 8639, 0 } }, - { "↾", { 8638, 0 } }, - { "▀", { 9600, 0 } }, - { "⌜", { 8988, 0 } }, - { "⌜", { 8988, 0 } }, - { "⌏", { 8975, 0 } }, - { "◸", { 9720, 0 } }, - { "ū", { 363, 0 } }, - { "¨", { 168, 0 } }, - { "ų", { 371, 0 } }, - { "𝕦", { 120166, 0 } }, - { "↑", { 8593, 0 } }, - { "↕", { 8597, 0 } }, - { "↿", { 8639, 0 } }, - { "↾", { 8638, 0 } }, - { "⊎", { 8846, 0 } }, - { "υ", { 965, 0 } }, - { "ϒ", { 978, 0 } }, - { "υ", { 965, 0 } }, - { "⇈", { 8648, 0 } }, - { "⌝", { 8989, 0 } }, - { "⌝", { 8989, 0 } }, - { "⌎", { 8974, 0 } }, - { "ů", { 367, 0 } }, - { "◹", { 9721, 0 } }, - { "𝓊", { 120010, 0 } }, - { "⋰", { 8944, 0 } }, - { "ũ", { 361, 0 } }, - { "▵", { 9653, 0 } }, - { "▴", { 9652, 0 } }, - { "⇈", { 8648, 0 } }, - { "ü", { 252, 0 } }, - { "⦧", { 10663, 0 } }, - { "⇕", { 8661, 0 } }, - { "⫨", { 10984, 0 } }, - { "⫩", { 10985, 0 } }, - { "⊨", { 8872, 0 } }, - { "⦜", { 10652, 0 } }, - { "ϵ", { 1013, 0 } }, - { "ϰ", { 1008, 0 } }, - { "∅", { 8709, 0 } }, - { "ϕ", { 981, 0 } }, - { "ϖ", { 982, 0 } }, - { "∝", { 8733, 0 } }, - { "↕", { 8597, 0 } }, - { "ϱ", { 1009, 0 } }, - { "ς", { 962, 0 } }, - { "⊊︀", { 8842, 65024 } }, - { "⫋︀", { 10955, 65024 } }, - { "⊋︀", { 8843, 65024 } }, - { "⫌︀", { 10956, 65024 } }, - { "ϑ", { 977, 0 } }, - { "⊲", { 8882, 0 } }, - { "⊳", { 8883, 0 } }, - { "в", { 1074, 0 } }, - { "⊢", { 8866, 0 } }, - { "∨", { 8744, 0 } }, - { "⊻", { 8891, 0 } }, - { "≚", { 8794, 0 } }, - { "⋮", { 8942, 0 } }, - { "|", { 124, 0 } }, - { "|", { 124, 0 } }, - { "𝔳", { 120115, 0 } }, - { "⊲", { 8882, 0 } }, - { "⊂⃒", { 8834, 8402 } }, - { "⊃⃒", { 8835, 8402 } }, - { "𝕧", { 120167, 0 } }, - { "∝", { 8733, 0 } }, - { "⊳", { 8883, 0 } }, - { "𝓋", { 120011, 0 } }, - { "⫋︀", { 10955, 65024 } }, - { "⊊︀", { 8842, 65024 } }, - { "⫌︀", { 10956, 65024 } }, - { "⊋︀", { 8843, 65024 } }, - { "⦚", { 10650, 0 } }, - { "ŵ", { 373, 0 } }, - { "⩟", { 10847, 0 } }, - { "∧", { 8743, 0 } }, - { "≙", { 8793, 0 } }, - { "℘", { 8472, 0 } }, - { "𝔴", { 120116, 0 } }, - { "𝕨", { 120168, 0 } }, - { "℘", { 8472, 0 } }, - { "≀", { 8768, 0 } }, - { "≀", { 8768, 0 } }, - { "𝓌", { 120012, 0 } }, - { "⋂", { 8898, 0 } }, - { "◯", { 9711, 0 } }, - { "⋃", { 8899, 0 } }, - { "▽", { 9661, 0 } }, - { "𝔵", { 120117, 0 } }, - { "⟺", { 10234, 0 } }, - { "⟷", { 10231, 0 } }, - { "ξ", { 958, 0 } }, - { "⟸", { 10232, 0 } }, - { "⟵", { 10229, 0 } }, - { "⟼", { 10236, 0 } }, - { "⋻", { 8955, 0 } }, - { "⨀", { 10752, 0 } }, - { "𝕩", { 120169, 0 } }, - { "⨁", { 10753, 0 } }, - { "⨂", { 10754, 0 } }, - { "⟹", { 10233, 0 } }, - { "⟶", { 10230, 0 } }, - { "𝓍", { 120013, 0 } }, - { "⨆", { 10758, 0 } }, - { "⨄", { 10756, 0 } }, - { "△", { 9651, 0 } }, - { "⋁", { 8897, 0 } }, - { "⋀", { 8896, 0 } }, - { "ý", { 253, 0 } }, - { "я", { 1103, 0 } }, - { "ŷ", { 375, 0 } }, - { "ы", { 1099, 0 } }, - { "¥", { 165, 0 } }, - { "𝔶", { 120118, 0 } }, - { "ї", { 1111, 0 } }, - { "𝕪", { 120170, 0 } }, - { "𝓎", { 120014, 0 } }, - { "ю", { 1102, 0 } }, - { "ÿ", { 255, 0 } }, - { "ź", { 378, 0 } }, - { "ž", { 382, 0 } }, - { "з", { 1079, 0 } }, - { "ż", { 380, 0 } }, - { "ℨ", { 8488, 0 } }, - { "ζ", { 950, 0 } }, - { "𝔷", { 120119, 0 } }, - { "ж", { 1078, 0 } }, - { "⇝", { 8669, 0 } }, - { "𝕫", { 120171, 0 } }, - { "𝓏", { 120015, 0 } }, - { "‍", { 8205, 0 } }, - { "‌", { 8204, 0 } } -}; - - -struct entity_key { - const char* name; - size_t name_size; -}; - -static int -entity_cmp(const void* p_key, const void* p_entity) -{ - struct entity_key* key = (struct entity_key*) p_key; - struct entity* ent = (struct entity*) p_entity; - - return strncmp(key->name, ent->name, key->name_size); -} - -const struct entity* -entity_lookup(const char* name, size_t name_size) -{ - struct entity_key key = { name, name_size }; - - return bsearch(&key, - entity_table, - sizeof(entity_table) / sizeof(entity_table[0]), - sizeof(struct entity), - entity_cmp); -} diff --git a/src/entity.h b/src/entity.h deleted file mode 100644 index 36395fe..0000000 --- a/src/entity.h +++ /dev/null @@ -1,42 +0,0 @@ -/* - * MD4C: Markdown parser for C - * (http://github.com/mity/md4c) - * - * Copyright (c) 2016-2019 Martin Mitas - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS - * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ - -#ifndef MD4C_ENTITY_H -#define MD4C_ENTITY_H - -#include <stdlib.h> - - -/* Most entities are formed by single Unicode codepoint, few by two codepoints. - * Single-codepoint entities have codepoints[1] set to zero. */ -struct entity { - const char* name; - unsigned codepoints[2]; -}; - -const struct entity* entity_lookup(const char* name, size_t name_size); - - -#endif /* MD4C_ENTITY_H */ diff --git a/src/main.cpp b/src/main.cpp new file mode 100644 index 0000000..98f19c5 --- /dev/null +++ b/src/main.cpp @@ -0,0 +1,131 @@ +#include "ContentManager.hpp" +#include "EventManager.hpp" +#include "Filesystem.hpp" +#include "TemplateManager.hpp" + +#include "content/blog/Blog.hpp" + +#include <event2/buffer.h> + +#include <inja/inja.hpp> +#include <iostream> + +static void example_inja_render(struct evhttp_request * const req, EvUserData ud) +{ + (void)ud; + + inja::Environment env; + nlohmann::json data; + data["name"] = "Peter"; + data["city"] = "Brunswick"; + data["age"] = 29; + data["names"] = {"Jeff", "Seb"}; + data["brother"]["name"] = "Chris"; + data["brother"]["daughters"] = {"Maria", "Helen"}; + data["brother"]["daughter0"] = {{"name", "Maria"}}; + data["is_happy"] = true; + data["is_sad"] = false; + data["relatives"]["mother"] = "Maria"; + data["relatives"]["brother"] = "Chris"; + data["relatives"]["sister"] = "Jenny"; + data["vars"] = {2, 3, 4, 0, -1, -2, -3}; + + auto reply = env.render( + "<html><body>\n" + "Hello {{ name }}! I come from {{ city }}.<br>\n" + "Hello {{ names.1 }}!<br>\n" + "Hello {{ brother.name }}!<br>\n" + "Hello {{ brother.daughter0.name }}!<br>\n" + "{{ \"{{ no_value }}\" }}<br>\n" + "Hello{# This is a comment #}!<br>\n" + "{# --- #Todo --- #}<br>\n" + "{% for name in names %}a{% endfor %}<br>\n" + "Hello {% for name in names %}{{ name }} {% endfor %}!<br>\n" + "Hello {% for name in names %}{{ loop.index }}: {{ name }}, {% " + "endfor %}!<br>\n" + "{% for type, name in relatives %}{{ type }}: {{ name }}, {% endfor " + "%}<br>\n" + "{% for v in vars %}{% if v > 0 %}+{% endif %}{% endfor %}<br>\n" + "{% for name in names %}{{ loop.index }}: {{ name }}{% if not " + "loop.is_last %}, {% endif %}{% endfor %}!<br>\n" + "{% for name in names %}{{ loop.index }}: {{ name }}{% if " + "loop.is_last == false %}, {% endif %}{% endfor %}!<br>\n" + "{% for name in names %}a{% endfor %}<br>\n" + "</body></html>\n", + data); + + evhttp_add_header(evhttp_request_get_output_headers(req), "Content-Type", "text/html"); + + struct evbuffer * const output = evbuffer_new(); + if (output != nullptr) + { + evbuffer_add(output, reply.c_str(), reply.size()); + evhttp_send_reply(req, 200, "OK", output); + evbuffer_free(output); + } +} + +int main(int argc, char ** argv) +{ + char const * host = "127.0.0.1"; + uint16_t port = 9000; + + if (argc <= 1 || argc > 5) + { + std::cout << "usage: cpp-web [HOST] [PORT] [STATIC-WWWROOT] [DYNAMIC-WWWROOT]" << std::endl; + if (argc > 5) + { + return 1; + } + } + + if (argc > 1) + { + host = argv[1]; + } + if (argc > 2) + { + port = atoi(argv[2]); + } + + std::shared_ptr<Filesystem> static_fs = std::make_shared<Filesystem>(); + if (argc > 3) + { + if (static_fs->Scan(argv[3]) != true) + { + return 1; + } + } + static_fs->Scan("./wwwroot", {"html", "tmpl"}, true); + + Filesystem dynamic_fs; + if (argc > 4) + { + if (dynamic_fs.Scan(argv[4]) != true) + { + return 1; + } + } + dynamic_fs.Scan("./wwwroot", {"html", "tmpl"}, false); + + std::shared_ptr<TemplateManager> tmgr = std::make_shared<TemplateManager>(); + tmgr->ParseTemplates(dynamic_fs); + + std::shared_ptr<ContentManager> ctmgr = std::make_shared<ContentManager>(); + ctmgr->SetStaticFilesystem(static_fs); + ctmgr->SetTemplateSystem(tmgr); + ctmgr->RegisterModule(std::make_shared<Blog>("/blog", "index.html")); + + if (ctmgr->InitAll() == false) + { + std::cout << "InitAll() failed." << std::endl; + return 1; + } + + EventManager evmgr(ctmgr); + // evmgr.SetDefaultCallback(example_inja_render, {}); + evmgr.AddCallback("/bla", example_inja_render, {}); + evmgr.Init(host, port); + + // ctmgr.ShutdownAll(); +} diff --git a/src/md4c-html.c b/src/md4c-html.c deleted file mode 100644 index d604aec..0000000 --- a/src/md4c-html.c +++ /dev/null @@ -1,573 +0,0 @@ -/* - * MD4C: Markdown parser for C - * (http://github.com/mity/md4c) - * - * Copyright (c) 2016-2019 Martin Mitas - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS - * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ - -#include <stdio.h> -#include <string.h> - -#include "md4c-html.h" -#include "entity.h" - - -#if !defined(__STDC_VERSION__) || __STDC_VERSION__ < 199409L - /* C89/90 or old compilers in general may not understand "inline". */ - #if defined __GNUC__ - #define inline __inline__ - #elif defined _MSC_VER - #define inline __inline - #else - #define inline - #endif -#endif - -#ifdef _WIN32 - #define snprintf _snprintf -#endif - - - -typedef struct MD_HTML_tag MD_HTML; -struct MD_HTML_tag { - void (*process_output)(const MD_CHAR*, MD_SIZE, void*); - void* userdata; - unsigned flags; - int image_nesting_level; - char escape_map[256]; -}; - -#define NEED_HTML_ESC_FLAG 0x1 -#define NEED_URL_ESC_FLAG 0x2 - - -/***************************************** - *** HTML rendering helper functions *** - *****************************************/ - -#define ISDIGIT(ch) ('0' <= (ch) && (ch) <= '9') -#define ISLOWER(ch) ('a' <= (ch) && (ch) <= 'z') -#define ISUPPER(ch) ('A' <= (ch) && (ch) <= 'Z') -#define ISALNUM(ch) (ISLOWER(ch) || ISUPPER(ch) || ISDIGIT(ch)) - - -static inline void -render_verbatim(MD_HTML* r, const MD_CHAR* text, MD_SIZE size) -{ - r->process_output(text, size, r->userdata); -} - -/* Keep this as a macro. Most compiler should then be smart enough to replace - * the strlen() call with a compile-time constant if the string is a C literal. */ -#define RENDER_VERBATIM(r, verbatim) \ - render_verbatim((r), (verbatim), (MD_SIZE) (strlen(verbatim))) - - -static void -render_html_escaped(MD_HTML* r, const MD_CHAR* data, MD_SIZE size) -{ - MD_OFFSET beg = 0; - MD_OFFSET off = 0; - - /* Some characters need to be escaped in normal HTML text. */ - #define NEED_HTML_ESC(ch) (r->escape_map[(unsigned char)(ch)] & NEED_HTML_ESC_FLAG) - - while(1) { - /* Optimization: Use some loop unrolling. */ - while(off + 3 < size && !NEED_HTML_ESC(data[off+0]) && !NEED_HTML_ESC(data[off+1]) - && !NEED_HTML_ESC(data[off+2]) && !NEED_HTML_ESC(data[off+3])) - off += 4; - while(off < size && !NEED_HTML_ESC(data[off])) - off++; - - if(off > beg) - render_verbatim(r, data + beg, off - beg); - - if(off < size) { - switch(data[off]) { - case '&': RENDER_VERBATIM(r, "&"); break; - case '<': RENDER_VERBATIM(r, "<"); break; - case '>': RENDER_VERBATIM(r, ">"); break; - case '"': RENDER_VERBATIM(r, """); break; - } - off++; - } else { - break; - } - beg = off; - } -} - -static void -render_url_escaped(MD_HTML* r, const MD_CHAR* data, MD_SIZE size) -{ - static const MD_CHAR hex_chars[] = "0123456789ABCDEF"; - MD_OFFSET beg = 0; - MD_OFFSET off = 0; - - /* Some characters need to be escaped in URL attributes. */ - #define NEED_URL_ESC(ch) (r->escape_map[(unsigned char)(ch)] & NEED_URL_ESC_FLAG) - - while(1) { - while(off < size && !NEED_URL_ESC(data[off])) - off++; - if(off > beg) - render_verbatim(r, data + beg, off - beg); - - if(off < size) { - char hex[3]; - - switch(data[off]) { - case '&': RENDER_VERBATIM(r, "&"); break; - default: - hex[0] = '%'; - hex[1] = hex_chars[((unsigned)data[off] >> 4) & 0xf]; - hex[2] = hex_chars[((unsigned)data[off] >> 0) & 0xf]; - render_verbatim(r, hex, 3); - break; - } - off++; - } else { - break; - } - - beg = off; - } -} - -static unsigned -hex_val(char ch) -{ - if('0' <= ch && ch <= '9') - return ch - '0'; - if('A' <= ch && ch <= 'Z') - return ch - 'A' + 10; - else - return ch - 'a' + 10; -} - -static void -render_utf8_codepoint(MD_HTML* r, unsigned codepoint, - void (*fn_append)(MD_HTML*, const MD_CHAR*, MD_SIZE)) -{ - static const MD_CHAR utf8_replacement_char[] = { 0xef, 0xbf, 0xbd }; - - unsigned char utf8[4]; - size_t n; - - if(codepoint <= 0x7f) { - n = 1; - utf8[0] = codepoint; - } else if(codepoint <= 0x7ff) { - n = 2; - utf8[0] = 0xc0 | ((codepoint >> 6) & 0x1f); - utf8[1] = 0x80 + ((codepoint >> 0) & 0x3f); - } else if(codepoint <= 0xffff) { - n = 3; - utf8[0] = 0xe0 | ((codepoint >> 12) & 0xf); - utf8[1] = 0x80 + ((codepoint >> 6) & 0x3f); - utf8[2] = 0x80 + ((codepoint >> 0) & 0x3f); - } else { - n = 4; - utf8[0] = 0xf0 | ((codepoint >> 18) & 0x7); - utf8[1] = 0x80 + ((codepoint >> 12) & 0x3f); - utf8[2] = 0x80 + ((codepoint >> 6) & 0x3f); - utf8[3] = 0x80 + ((codepoint >> 0) & 0x3f); - } - - if(0 < codepoint && codepoint <= 0x10ffff) - fn_append(r, (char*)utf8, (MD_SIZE)n); - else - fn_append(r, utf8_replacement_char, 3); -} - -/* Translate entity to its UTF-8 equivalent, or output the verbatim one - * if such entity is unknown (or if the translation is disabled). */ -static void -render_entity(MD_HTML* r, const MD_CHAR* text, MD_SIZE size, - void (*fn_append)(MD_HTML*, const MD_CHAR*, MD_SIZE)) -{ - if(r->flags & MD_HTML_FLAG_VERBATIM_ENTITIES) { - render_verbatim(r, text, size); - return; - } - - /* We assume UTF-8 output is what is desired. */ - if(size > 3 && text[1] == '#') { - unsigned codepoint = 0; - - if(text[2] == 'x' || text[2] == 'X') { - /* Hexadecimal entity (e.g. "�")). */ - MD_SIZE i; - for(i = 3; i < size-1; i++) - codepoint = 16 * codepoint + hex_val(text[i]); - } else { - /* Decimal entity (e.g. "&1234;") */ - MD_SIZE i; - for(i = 2; i < size-1; i++) - codepoint = 10 * codepoint + (text[i] - '0'); - } - - render_utf8_codepoint(r, codepoint, fn_append); - return; - } else { - /* Named entity (e.g. " "). */ - const struct entity* ent; - - ent = entity_lookup(text, size); - if(ent != NULL) { - render_utf8_codepoint(r, ent->codepoints[0], fn_append); - if(ent->codepoints[1]) - render_utf8_codepoint(r, ent->codepoints[1], fn_append); - return; - } - } - - fn_append(r, text, size); -} - -static void -render_attribute(MD_HTML* r, const MD_ATTRIBUTE* attr, - void (*fn_append)(MD_HTML*, const MD_CHAR*, MD_SIZE)) -{ - int i; - - for(i = 0; attr->substr_offsets[i] < attr->size; i++) { - MD_TEXTTYPE type = attr->substr_types[i]; - MD_OFFSET off = attr->substr_offsets[i]; - MD_SIZE size = attr->substr_offsets[i+1] - off; - const MD_CHAR* text = attr->text + off; - - switch(type) { - case MD_TEXT_NULLCHAR: render_utf8_codepoint(r, 0x0000, render_verbatim); break; - case MD_TEXT_ENTITY: render_entity(r, text, size, fn_append); break; - default: fn_append(r, text, size); break; - } - } -} - - -static void -render_open_ol_block(MD_HTML* r, const MD_BLOCK_OL_DETAIL* det) -{ - char buf[64]; - - if(det->start == 1) { - RENDER_VERBATIM(r, "<ol>\n"); - return; - } - - snprintf(buf, sizeof(buf), "<ol start=\"%u\">\n", det->start); - RENDER_VERBATIM(r, buf); -} - -static void -render_open_li_block(MD_HTML* r, const MD_BLOCK_LI_DETAIL* det) -{ - if(det->is_task) { - RENDER_VERBATIM(r, "<li class=\"task-list-item\">" - "<input type=\"checkbox\" class=\"task-list-item-checkbox\" disabled"); - if(det->task_mark == 'x' || det->task_mark == 'X') - RENDER_VERBATIM(r, " checked"); - RENDER_VERBATIM(r, ">"); - } else { - RENDER_VERBATIM(r, "<li>"); - } -} - -static void -render_open_code_block(MD_HTML* r, const MD_BLOCK_CODE_DETAIL* det) -{ - RENDER_VERBATIM(r, "<pre><code"); - - /* If known, output the HTML 5 attribute class="language-LANGNAME". */ - if(det->lang.text != NULL) { - RENDER_VERBATIM(r, " class=\"language-"); - render_attribute(r, &det->lang, render_html_escaped); - RENDER_VERBATIM(r, "\""); - } - - RENDER_VERBATIM(r, ">"); -} - -static void -render_open_td_block(MD_HTML* r, const MD_CHAR* cell_type, const MD_BLOCK_TD_DETAIL* det) -{ - RENDER_VERBATIM(r, "<"); - RENDER_VERBATIM(r, cell_type); - - switch(det->align) { - case MD_ALIGN_LEFT: RENDER_VERBATIM(r, " align=\"left\">"); break; - case MD_ALIGN_CENTER: RENDER_VERBATIM(r, " align=\"center\">"); break; - case MD_ALIGN_RIGHT: RENDER_VERBATIM(r, " align=\"right\">"); break; - default: RENDER_VERBATIM(r, ">"); break; - } -} - -static void -render_open_a_span(MD_HTML* r, const MD_SPAN_A_DETAIL* det) -{ - RENDER_VERBATIM(r, "<a href=\""); - render_attribute(r, &det->href, render_url_escaped); - - if(det->title.text != NULL) { - RENDER_VERBATIM(r, "\" title=\""); - render_attribute(r, &det->title, render_html_escaped); - } - - RENDER_VERBATIM(r, "\">"); -} - -static void -render_open_img_span(MD_HTML* r, const MD_SPAN_IMG_DETAIL* det) -{ - RENDER_VERBATIM(r, "<img src=\""); - render_attribute(r, &det->src, render_url_escaped); - - RENDER_VERBATIM(r, "\" alt=\""); - - r->image_nesting_level++; -} - -static void -render_close_img_span(MD_HTML* r, const MD_SPAN_IMG_DETAIL* det) -{ - if(det->title.text != NULL) { - RENDER_VERBATIM(r, "\" title=\""); - render_attribute(r, &det->title, render_html_escaped); - } - - RENDER_VERBATIM(r, (r->flags & MD_HTML_FLAG_XHTML) ? "\" />" : "\">"); - - r->image_nesting_level--; -} - -static void -render_open_wikilink_span(MD_HTML* r, const MD_SPAN_WIKILINK_DETAIL* det) -{ - RENDER_VERBATIM(r, "<x-wikilink data-target=\""); - render_attribute(r, &det->target, render_html_escaped); - - RENDER_VERBATIM(r, "\">"); -} - - -/************************************** - *** HTML renderer implementation *** - **************************************/ - -static int -enter_block_callback(MD_BLOCKTYPE type, void* detail, void* userdata) -{ - static const MD_CHAR* head[6] = { "<h1>", "<h2>", "<h3>", "<h4>", "<h5>", "<h6>" }; - MD_HTML* r = (MD_HTML*) userdata; - - switch(type) { - case MD_BLOCK_DOC: /* noop */ break; - case MD_BLOCK_QUOTE: RENDER_VERBATIM(r, "<blockquote>\n"); break; - case MD_BLOCK_UL: RENDER_VERBATIM(r, "<ul>\n"); break; - case MD_BLOCK_OL: render_open_ol_block(r, (const MD_BLOCK_OL_DETAIL*)detail); break; - case MD_BLOCK_LI: render_open_li_block(r, (const MD_BLOCK_LI_DETAIL*)detail); break; - case MD_BLOCK_HR: RENDER_VERBATIM(r, (r->flags & MD_HTML_FLAG_XHTML) ? "<hr />\n" : "<hr>\n"); break; - case MD_BLOCK_H: RENDER_VERBATIM(r, head[((MD_BLOCK_H_DETAIL*)detail)->level - 1]); break; - case MD_BLOCK_CODE: render_open_code_block(r, (const MD_BLOCK_CODE_DETAIL*) detail); break; - case MD_BLOCK_HTML: /* noop */ break; - case MD_BLOCK_P: RENDER_VERBATIM(r, "<p>"); break; - case MD_BLOCK_TABLE: RENDER_VERBATIM(r, "<table>\n"); break; - case MD_BLOCK_THEAD: RENDER_VERBATIM(r, "<thead>\n"); break; - case MD_BLOCK_TBODY: RENDER_VERBATIM(r, "<tbody>\n"); break; - case MD_BLOCK_TR: RENDER_VERBATIM(r, "<tr>\n"); break; - case MD_BLOCK_TH: render_open_td_block(r, "th", (MD_BLOCK_TD_DETAIL*)detail); break; - case MD_BLOCK_TD: render_open_td_block(r, "td", (MD_BLOCK_TD_DETAIL*)detail); break; - } - - return 0; -} - -static int -leave_block_callback(MD_BLOCKTYPE type, void* detail, void* userdata) -{ - static const MD_CHAR* head[6] = { "</h1>\n", "</h2>\n", "</h3>\n", "</h4>\n", "</h5>\n", "</h6>\n" }; - MD_HTML* r = (MD_HTML*) userdata; - - switch(type) { - case MD_BLOCK_DOC: /*noop*/ break; - case MD_BLOCK_QUOTE: RENDER_VERBATIM(r, "</blockquote>\n"); break; - case MD_BLOCK_UL: RENDER_VERBATIM(r, "</ul>\n"); break; - case MD_BLOCK_OL: RENDER_VERBATIM(r, "</ol>\n"); break; - case MD_BLOCK_LI: RENDER_VERBATIM(r, "</li>\n"); break; - case MD_BLOCK_HR: /*noop*/ break; - case MD_BLOCK_H: RENDER_VERBATIM(r, head[((MD_BLOCK_H_DETAIL*)detail)->level - 1]); break; - case MD_BLOCK_CODE: RENDER_VERBATIM(r, "</code></pre>\n"); break; - case MD_BLOCK_HTML: /* noop */ break; - case MD_BLOCK_P: RENDER_VERBATIM(r, "</p>\n"); break; - case MD_BLOCK_TABLE: RENDER_VERBATIM(r, "</table>\n"); break; - case MD_BLOCK_THEAD: RENDER_VERBATIM(r, "</thead>\n"); break; - case MD_BLOCK_TBODY: RENDER_VERBATIM(r, "</tbody>\n"); break; - case MD_BLOCK_TR: RENDER_VERBATIM(r, "</tr>\n"); break; - case MD_BLOCK_TH: RENDER_VERBATIM(r, "</th>\n"); break; - case MD_BLOCK_TD: RENDER_VERBATIM(r, "</td>\n"); break; - } - - return 0; -} - -static int -enter_span_callback(MD_SPANTYPE type, void* detail, void* userdata) -{ - MD_HTML* r = (MD_HTML*) userdata; - - if(r->image_nesting_level > 0) { - /* We are inside a Markdown image label. Markdown allows to use any - * emphasis and other rich contents in that context similarly as in - * any link label. - * - * However, unlike in the case of links (where that contents becomes - * contents of the <a>...</a> tag), in the case of images the contents - * is supposed to fall into the attribute alt: <img alt="...">. - * - * In that context we naturally cannot output nested HTML tags. So lets - * suppress them and only output the plain text (i.e. what falls into - * text() callback). - * - * This make-it-a-plain-text approach is the recommended practice by - * CommonMark specification (for HTML output). - */ - return 0; - } - - switch(type) { - case MD_SPAN_EM: RENDER_VERBATIM(r, "<em>"); break; - case MD_SPAN_STRONG: RENDER_VERBATIM(r, "<strong>"); break; - case MD_SPAN_U: RENDER_VERBATIM(r, "<u>"); break; - case MD_SPAN_A: render_open_a_span(r, (MD_SPAN_A_DETAIL*) detail); break; - case MD_SPAN_IMG: render_open_img_span(r, (MD_SPAN_IMG_DETAIL*) detail); break; - case MD_SPAN_CODE: RENDER_VERBATIM(r, "<code>"); break; - case MD_SPAN_DEL: RENDER_VERBATIM(r, "<del>"); break; - case MD_SPAN_LATEXMATH: RENDER_VERBATIM(r, "<x-equation>"); break; - case MD_SPAN_LATEXMATH_DISPLAY: RENDER_VERBATIM(r, "<x-equation type=\"display\">"); break; - case MD_SPAN_WIKILINK: render_open_wikilink_span(r, (MD_SPAN_WIKILINK_DETAIL*) detail); break; - } - - return 0; -} - -static int -leave_span_callback(MD_SPANTYPE type, void* detail, void* userdata) -{ - MD_HTML* r = (MD_HTML*) userdata; - - if(r->image_nesting_level > 0) { - /* Ditto as in enter_span_callback(), except we have to allow the - * end of the <img> tag. */ - if(r->image_nesting_level == 1 && type == MD_SPAN_IMG) - render_close_img_span(r, (MD_SPAN_IMG_DETAIL*) detail); - return 0; - } - - switch(type) { - case MD_SPAN_EM: RENDER_VERBATIM(r, "</em>"); break; - case MD_SPAN_STRONG: RENDER_VERBATIM(r, "</strong>"); break; - case MD_SPAN_U: RENDER_VERBATIM(r, "</u>"); break; - case MD_SPAN_A: RENDER_VERBATIM(r, "</a>"); break; - case MD_SPAN_IMG: /*noop, handled above*/ break; - case MD_SPAN_CODE: RENDER_VERBATIM(r, "</code>"); break; - case MD_SPAN_DEL: RENDER_VERBATIM(r, "</del>"); break; - case MD_SPAN_LATEXMATH: /*fall through*/ - case MD_SPAN_LATEXMATH_DISPLAY: RENDER_VERBATIM(r, "</x-equation>"); break; - case MD_SPAN_WIKILINK: RENDER_VERBATIM(r, "</x-wikilink>"); break; - } - - return 0; -} - -static int -text_callback(MD_TEXTTYPE type, const MD_CHAR* text, MD_SIZE size, void* userdata) -{ - MD_HTML* r = (MD_HTML*) userdata; - - switch(type) { - case MD_TEXT_NULLCHAR: render_utf8_codepoint(r, 0x0000, render_verbatim); break; - case MD_TEXT_BR: RENDER_VERBATIM(r, (r->image_nesting_level == 0 - ? ((r->flags & MD_HTML_FLAG_XHTML) ? "<br />\n" : "<br>\n") - : " ")); - break; - case MD_TEXT_SOFTBR: RENDER_VERBATIM(r, (r->image_nesting_level == 0 ? "\n" : " ")); break; - case MD_TEXT_HTML: render_verbatim(r, text, size); break; - case MD_TEXT_ENTITY: render_entity(r, text, size, render_html_escaped); break; - default: render_html_escaped(r, text, size); break; - } - - return 0; -} - -static void -debug_log_callback(const char* msg, void* userdata) -{ - MD_HTML* r = (MD_HTML*) userdata; - if(r->flags & MD_HTML_FLAG_DEBUG) - fprintf(stderr, "MD4C: %s\n", msg); -} - -int -md_html(const MD_CHAR* input, MD_SIZE input_size, - void (*process_output)(const MD_CHAR*, MD_SIZE, void*), - void* userdata, unsigned parser_flags, unsigned renderer_flags) -{ - MD_HTML render = { process_output, userdata, renderer_flags, 0, { 0 } }; - int i; - - MD_PARSER parser = { - 0, - parser_flags, - enter_block_callback, - leave_block_callback, - enter_span_callback, - leave_span_callback, - text_callback, - debug_log_callback, - NULL - }; - - /* Build map of characters which need escaping. */ - for(i = 0; i < 256; i++) { - unsigned char ch = (unsigned char) i; - - if(strchr("\"&<>", ch) != NULL) - render.escape_map[i] |= NEED_HTML_ESC_FLAG; - - if(!ISALNUM(ch) && strchr("~-_.+!*(),%#@?=;:/,+$", ch) == NULL) - render.escape_map[i] |= NEED_URL_ESC_FLAG; - } - - /* Consider skipping UTF-8 byte order mark (BOM). */ - if(renderer_flags & MD_HTML_FLAG_SKIP_UTF8_BOM && sizeof(MD_CHAR) == 1) { - static const MD_CHAR bom[3] = { 0xef, 0xbb, 0xbf }; - if(input_size >= sizeof(bom) && memcmp(input, bom, sizeof(bom)) == 0) { - input += sizeof(bom); - input_size -= sizeof(bom); - } - } - - return md_parse(input, input_size, &parser, (void*) &render); -} - diff --git a/src/md4c-html.h b/src/md4c-html.h deleted file mode 100644 index 1042d29..0000000 --- a/src/md4c-html.h +++ /dev/null @@ -1,68 +0,0 @@ -/* - * MD4C: Markdown parser for C - * (http://github.com/mity/md4c) - * - * Copyright (c) 2016-2017 Martin Mitas - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS - * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ - -#ifndef MD4C_HTML_H -#define MD4C_HTML_H - -#include "md4c.h" - -#ifdef __cplusplus - extern "C" { -#endif - - -/* If set, debug output from md_parse() is sent to stderr. */ -#define MD_HTML_FLAG_DEBUG 0x0001 -#define MD_HTML_FLAG_VERBATIM_ENTITIES 0x0002 -#define MD_HTML_FLAG_SKIP_UTF8_BOM 0x0004 -#define MD_HTML_FLAG_XHTML 0x0008 - - -/* Render Markdown into HTML. - * - * Note only contents of <body> tag is generated. Caller must generate - * HTML header/footer manually before/after calling md_html(). - * - * Params input and input_size specify the Markdown input. - * Callback process_output() gets called with chunks of HTML output. - * (Typical implementation may just output the bytes to a file or append to - * some buffer). - * Param userdata is just propgated back to process_output() callback. - * Param parser_flags are flags from md4c.h propagated to md_parse(). - * Param render_flags is bitmask of MD_HTML_FLAG_xxxx. - * - * Returns -1 on error (if md_parse() fails.) - * Returns 0 on success. - */ -int md_html(const MD_CHAR* input, MD_SIZE input_size, - void (*process_output)(const MD_CHAR*, MD_SIZE, void*), - void* userdata, unsigned parser_flags, unsigned renderer_flags); - - -#ifdef __cplusplus - } /* extern "C" { */ -#endif - -#endif /* MD4C_HTML_H */ diff --git a/src/md4c-html.pc.in b/src/md4c-html.pc.in deleted file mode 100644 index 504bb52..0000000 --- a/src/md4c-html.pc.in +++ /dev/null @@ -1,13 +0,0 @@ -prefix=@CMAKE_INSTALL_PREFIX@ -exec_prefix=@CMAKE_INSTALL_PREFIX@ -libdir=${exec_prefix}/@CMAKE_INSTALL_LIBDIR@ -includedir=${prefix}/@CMAKE_INSTALL_INCLUDEDIR@ - -Name: @PROJECT_NAME@ HTML renderer -Description: Markdown to HTML converter library. -Version: @PROJECT_VERSION@ -URL: @PROJECT_URL@ - -Requires: md4c = @PROJECT_VERSION@ -Libs: -L${libdir} -lmd4c-html -Cflags: -I${includedir} diff --git a/src/md4c.c b/src/md4c.c deleted file mode 100644 index 40066b2..0000000 --- a/src/md4c.c +++ /dev/null @@ -1,6383 +0,0 @@ -/* - * MD4C: Markdown parser for C - * (http://github.com/mity/md4c) - * - * Copyright (c) 2016-2020 Martin Mitas - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS - * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ - -#include "md4c.h" - -#include <limits.h> -#include <stdio.h> -#include <stdlib.h> -#include <string.h> - - -/***************************** - *** Miscellaneous Stuff *** - *****************************/ - -#if !defined(__STDC_VERSION__) || __STDC_VERSION__ < 199409L - /* C89/90 or old compilers in general may not understand "inline". */ - #if defined __GNUC__ - #define inline __inline__ - #elif defined _MSC_VER - #define inline __inline - #else - #define inline - #endif -#endif - -/* Make the UTF-8 support the default. */ -#if !defined MD4C_USE_ASCII && !defined MD4C_USE_UTF8 && !defined MD4C_USE_UTF16 - #define MD4C_USE_UTF8 -#endif - -/* Magic for making wide literals with MD4C_USE_UTF16. */ -#ifdef _T - #undef _T -#endif -#if defined MD4C_USE_UTF16 - #define _T(x) L##x -#else - #define _T(x) x -#endif - -/* Misc. macros. */ -#define SIZEOF_ARRAY(a) (sizeof(a) / sizeof(a[0])) - -#define STRINGIZE_(x) #x -#define STRINGIZE(x) STRINGIZE_(x) - -#ifndef TRUE - #define TRUE 1 - #define FALSE 0 -#endif - -#define MD_LOG(msg) \ - do { \ - if(ctx->parser.debug_log != NULL) \ - ctx->parser.debug_log((msg), ctx->userdata); \ - } while(0) - -#ifdef DEBUG - #define MD_ASSERT(cond) \ - do { \ - if(!(cond)) { \ - MD_LOG(__FILE__ ":" STRINGIZE(__LINE__) ": " \ - "Assertion '" STRINGIZE(cond) "' failed."); \ - exit(1); \ - } \ - } while(0) - - #define MD_UNREACHABLE() MD_ASSERT(1 == 0) -#else - #ifdef __GNUC__ - #define MD_ASSERT(cond) do { if(!(cond)) __builtin_unreachable(); } while(0) - #define MD_UNREACHABLE() do { __builtin_unreachable(); } while(0) - #elif defined _MSC_VER && _MSC_VER > 120 - #define MD_ASSERT(cond) do { __assume(cond); } while(0) - #define MD_UNREACHABLE() do { __assume(0); } while(0) - #else - #define MD_ASSERT(cond) do {} while(0) - #define MD_UNREACHABLE() do {} while(0) - #endif -#endif - -/* For falling through case labels in switch statements. */ -#if defined __clang__ && __clang_major__ >= 12 - #define MD_FALLTHROUGH() __attribute__((fallthrough)) -#elif defined __GNUC__ && __GNUC__ >= 7 - #define MD_FALLTHROUGH() __attribute__((fallthrough)) -#else - #define MD_FALLTHROUGH() ((void)0) -#endif - -/* Suppress "unused parameter" warnings. */ -#define MD_UNUSED(x) ((void)x) - - -/************************ - *** Internal Types *** - ************************/ - -/* These are omnipresent so lets save some typing. */ -#define CHAR MD_CHAR -#define SZ MD_SIZE -#define OFF MD_OFFSET - -typedef struct MD_MARK_tag MD_MARK; -typedef struct MD_BLOCK_tag MD_BLOCK; -typedef struct MD_CONTAINER_tag MD_CONTAINER; -typedef struct MD_REF_DEF_tag MD_REF_DEF; - - -/* During analyzes of inline marks, we need to manage some "mark chains", - * of (yet unresolved) openers. This structure holds start/end of the chain. - * The chain internals are then realized through MD_MARK::prev and ::next. - */ -typedef struct MD_MARKCHAIN_tag MD_MARKCHAIN; -struct MD_MARKCHAIN_tag { - int head; /* Index of first mark in the chain, or -1 if empty. */ - int tail; /* Index of last mark in the chain, or -1 if empty. */ -}; - -/* Context propagated through all the parsing. */ -typedef struct MD_CTX_tag MD_CTX; -struct MD_CTX_tag { - /* Immutable stuff (parameters of md_parse()). */ - const CHAR* text; - SZ size; - MD_PARSER parser; - void* userdata; - - /* When this is true, it allows some optimizations. */ - int doc_ends_with_newline; - - /* Helper temporary growing buffer. */ - CHAR* buffer; - unsigned alloc_buffer; - - /* Reference definitions. */ - MD_REF_DEF* ref_defs; - int n_ref_defs; - int alloc_ref_defs; - void** ref_def_hashtable; - int ref_def_hashtable_size; - - /* Stack of inline/span markers. - * This is only used for parsing a single block contents but by storing it - * here we may reuse the stack for subsequent blocks; i.e. we have fewer - * (re)allocations. */ - MD_MARK* marks; - int n_marks; - int alloc_marks; - -#if defined MD4C_USE_UTF16 - char mark_char_map[128]; -#else - char mark_char_map[256]; -#endif - - /* For resolving of inline spans. */ - MD_MARKCHAIN mark_chains[13]; -#define PTR_CHAIN (ctx->mark_chains[0]) -#define TABLECELLBOUNDARIES (ctx->mark_chains[1]) -#define ASTERISK_OPENERS_extraword_mod3_0 (ctx->mark_chains[2]) -#define ASTERISK_OPENERS_extraword_mod3_1 (ctx->mark_chains[3]) -#define ASTERISK_OPENERS_extraword_mod3_2 (ctx->mark_chains[4]) -#define ASTERISK_OPENERS_intraword_mod3_0 (ctx->mark_chains[5]) -#define ASTERISK_OPENERS_intraword_mod3_1 (ctx->mark_chains[6]) -#define ASTERISK_OPENERS_intraword_mod3_2 (ctx->mark_chains[7]) -#define UNDERSCORE_OPENERS (ctx->mark_chains[8]) -#define TILDE_OPENERS_1 (ctx->mark_chains[9]) -#define TILDE_OPENERS_2 (ctx->mark_chains[10]) -#define BRACKET_OPENERS (ctx->mark_chains[11]) -#define DOLLAR_OPENERS (ctx->mark_chains[12]) -#define OPENERS_CHAIN_FIRST 2 -#define OPENERS_CHAIN_LAST 12 - - int n_table_cell_boundaries; - - /* For resolving links. */ - int unresolved_link_head; - int unresolved_link_tail; - - /* For resolving raw HTML. */ - OFF html_comment_horizon; - OFF html_proc_instr_horizon; - OFF html_decl_horizon; - OFF html_cdata_horizon; - - /* For block analysis. - * Notes: - * -- It holds MD_BLOCK as well as MD_LINE structures. After each - * MD_BLOCK, its (multiple) MD_LINE(s) follow. - * -- For MD_BLOCK_HTML and MD_BLOCK_CODE, MD_VERBATIMLINE(s) are used - * instead of MD_LINE(s). - */ - void* block_bytes; - MD_BLOCK* current_block; - int n_block_bytes; - int alloc_block_bytes; - - /* For container block analysis. */ - MD_CONTAINER* containers; - int n_containers; - int alloc_containers; - - /* Minimal indentation to call the block "indented code block". */ - unsigned code_indent_offset; - - /* Contextual info for line analysis. */ - SZ code_fence_length; /* For checking closing fence length. */ - int html_block_type; /* For checking closing raw HTML condition. */ - int last_line_has_list_loosening_effect; - int last_list_item_starts_with_two_blank_lines; -}; - -enum MD_LINETYPE_tag { - MD_LINE_BLANK, - MD_LINE_HR, - MD_LINE_ATXHEADER, - MD_LINE_SETEXTHEADER, - MD_LINE_SETEXTUNDERLINE, - MD_LINE_INDENTEDCODE, - MD_LINE_FENCEDCODE, - MD_LINE_HTML, - MD_LINE_TEXT, - MD_LINE_TABLE, - MD_LINE_TABLEUNDERLINE -}; -typedef enum MD_LINETYPE_tag MD_LINETYPE; - -typedef struct MD_LINE_ANALYSIS_tag MD_LINE_ANALYSIS; -struct MD_LINE_ANALYSIS_tag { - MD_LINETYPE type : 16; - unsigned data : 16; - OFF beg; - OFF end; - unsigned indent; /* Indentation level. */ -}; - -typedef struct MD_LINE_tag MD_LINE; -struct MD_LINE_tag { - OFF beg; - OFF end; -}; - -typedef struct MD_VERBATIMLINE_tag MD_VERBATIMLINE; -struct MD_VERBATIMLINE_tag { - OFF beg; - OFF end; - OFF indent; -}; - - -/***************** - *** Helpers *** - *****************/ - -/* Character accessors. */ -#define CH(off) (ctx->text[(off)]) -#define STR(off) (ctx->text + (off)) - -/* Character classification. - * Note we assume ASCII compatibility of code points < 128 here. */ -#define ISIN_(ch, ch_min, ch_max) ((ch_min) <= (unsigned)(ch) && (unsigned)(ch) <= (ch_max)) -#define ISANYOF_(ch, palette) ((ch) != _T('\0') && md_strchr((palette), (ch)) != NULL) -#define ISANYOF2_(ch, ch1, ch2) ((ch) == (ch1) || (ch) == (ch2)) -#define ISANYOF3_(ch, ch1, ch2, ch3) ((ch) == (ch1) || (ch) == (ch2) || (ch) == (ch3)) -#define ISASCII_(ch) ((unsigned)(ch) <= 127) -#define ISBLANK_(ch) (ISANYOF2_((ch), _T(' '), _T('\t'))) -#define ISNEWLINE_(ch) (ISANYOF2_((ch), _T('\r'), _T('\n'))) -#define ISWHITESPACE_(ch) (ISBLANK_(ch) || ISANYOF2_((ch), _T('\v'), _T('\f'))) -#define ISCNTRL_(ch) ((unsigned)(ch) <= 31 || (unsigned)(ch) == 127) -#define ISPUNCT_(ch) (ISIN_(ch, 33, 47) || ISIN_(ch, 58, 64) || ISIN_(ch, 91, 96) || ISIN_(ch, 123, 126)) -#define ISUPPER_(ch) (ISIN_(ch, _T('A'), _T('Z'))) -#define ISLOWER_(ch) (ISIN_(ch, _T('a'), _T('z'))) -#define ISALPHA_(ch) (ISUPPER_(ch) || ISLOWER_(ch)) -#define ISDIGIT_(ch) (ISIN_(ch, _T('0'), _T('9'))) -#define ISXDIGIT_(ch) (ISDIGIT_(ch) || ISIN_(ch, _T('A'), _T('F')) || ISIN_(ch, _T('a'), _T('f'))) -#define ISALNUM_(ch) (ISALPHA_(ch) || ISDIGIT_(ch)) - -#define ISANYOF(off, palette) ISANYOF_(CH(off), (palette)) -#define ISANYOF2(off, ch1, ch2) ISANYOF2_(CH(off), (ch1), (ch2)) -#define ISANYOF3(off, ch1, ch2, ch3) ISANYOF3_(CH(off), (ch1), (ch2), (ch3)) -#define ISASCII(off) ISASCII_(CH(off)) -#define ISBLANK(off) ISBLANK_(CH(off)) -#define ISNEWLINE(off) ISNEWLINE_(CH(off)) -#define ISWHITESPACE(off) ISWHITESPACE_(CH(off)) -#define ISCNTRL(off) ISCNTRL_(CH(off)) -#define ISPUNCT(off) ISPUNCT_(CH(off)) -#define ISUPPER(off) ISUPPER_(CH(off)) -#define ISLOWER(off) ISLOWER_(CH(off)) -#define ISALPHA(off) ISALPHA_(CH(off)) -#define ISDIGIT(off) ISDIGIT_(CH(off)) -#define ISXDIGIT(off) ISXDIGIT_(CH(off)) -#define ISALNUM(off) ISALNUM_(CH(off)) - - -#if defined MD4C_USE_UTF16 - #define md_strchr wcschr -#else - #define md_strchr strchr -#endif - - -/* Case insensitive check of string equality. */ -static inline int -md_ascii_case_eq(const CHAR* s1, const CHAR* s2, SZ n) -{ - OFF i; - for(i = 0; i < n; i++) { - CHAR ch1 = s1[i]; - CHAR ch2 = s2[i]; - - if(ISLOWER_(ch1)) - ch1 += ('A'-'a'); - if(ISLOWER_(ch2)) - ch2 += ('A'-'a'); - if(ch1 != ch2) - return FALSE; - } - return TRUE; -} - -static inline int -md_ascii_eq(const CHAR* s1, const CHAR* s2, SZ n) -{ - return memcmp(s1, s2, n * sizeof(CHAR)) == 0; -} - -static int -md_text_with_null_replacement(MD_CTX* ctx, MD_TEXTTYPE type, const CHAR* str, SZ size) -{ - OFF off = 0; - int ret = 0; - - while(1) { - while(off < size && str[off] != _T('\0')) - off++; - - if(off > 0) { - ret = ctx->parser.text(type, str, off, ctx->userdata); - if(ret != 0) - return ret; - - str += off; - size -= off; - off = 0; - } - - if(off >= size) - return 0; - - ret = ctx->parser.text(MD_TEXT_NULLCHAR, _T(""), 1, ctx->userdata); - if(ret != 0) - return ret; - off++; - } -} - - -#define MD_CHECK(func) \ - do { \ - ret = (func); \ - if(ret < 0) \ - goto abort; \ - } while(0) - - -#define MD_TEMP_BUFFER(sz) \ - do { \ - if(sz > ctx->alloc_buffer) { \ - CHAR* new_buffer; \ - SZ new_size = ((sz) + (sz) / 2 + 128) & ~127; \ - \ - new_buffer = realloc(ctx->buffer, new_size); \ - if(new_buffer == NULL) { \ - MD_LOG("realloc() failed."); \ - ret = -1; \ - goto abort; \ - } \ - \ - ctx->buffer = new_buffer; \ - ctx->alloc_buffer = new_size; \ - } \ - } while(0) - - -#define MD_ENTER_BLOCK(type, arg) \ - do { \ - ret = ctx->parser.enter_block((type), (arg), ctx->userdata); \ - if(ret != 0) { \ - MD_LOG("Aborted from enter_block() callback."); \ - goto abort; \ - } \ - } while(0) - -#define MD_LEAVE_BLOCK(type, arg) \ - do { \ - ret = ctx->parser.leave_block((type), (arg), ctx->userdata); \ - if(ret != 0) { \ - MD_LOG("Aborted from leave_block() callback."); \ - goto abort; \ - } \ - } while(0) - -#define MD_ENTER_SPAN(type, arg) \ - do { \ - ret = ctx->parser.enter_span((type), (arg), ctx->userdata); \ - if(ret != 0) { \ - MD_LOG("Aborted from enter_span() callback."); \ - goto abort; \ - } \ - } while(0) - -#define MD_LEAVE_SPAN(type, arg) \ - do { \ - ret = ctx->parser.leave_span((type), (arg), ctx->userdata); \ - if(ret != 0) { \ - MD_LOG("Aborted from leave_span() callback."); \ - goto abort; \ - } \ - } while(0) - -#define MD_TEXT(type, str, size) \ - do { \ - if(size > 0) { \ - ret = ctx->parser.text((type), (str), (size), ctx->userdata); \ - if(ret != 0) { \ - MD_LOG("Aborted from text() callback."); \ - goto abort; \ - } \ - } \ - } while(0) - -#define MD_TEXT_INSECURE(type, str, size) \ - do { \ - if(size > 0) { \ - ret = md_text_with_null_replacement(ctx, type, str, size); \ - if(ret != 0) { \ - MD_LOG("Aborted from text() callback."); \ - goto abort; \ - } \ - } \ - } while(0) - - - -/************************* - *** Unicode Support *** - *************************/ - -typedef struct MD_UNICODE_FOLD_INFO_tag MD_UNICODE_FOLD_INFO; -struct MD_UNICODE_FOLD_INFO_tag { - unsigned codepoints[3]; - unsigned n_codepoints; -}; - - -#if defined MD4C_USE_UTF16 || defined MD4C_USE_UTF8 - /* Binary search over sorted "map" of codepoints. Consecutive sequences - * of codepoints may be encoded in the map by just using the - * (MIN_CODEPOINT | 0x40000000) and (MAX_CODEPOINT | 0x80000000). - * - * Returns index of the found record in the map (in the case of ranges, - * the minimal value is used); or -1 on failure. */ - static int - md_unicode_bsearch__(unsigned codepoint, const unsigned* map, size_t map_size) - { - int beg, end; - int pivot_beg, pivot_end; - - beg = 0; - end = (int) map_size-1; - while(beg <= end) { - /* Pivot may be a range, not just a single value. */ - pivot_beg = pivot_end = (beg + end) / 2; - if(map[pivot_end] & 0x40000000) - pivot_end++; - if(map[pivot_beg] & 0x80000000) - pivot_beg--; - - if(codepoint < (map[pivot_beg] & 0x00ffffff)) - end = pivot_beg - 1; - else if(codepoint > (map[pivot_end] & 0x00ffffff)) - beg = pivot_end + 1; - else - return pivot_beg; - } - - return -1; - } - - static int - md_is_unicode_whitespace__(unsigned codepoint) - { -#define R(cp_min, cp_max) ((cp_min) | 0x40000000), ((cp_max) | 0x80000000) -#define S(cp) (cp) - /* Unicode "Zs" category. - * (generated by scripts/build_whitespace_map.py) */ - static const unsigned WHITESPACE_MAP[] = { - S(0x0020), S(0x00a0), S(0x1680), R(0x2000,0x200a), S(0x202f), S(0x205f), S(0x3000) - }; -#undef R -#undef S - - /* The ASCII ones are the most frequently used ones, also CommonMark - * specification requests few more in this range. */ - if(codepoint <= 0x7f) - return ISWHITESPACE_(codepoint); - - return (md_unicode_bsearch__(codepoint, WHITESPACE_MAP, SIZEOF_ARRAY(WHITESPACE_MAP)) >= 0); - } - - static int - md_is_unicode_punct__(unsigned codepoint) - { -#define R(cp_min, cp_max) ((cp_min) | 0x40000000), ((cp_max) | 0x80000000) -#define S(cp) (cp) - /* Unicode "Pc", "Pd", "Pe", "Pf", "Pi", "Po", "Ps" categories. - * (generated by scripts/build_punct_map.py) */ - static const unsigned PUNCT_MAP[] = { - R(0x0021,0x0023), R(0x0025,0x002a), R(0x002c,0x002f), R(0x003a,0x003b), R(0x003f,0x0040), - R(0x005b,0x005d), S(0x005f), S(0x007b), S(0x007d), S(0x00a1), S(0x00a7), S(0x00ab), R(0x00b6,0x00b7), - S(0x00bb), S(0x00bf), S(0x037e), S(0x0387), R(0x055a,0x055f), R(0x0589,0x058a), S(0x05be), S(0x05c0), - S(0x05c3), S(0x05c6), R(0x05f3,0x05f4), R(0x0609,0x060a), R(0x060c,0x060d), S(0x061b), R(0x061e,0x061f), - R(0x066a,0x066d), S(0x06d4), R(0x0700,0x070d), R(0x07f7,0x07f9), R(0x0830,0x083e), S(0x085e), - R(0x0964,0x0965), S(0x0970), S(0x09fd), S(0x0a76), S(0x0af0), S(0x0c77), S(0x0c84), S(0x0df4), S(0x0e4f), - R(0x0e5a,0x0e5b), R(0x0f04,0x0f12), S(0x0f14), R(0x0f3a,0x0f3d), S(0x0f85), R(0x0fd0,0x0fd4), - R(0x0fd9,0x0fda), R(0x104a,0x104f), S(0x10fb), R(0x1360,0x1368), S(0x1400), S(0x166e), R(0x169b,0x169c), - R(0x16eb,0x16ed), R(0x1735,0x1736), R(0x17d4,0x17d6), R(0x17d8,0x17da), R(0x1800,0x180a), - R(0x1944,0x1945), R(0x1a1e,0x1a1f), R(0x1aa0,0x1aa6), R(0x1aa8,0x1aad), R(0x1b5a,0x1b60), - R(0x1bfc,0x1bff), R(0x1c3b,0x1c3f), R(0x1c7e,0x1c7f), R(0x1cc0,0x1cc7), S(0x1cd3), R(0x2010,0x2027), - R(0x2030,0x2043), R(0x2045,0x2051), R(0x2053,0x205e), R(0x207d,0x207e), R(0x208d,0x208e), - R(0x2308,0x230b), R(0x2329,0x232a), R(0x2768,0x2775), R(0x27c5,0x27c6), R(0x27e6,0x27ef), - R(0x2983,0x2998), R(0x29d8,0x29db), R(0x29fc,0x29fd), R(0x2cf9,0x2cfc), R(0x2cfe,0x2cff), S(0x2d70), - R(0x2e00,0x2e2e), R(0x2e30,0x2e4f), S(0x2e52), R(0x3001,0x3003), R(0x3008,0x3011), R(0x3014,0x301f), - S(0x3030), S(0x303d), S(0x30a0), S(0x30fb), R(0xa4fe,0xa4ff), R(0xa60d,0xa60f), S(0xa673), S(0xa67e), - R(0xa6f2,0xa6f7), R(0xa874,0xa877), R(0xa8ce,0xa8cf), R(0xa8f8,0xa8fa), S(0xa8fc), R(0xa92e,0xa92f), - S(0xa95f), R(0xa9c1,0xa9cd), R(0xa9de,0xa9df), R(0xaa5c,0xaa5f), R(0xaade,0xaadf), R(0xaaf0,0xaaf1), - S(0xabeb), R(0xfd3e,0xfd3f), R(0xfe10,0xfe19), R(0xfe30,0xfe52), R(0xfe54,0xfe61), S(0xfe63), S(0xfe68), - R(0xfe6a,0xfe6b), R(0xff01,0xff03), R(0xff05,0xff0a), R(0xff0c,0xff0f), R(0xff1a,0xff1b), - R(0xff1f,0xff20), R(0xff3b,0xff3d), S(0xff3f), S(0xff5b), S(0xff5d), R(0xff5f,0xff65), R(0x10100,0x10102), - S(0x1039f), S(0x103d0), S(0x1056f), S(0x10857), S(0x1091f), S(0x1093f), R(0x10a50,0x10a58), S(0x10a7f), - R(0x10af0,0x10af6), R(0x10b39,0x10b3f), R(0x10b99,0x10b9c), S(0x10ead), R(0x10f55,0x10f59), - R(0x11047,0x1104d), R(0x110bb,0x110bc), R(0x110be,0x110c1), R(0x11140,0x11143), R(0x11174,0x11175), - R(0x111c5,0x111c8), S(0x111cd), S(0x111db), R(0x111dd,0x111df), R(0x11238,0x1123d), S(0x112a9), - R(0x1144b,0x1144f), R(0x1145a,0x1145b), S(0x1145d), S(0x114c6), R(0x115c1,0x115d7), R(0x11641,0x11643), - R(0x11660,0x1166c), R(0x1173c,0x1173e), S(0x1183b), R(0x11944,0x11946), S(0x119e2), R(0x11a3f,0x11a46), - R(0x11a9a,0x11a9c), R(0x11a9e,0x11aa2), R(0x11c41,0x11c45), R(0x11c70,0x11c71), R(0x11ef7,0x11ef8), - S(0x11fff), R(0x12470,0x12474), R(0x16a6e,0x16a6f), S(0x16af5), R(0x16b37,0x16b3b), S(0x16b44), - R(0x16e97,0x16e9a), S(0x16fe2), S(0x1bc9f), R(0x1da87,0x1da8b), R(0x1e95e,0x1e95f) - }; -#undef R -#undef S - - /* The ASCII ones are the most frequently used ones, also CommonMark - * specification requests few more in this range. */ - if(codepoint <= 0x7f) - return ISPUNCT_(codepoint); - - return (md_unicode_bsearch__(codepoint, PUNCT_MAP, SIZEOF_ARRAY(PUNCT_MAP)) >= 0); - } - - static void - md_get_unicode_fold_info(unsigned codepoint, MD_UNICODE_FOLD_INFO* info) - { -#define R(cp_min, cp_max) ((cp_min) | 0x40000000), ((cp_max) | 0x80000000) -#define S(cp) (cp) - /* Unicode "Pc", "Pd", "Pe", "Pf", "Pi", "Po", "Ps" categories. - * (generated by scripts/build_folding_map.py) */ - static const unsigned FOLD_MAP_1[] = { - R(0x0041,0x005a), S(0x00b5), R(0x00c0,0x00d6), R(0x00d8,0x00de), R(0x0100,0x012e), R(0x0132,0x0136), - R(0x0139,0x0147), R(0x014a,0x0176), S(0x0178), R(0x0179,0x017d), S(0x017f), S(0x0181), S(0x0182), - S(0x0184), S(0x0186), S(0x0187), S(0x0189), S(0x018a), S(0x018b), S(0x018e), S(0x018f), S(0x0190), - S(0x0191), S(0x0193), S(0x0194), S(0x0196), S(0x0197), S(0x0198), S(0x019c), S(0x019d), S(0x019f), - R(0x01a0,0x01a4), S(0x01a6), S(0x01a7), S(0x01a9), S(0x01ac), S(0x01ae), S(0x01af), S(0x01b1), S(0x01b2), - S(0x01b3), S(0x01b5), S(0x01b7), S(0x01b8), S(0x01bc), S(0x01c4), S(0x01c5), S(0x01c7), S(0x01c8), - S(0x01ca), R(0x01cb,0x01db), R(0x01de,0x01ee), S(0x01f1), S(0x01f2), S(0x01f4), S(0x01f6), S(0x01f7), - R(0x01f8,0x021e), S(0x0220), R(0x0222,0x0232), S(0x023a), S(0x023b), S(0x023d), S(0x023e), S(0x0241), - S(0x0243), S(0x0244), S(0x0245), R(0x0246,0x024e), S(0x0345), S(0x0370), S(0x0372), S(0x0376), S(0x037f), - S(0x0386), R(0x0388,0x038a), S(0x038c), S(0x038e), S(0x038f), R(0x0391,0x03a1), R(0x03a3,0x03ab), - S(0x03c2), S(0x03cf), S(0x03d0), S(0x03d1), S(0x03d5), S(0x03d6), R(0x03d8,0x03ee), S(0x03f0), S(0x03f1), - S(0x03f4), S(0x03f5), S(0x03f7), S(0x03f9), S(0x03fa), R(0x03fd,0x03ff), R(0x0400,0x040f), - R(0x0410,0x042f), R(0x0460,0x0480), R(0x048a,0x04be), S(0x04c0), R(0x04c1,0x04cd), R(0x04d0,0x052e), - R(0x0531,0x0556), R(0x10a0,0x10c5), S(0x10c7), S(0x10cd), R(0x13f8,0x13fd), S(0x1c80), S(0x1c81), - S(0x1c82), S(0x1c83), S(0x1c84), S(0x1c85), S(0x1c86), S(0x1c87), S(0x1c88), R(0x1c90,0x1cba), - R(0x1cbd,0x1cbf), R(0x1e00,0x1e94), S(0x1e9b), R(0x1ea0,0x1efe), R(0x1f08,0x1f0f), R(0x1f18,0x1f1d), - R(0x1f28,0x1f2f), R(0x1f38,0x1f3f), R(0x1f48,0x1f4d), S(0x1f59), S(0x1f5b), S(0x1f5d), S(0x1f5f), - R(0x1f68,0x1f6f), S(0x1fb8), S(0x1fb9), S(0x1fba), S(0x1fbb), S(0x1fbe), R(0x1fc8,0x1fcb), S(0x1fd8), - S(0x1fd9), S(0x1fda), S(0x1fdb), S(0x1fe8), S(0x1fe9), S(0x1fea), S(0x1feb), S(0x1fec), S(0x1ff8), - S(0x1ff9), S(0x1ffa), S(0x1ffb), S(0x2126), S(0x212a), S(0x212b), S(0x2132), R(0x2160,0x216f), S(0x2183), - R(0x24b6,0x24cf), R(0x2c00,0x2c2e), S(0x2c60), S(0x2c62), S(0x2c63), S(0x2c64), R(0x2c67,0x2c6b), - S(0x2c6d), S(0x2c6e), S(0x2c6f), S(0x2c70), S(0x2c72), S(0x2c75), S(0x2c7e), S(0x2c7f), R(0x2c80,0x2ce2), - S(0x2ceb), S(0x2ced), S(0x2cf2), R(0xa640,0xa66c), R(0xa680,0xa69a), R(0xa722,0xa72e), R(0xa732,0xa76e), - S(0xa779), S(0xa77b), S(0xa77d), R(0xa77e,0xa786), S(0xa78b), S(0xa78d), S(0xa790), S(0xa792), - R(0xa796,0xa7a8), S(0xa7aa), S(0xa7ab), S(0xa7ac), S(0xa7ad), S(0xa7ae), S(0xa7b0), S(0xa7b1), S(0xa7b2), - S(0xa7b3), R(0xa7b4,0xa7be), S(0xa7c2), S(0xa7c4), S(0xa7c5), S(0xa7c6), S(0xa7c7), S(0xa7c9), S(0xa7f5), - R(0xab70,0xabbf), R(0xff21,0xff3a), R(0x10400,0x10427), R(0x104b0,0x104d3), R(0x10c80,0x10cb2), - R(0x118a0,0x118bf), R(0x16e40,0x16e5f), R(0x1e900,0x1e921) - }; - static const unsigned FOLD_MAP_1_DATA[] = { - 0x0061, 0x007a, 0x03bc, 0x00e0, 0x00f6, 0x00f8, 0x00fe, 0x0101, 0x012f, 0x0133, 0x0137, 0x013a, 0x0148, - 0x014b, 0x0177, 0x00ff, 0x017a, 0x017e, 0x0073, 0x0253, 0x0183, 0x0185, 0x0254, 0x0188, 0x0256, 0x0257, - 0x018c, 0x01dd, 0x0259, 0x025b, 0x0192, 0x0260, 0x0263, 0x0269, 0x0268, 0x0199, 0x026f, 0x0272, 0x0275, - 0x01a1, 0x01a5, 0x0280, 0x01a8, 0x0283, 0x01ad, 0x0288, 0x01b0, 0x028a, 0x028b, 0x01b4, 0x01b6, 0x0292, - 0x01b9, 0x01bd, 0x01c6, 0x01c6, 0x01c9, 0x01c9, 0x01cc, 0x01cc, 0x01dc, 0x01df, 0x01ef, 0x01f3, 0x01f3, - 0x01f5, 0x0195, 0x01bf, 0x01f9, 0x021f, 0x019e, 0x0223, 0x0233, 0x2c65, 0x023c, 0x019a, 0x2c66, 0x0242, - 0x0180, 0x0289, 0x028c, 0x0247, 0x024f, 0x03b9, 0x0371, 0x0373, 0x0377, 0x03f3, 0x03ac, 0x03ad, 0x03af, - 0x03cc, 0x03cd, 0x03ce, 0x03b1, 0x03c1, 0x03c3, 0x03cb, 0x03c3, 0x03d7, 0x03b2, 0x03b8, 0x03c6, 0x03c0, - 0x03d9, 0x03ef, 0x03ba, 0x03c1, 0x03b8, 0x03b5, 0x03f8, 0x03f2, 0x03fb, 0x037b, 0x037d, 0x0450, 0x045f, - 0x0430, 0x044f, 0x0461, 0x0481, 0x048b, 0x04bf, 0x04cf, 0x04c2, 0x04ce, 0x04d1, 0x052f, 0x0561, 0x0586, - 0x2d00, 0x2d25, 0x2d27, 0x2d2d, 0x13f0, 0x13f5, 0x0432, 0x0434, 0x043e, 0x0441, 0x0442, 0x0442, 0x044a, - 0x0463, 0xa64b, 0x10d0, 0x10fa, 0x10fd, 0x10ff, 0x1e01, 0x1e95, 0x1e61, 0x1ea1, 0x1eff, 0x1f00, 0x1f07, - 0x1f10, 0x1f15, 0x1f20, 0x1f27, 0x1f30, 0x1f37, 0x1f40, 0x1f45, 0x1f51, 0x1f53, 0x1f55, 0x1f57, 0x1f60, - 0x1f67, 0x1fb0, 0x1fb1, 0x1f70, 0x1f71, 0x03b9, 0x1f72, 0x1f75, 0x1fd0, 0x1fd1, 0x1f76, 0x1f77, 0x1fe0, - 0x1fe1, 0x1f7a, 0x1f7b, 0x1fe5, 0x1f78, 0x1f79, 0x1f7c, 0x1f7d, 0x03c9, 0x006b, 0x00e5, 0x214e, 0x2170, - 0x217f, 0x2184, 0x24d0, 0x24e9, 0x2c30, 0x2c5e, 0x2c61, 0x026b, 0x1d7d, 0x027d, 0x2c68, 0x2c6c, 0x0251, - 0x0271, 0x0250, 0x0252, 0x2c73, 0x2c76, 0x023f, 0x0240, 0x2c81, 0x2ce3, 0x2cec, 0x2cee, 0x2cf3, 0xa641, - 0xa66d, 0xa681, 0xa69b, 0xa723, 0xa72f, 0xa733, 0xa76f, 0xa77a, 0xa77c, 0x1d79, 0xa77f, 0xa787, 0xa78c, - 0x0265, 0xa791, 0xa793, 0xa797, 0xa7a9, 0x0266, 0x025c, 0x0261, 0x026c, 0x026a, 0x029e, 0x0287, 0x029d, - 0xab53, 0xa7b5, 0xa7bf, 0xa7c3, 0xa794, 0x0282, 0x1d8e, 0xa7c8, 0xa7ca, 0xa7f6, 0x13a0, 0x13ef, 0xff41, - 0xff5a, 0x10428, 0x1044f, 0x104d8, 0x104fb, 0x10cc0, 0x10cf2, 0x118c0, 0x118df, 0x16e60, 0x16e7f, 0x1e922, - 0x1e943 - }; - static const unsigned FOLD_MAP_2[] = { - S(0x00df), S(0x0130), S(0x0149), S(0x01f0), S(0x0587), S(0x1e96), S(0x1e97), S(0x1e98), S(0x1e99), - S(0x1e9a), S(0x1e9e), S(0x1f50), R(0x1f80,0x1f87), R(0x1f88,0x1f8f), R(0x1f90,0x1f97), R(0x1f98,0x1f9f), - R(0x1fa0,0x1fa7), R(0x1fa8,0x1faf), S(0x1fb2), S(0x1fb3), S(0x1fb4), S(0x1fb6), S(0x1fbc), S(0x1fc2), - S(0x1fc3), S(0x1fc4), S(0x1fc6), S(0x1fcc), S(0x1fd6), S(0x1fe4), S(0x1fe6), S(0x1ff2), S(0x1ff3), - S(0x1ff4), S(0x1ff6), S(0x1ffc), S(0xfb00), S(0xfb01), S(0xfb02), S(0xfb05), S(0xfb06), S(0xfb13), - S(0xfb14), S(0xfb15), S(0xfb16), S(0xfb17) - }; - static const unsigned FOLD_MAP_2_DATA[] = { - 0x0073,0x0073, 0x0069,0x0307, 0x02bc,0x006e, 0x006a,0x030c, 0x0565,0x0582, 0x0068,0x0331, 0x0074,0x0308, - 0x0077,0x030a, 0x0079,0x030a, 0x0061,0x02be, 0x0073,0x0073, 0x03c5,0x0313, 0x1f00,0x03b9, 0x1f07,0x03b9, - 0x1f00,0x03b9, 0x1f07,0x03b9, 0x1f20,0x03b9, 0x1f27,0x03b9, 0x1f20,0x03b9, 0x1f27,0x03b9, 0x1f60,0x03b9, - 0x1f67,0x03b9, 0x1f60,0x03b9, 0x1f67,0x03b9, 0x1f70,0x03b9, 0x03b1,0x03b9, 0x03ac,0x03b9, 0x03b1,0x0342, - 0x03b1,0x03b9, 0x1f74,0x03b9, 0x03b7,0x03b9, 0x03ae,0x03b9, 0x03b7,0x0342, 0x03b7,0x03b9, 0x03b9,0x0342, - 0x03c1,0x0313, 0x03c5,0x0342, 0x1f7c,0x03b9, 0x03c9,0x03b9, 0x03ce,0x03b9, 0x03c9,0x0342, 0x03c9,0x03b9, - 0x0066,0x0066, 0x0066,0x0069, 0x0066,0x006c, 0x0073,0x0074, 0x0073,0x0074, 0x0574,0x0576, 0x0574,0x0565, - 0x0574,0x056b, 0x057e,0x0576, 0x0574,0x056d - }; - static const unsigned FOLD_MAP_3[] = { - S(0x0390), S(0x03b0), S(0x1f52), S(0x1f54), S(0x1f56), S(0x1fb7), S(0x1fc7), S(0x1fd2), S(0x1fd3), - S(0x1fd7), S(0x1fe2), S(0x1fe3), S(0x1fe7), S(0x1ff7), S(0xfb03), S(0xfb04) - }; - static const unsigned FOLD_MAP_3_DATA[] = { - 0x03b9,0x0308,0x0301, 0x03c5,0x0308,0x0301, 0x03c5,0x0313,0x0300, 0x03c5,0x0313,0x0301, - 0x03c5,0x0313,0x0342, 0x03b1,0x0342,0x03b9, 0x03b7,0x0342,0x03b9, 0x03b9,0x0308,0x0300, - 0x03b9,0x0308,0x0301, 0x03b9,0x0308,0x0342, 0x03c5,0x0308,0x0300, 0x03c5,0x0308,0x0301, - 0x03c5,0x0308,0x0342, 0x03c9,0x0342,0x03b9, 0x0066,0x0066,0x0069, 0x0066,0x0066,0x006c - }; -#undef R -#undef S - static const struct { - const unsigned* map; - const unsigned* data; - size_t map_size; - unsigned n_codepoints; - } FOLD_MAP_LIST[] = { - { FOLD_MAP_1, FOLD_MAP_1_DATA, SIZEOF_ARRAY(FOLD_MAP_1), 1 }, - { FOLD_MAP_2, FOLD_MAP_2_DATA, SIZEOF_ARRAY(FOLD_MAP_2), 2 }, - { FOLD_MAP_3, FOLD_MAP_3_DATA, SIZEOF_ARRAY(FOLD_MAP_3), 3 } - }; - - int i; - - /* Fast path for ASCII characters. */ - if(codepoint <= 0x7f) { - info->codepoints[0] = codepoint; - if(ISUPPER_(codepoint)) - info->codepoints[0] += 'a' - 'A'; - info->n_codepoints = 1; - return; - } - - /* Try to locate the codepoint in any of the maps. */ - for(i = 0; i < (int) SIZEOF_ARRAY(FOLD_MAP_LIST); i++) { - int index; - - index = md_unicode_bsearch__(codepoint, FOLD_MAP_LIST[i].map, FOLD_MAP_LIST[i].map_size); - if(index >= 0) { - /* Found the mapping. */ - unsigned n_codepoints = FOLD_MAP_LIST[i].n_codepoints; - const unsigned* map = FOLD_MAP_LIST[i].map; - const unsigned* codepoints = FOLD_MAP_LIST[i].data + (index * n_codepoints); - - memcpy(info->codepoints, codepoints, sizeof(unsigned) * n_codepoints); - info->n_codepoints = n_codepoints; - - if(FOLD_MAP_LIST[i].map[index] != codepoint) { - /* The found mapping maps whole range of codepoints, - * i.e. we have to offset info->codepoints[0] accordingly. */ - if((map[index] & 0x00ffffff)+1 == codepoints[0]) { - /* Alternating type of the range. */ - info->codepoints[0] = codepoint + ((codepoint & 0x1) == (map[index] & 0x1) ? 1 : 0); - } else { - /* Range to range kind of mapping. */ - info->codepoints[0] += (codepoint - (map[index] & 0x00ffffff)); - } - } - - return; - } - } - - /* No mapping found. Map the codepoint to itself. */ - info->codepoints[0] = codepoint; - info->n_codepoints = 1; - } -#endif - - -#if defined MD4C_USE_UTF16 - #define IS_UTF16_SURROGATE_HI(word) (((WORD)(word) & 0xfc00) == 0xd800) - #define IS_UTF16_SURROGATE_LO(word) (((WORD)(word) & 0xfc00) == 0xdc00) - #define UTF16_DECODE_SURROGATE(hi, lo) (0x10000 + ((((unsigned)(hi) & 0x3ff) << 10) | (((unsigned)(lo) & 0x3ff) << 0))) - - static unsigned - md_decode_utf16le__(const CHAR* str, SZ str_size, SZ* p_size) - { - if(IS_UTF16_SURROGATE_HI(str[0])) { - if(1 < str_size && IS_UTF16_SURROGATE_LO(str[1])) { - if(p_size != NULL) - *p_size = 2; - return UTF16_DECODE_SURROGATE(str[0], str[1]); - } - } - - if(p_size != NULL) - *p_size = 1; - return str[0]; - } - - static unsigned - md_decode_utf16le_before__(MD_CTX* ctx, OFF off) - { - if(off > 2 && IS_UTF16_SURROGATE_HI(CH(off-2)) && IS_UTF16_SURROGATE_LO(CH(off-1))) - return UTF16_DECODE_SURROGATE(CH(off-2), CH(off-1)); - - return CH(off); - } - - /* No whitespace uses surrogates, so no decoding needed here. */ - #define ISUNICODEWHITESPACE_(codepoint) md_is_unicode_whitespace__(codepoint) - #define ISUNICODEWHITESPACE(off) md_is_unicode_whitespace__(CH(off)) - #define ISUNICODEWHITESPACEBEFORE(off) md_is_unicode_whitespace__(CH((off)-1)) - - #define ISUNICODEPUNCT(off) md_is_unicode_punct__(md_decode_utf16le__(STR(off), ctx->size - (off), NULL)) - #define ISUNICODEPUNCTBEFORE(off) md_is_unicode_punct__(md_decode_utf16le_before__(ctx, off)) - - static inline int - md_decode_unicode(const CHAR* str, OFF off, SZ str_size, SZ* p_char_size) - { - return md_decode_utf16le__(str+off, str_size-off, p_char_size); - } -#elif defined MD4C_USE_UTF8 - #define IS_UTF8_LEAD1(byte) ((unsigned char)(byte) <= 0x7f) - #define IS_UTF8_LEAD2(byte) (((unsigned char)(byte) & 0xe0) == 0xc0) - #define IS_UTF8_LEAD3(byte) (((unsigned char)(byte) & 0xf0) == 0xe0) - #define IS_UTF8_LEAD4(byte) (((unsigned char)(byte) & 0xf8) == 0xf0) - #define IS_UTF8_TAIL(byte) (((unsigned char)(byte) & 0xc0) == 0x80) - - static unsigned - md_decode_utf8__(const CHAR* str, SZ str_size, SZ* p_size) - { - if(!IS_UTF8_LEAD1(str[0])) { - if(IS_UTF8_LEAD2(str[0])) { - if(1 < str_size && IS_UTF8_TAIL(str[1])) { - if(p_size != NULL) - *p_size = 2; - - return (((unsigned int)str[0] & 0x1f) << 6) | - (((unsigned int)str[1] & 0x3f) << 0); - } - } else if(IS_UTF8_LEAD3(str[0])) { - if(2 < str_size && IS_UTF8_TAIL(str[1]) && IS_UTF8_TAIL(str[2])) { - if(p_size != NULL) - *p_size = 3; - - return (((unsigned int)str[0] & 0x0f) << 12) | - (((unsigned int)str[1] & 0x3f) << 6) | - (((unsigned int)str[2] & 0x3f) << 0); - } - } else if(IS_UTF8_LEAD4(str[0])) { - if(3 < str_size && IS_UTF8_TAIL(str[1]) && IS_UTF8_TAIL(str[2]) && IS_UTF8_TAIL(str[3])) { - if(p_size != NULL) - *p_size = 4; - - return (((unsigned int)str[0] & 0x07) << 18) | - (((unsigned int)str[1] & 0x3f) << 12) | - (((unsigned int)str[2] & 0x3f) << 6) | - (((unsigned int)str[3] & 0x3f) << 0); - } - } - } - - if(p_size != NULL) - *p_size = 1; - return (unsigned) str[0]; - } - - static unsigned - md_decode_utf8_before__(MD_CTX* ctx, OFF off) - { - if(!IS_UTF8_LEAD1(CH(off-1))) { - if(off > 1 && IS_UTF8_LEAD2(CH(off-2)) && IS_UTF8_TAIL(CH(off-1))) - return (((unsigned int)CH(off-2) & 0x1f) << 6) | - (((unsigned int)CH(off-1) & 0x3f) << 0); - - if(off > 2 && IS_UTF8_LEAD3(CH(off-3)) && IS_UTF8_TAIL(CH(off-2)) && IS_UTF8_TAIL(CH(off-1))) - return (((unsigned int)CH(off-3) & 0x0f) << 12) | - (((unsigned int)CH(off-2) & 0x3f) << 6) | - (((unsigned int)CH(off-1) & 0x3f) << 0); - - if(off > 3 && IS_UTF8_LEAD4(CH(off-4)) && IS_UTF8_TAIL(CH(off-3)) && IS_UTF8_TAIL(CH(off-2)) && IS_UTF8_TAIL(CH(off-1))) - return (((unsigned int)CH(off-4) & 0x07) << 18) | - (((unsigned int)CH(off-3) & 0x3f) << 12) | - (((unsigned int)CH(off-2) & 0x3f) << 6) | - (((unsigned int)CH(off-1) & 0x3f) << 0); - } - - return (unsigned) CH(off-1); - } - - #define ISUNICODEWHITESPACE_(codepoint) md_is_unicode_whitespace__(codepoint) - #define ISUNICODEWHITESPACE(off) md_is_unicode_whitespace__(md_decode_utf8__(STR(off), ctx->size - (off), NULL)) - #define ISUNICODEWHITESPACEBEFORE(off) md_is_unicode_whitespace__(md_decode_utf8_before__(ctx, off)) - - #define ISUNICODEPUNCT(off) md_is_unicode_punct__(md_decode_utf8__(STR(off), ctx->size - (off), NULL)) - #define ISUNICODEPUNCTBEFORE(off) md_is_unicode_punct__(md_decode_utf8_before__(ctx, off)) - - static inline unsigned - md_decode_unicode(const CHAR* str, OFF off, SZ str_size, SZ* p_char_size) - { - return md_decode_utf8__(str+off, str_size-off, p_char_size); - } -#else - #define ISUNICODEWHITESPACE_(codepoint) ISWHITESPACE_(codepoint) - #define ISUNICODEWHITESPACE(off) ISWHITESPACE(off) - #define ISUNICODEWHITESPACEBEFORE(off) ISWHITESPACE((off)-1) - - #define ISUNICODEPUNCT(off) ISPUNCT(off) - #define ISUNICODEPUNCTBEFORE(off) ISPUNCT((off)-1) - - static inline void - md_get_unicode_fold_info(unsigned codepoint, MD_UNICODE_FOLD_INFO* info) - { - info->codepoints[0] = codepoint; - if(ISUPPER_(codepoint)) - info->codepoints[0] += 'a' - 'A'; - info->n_codepoints = 1; - } - - static inline unsigned - md_decode_unicode(const CHAR* str, OFF off, SZ str_size, SZ* p_size) - { - *p_size = 1; - return (unsigned) str[off]; - } -#endif - - -/************************************* - *** Helper string manipulations *** - *************************************/ - -/* Fill buffer with copy of the string between 'beg' and 'end' but replace any - * line breaks with given replacement character. - * - * NOTE: Caller is responsible to make sure the buffer is large enough. - * (Given the output is always shorter then input, (end - beg) is good idea - * what the caller should allocate.) - */ -static void -md_merge_lines(MD_CTX* ctx, OFF beg, OFF end, const MD_LINE* lines, int n_lines, - CHAR line_break_replacement_char, CHAR* buffer, SZ* p_size) -{ - CHAR* ptr = buffer; - int line_index = 0; - OFF off = beg; - - MD_UNUSED(n_lines); - - while(1) { - const MD_LINE* line = &lines[line_index]; - OFF line_end = line->end; - if(end < line_end) - line_end = end; - - while(off < line_end) { - *ptr = CH(off); - ptr++; - off++; - } - - if(off >= end) { - *p_size = (MD_SIZE)(ptr - buffer); - return; - } - - *ptr = line_break_replacement_char; - ptr++; - - line_index++; - off = lines[line_index].beg; - } -} - -/* Wrapper of md_merge_lines() which allocates new buffer for the output string. - */ -static int -md_merge_lines_alloc(MD_CTX* ctx, OFF beg, OFF end, const MD_LINE* lines, int n_lines, - CHAR line_break_replacement_char, CHAR** p_str, SZ* p_size) -{ - CHAR* buffer; - - buffer = (CHAR*) malloc(sizeof(CHAR) * (end - beg)); - if(buffer == NULL) { - MD_LOG("malloc() failed."); - return -1; - } - - md_merge_lines(ctx, beg, end, lines, n_lines, - line_break_replacement_char, buffer, p_size); - - *p_str = buffer; - return 0; -} - -static OFF -md_skip_unicode_whitespace(const CHAR* label, OFF off, SZ size) -{ - SZ char_size; - unsigned codepoint; - - while(off < size) { - codepoint = md_decode_unicode(label, off, size, &char_size); - if(!ISUNICODEWHITESPACE_(codepoint) && !ISNEWLINE_(label[off])) - break; - off += char_size; - } - - return off; -} - - -/****************************** - *** Recognizing raw HTML *** - ******************************/ - -/* md_is_html_tag() may be called when processing inlines (inline raw HTML) - * or when breaking document to blocks (checking for start of HTML block type 7). - * - * When breaking document to blocks, we do not yet know line boundaries, but - * in that case the whole tag has to live on a single line. We distinguish this - * by n_lines == 0. - */ -static int -md_is_html_tag(MD_CTX* ctx, const MD_LINE* lines, int n_lines, OFF beg, OFF max_end, OFF* p_end) -{ - int attr_state; - OFF off = beg; - OFF line_end = (n_lines > 0) ? lines[0].end : ctx->size; - int i = 0; - - MD_ASSERT(CH(beg) == _T('<')); - - if(off + 1 >= line_end) - return FALSE; - off++; - - /* For parsing attributes, we need a little state automaton below. - * State -1: no attributes are allowed. - * State 0: attribute could follow after some whitespace. - * State 1: after a whitespace (attribute name may follow). - * State 2: after attribute name ('=' MAY follow). - * State 3: after '=' (value specification MUST follow). - * State 41: in middle of unquoted attribute value. - * State 42: in middle of single-quoted attribute value. - * State 43: in middle of double-quoted attribute value. - */ - attr_state = 0; - - if(CH(off) == _T('/')) { - /* Closer tag "</ ... >". No attributes may be present. */ - attr_state = -1; - off++; - } - - /* Tag name */ - if(off >= line_end || !ISALPHA(off)) - return FALSE; - off++; - while(off < line_end && (ISALNUM(off) || CH(off) == _T('-'))) - off++; - - /* (Optional) attributes (if not closer), (optional) '/' (if not closer) - * and final '>'. */ - while(1) { - while(off < line_end && !ISNEWLINE(off)) { - if(attr_state > 40) { - if(attr_state == 41 && (ISBLANK(off) || ISANYOF(off, _T("\"'=<>`")))) { - attr_state = 0; - off--; /* Put the char back for re-inspection in the new state. */ - } else if(attr_state == 42 && CH(off) == _T('\'')) { - attr_state = 0; - } else if(attr_state == 43 && CH(off) == _T('"')) { - attr_state = 0; - } - off++; - } else if(ISWHITESPACE(off)) { - if(attr_state == 0) - attr_state = 1; - off++; - } else if(attr_state <= 2 && CH(off) == _T('>')) { - /* End. */ - goto done; - } else if(attr_state <= 2 && CH(off) == _T('/') && off+1 < line_end && CH(off+1) == _T('>')) { - /* End with digraph '/>' */ - off++; - goto done; - } else if((attr_state == 1 || attr_state == 2) && (ISALPHA(off) || CH(off) == _T('_') || CH(off) == _T(':'))) { - off++; - /* Attribute name */ - while(off < line_end && (ISALNUM(off) || ISANYOF(off, _T("_.:-")))) - off++; - attr_state = 2; - } else if(attr_state == 2 && CH(off) == _T('=')) { - /* Attribute assignment sign */ - off++; - attr_state = 3; - } else if(attr_state == 3) { - /* Expecting start of attribute value. */ - if(CH(off) == _T('"')) - attr_state = 43; - else if(CH(off) == _T('\'')) - attr_state = 42; - else if(!ISANYOF(off, _T("\"'=<>`")) && !ISNEWLINE(off)) - attr_state = 41; - else - return FALSE; - off++; - } else { - /* Anything unexpected. */ - return FALSE; - } - } - - /* We have to be on a single line. See definition of start condition - * of HTML block, type 7. */ - if(n_lines == 0) - return FALSE; - - i++; - if(i >= n_lines) - return FALSE; - - off = lines[i].beg; - line_end = lines[i].end; - - if(attr_state == 0 || attr_state == 41) - attr_state = 1; - - if(off >= max_end) - return FALSE; - } - -done: - if(off >= max_end) - return FALSE; - - *p_end = off+1; - return TRUE; -} - -static int -md_scan_for_html_closer(MD_CTX* ctx, const MD_CHAR* str, MD_SIZE len, - const MD_LINE* lines, int n_lines, - OFF beg, OFF max_end, OFF* p_end, - OFF* p_scan_horizon) -{ - OFF off = beg; - int i = 0; - - if(off < *p_scan_horizon && *p_scan_horizon >= max_end - len) { - /* We have already scanned the range up to the max_end so we know - * there is nothing to see. */ - return FALSE; - } - - while(TRUE) { - while(off + len <= lines[i].end && off + len <= max_end) { - if(md_ascii_eq(STR(off), str, len)) { - /* Success. */ - *p_end = off + len; - return TRUE; - } - off++; - } - - i++; - if(off >= max_end || i >= n_lines) { - /* Failure. */ - *p_scan_horizon = off; - return FALSE; - } - - off = lines[i].beg; - } -} - -static int -md_is_html_comment(MD_CTX* ctx, const MD_LINE* lines, int n_lines, OFF beg, OFF max_end, OFF* p_end) -{ - OFF off = beg; - - MD_ASSERT(CH(beg) == _T('<')); - - if(off + 4 >= lines[0].end) - return FALSE; - if(CH(off+1) != _T('!') || CH(off+2) != _T('-') || CH(off+3) != _T('-')) - return FALSE; - off += 4; - - /* ">" and "->" must not follow the opening. */ - if(off < lines[0].end && CH(off) == _T('>')) - return FALSE; - if(off+1 < lines[0].end && CH(off) == _T('-') && CH(off+1) == _T('>')) - return FALSE; - - /* HTML comment must not contain "--", so we scan just for "--" instead - * of "-->" and verify manually that '>' follows. */ - if(md_scan_for_html_closer(ctx, _T("--"), 2, - lines, n_lines, off, max_end, p_end, &ctx->html_comment_horizon)) - { - if(*p_end < max_end && CH(*p_end) == _T('>')) { - *p_end = *p_end + 1; - return TRUE; - } - } - - return FALSE; -} - -static int -md_is_html_processing_instruction(MD_CTX* ctx, const MD_LINE* lines, int n_lines, OFF beg, OFF max_end, OFF* p_end) -{ - OFF off = beg; - - if(off + 2 >= lines[0].end) - return FALSE; - if(CH(off+1) != _T('?')) - return FALSE; - off += 2; - - return md_scan_for_html_closer(ctx, _T("?>"), 2, - lines, n_lines, off, max_end, p_end, &ctx->html_proc_instr_horizon); -} - -static int -md_is_html_declaration(MD_CTX* ctx, const MD_LINE* lines, int n_lines, OFF beg, OFF max_end, OFF* p_end) -{ - OFF off = beg; - - if(off + 2 >= lines[0].end) - return FALSE; - if(CH(off+1) != _T('!')) - return FALSE; - off += 2; - - /* Declaration name. */ - if(off >= lines[0].end || !ISALPHA(off)) - return FALSE; - off++; - while(off < lines[0].end && ISALPHA(off)) - off++; - if(off < lines[0].end && !ISWHITESPACE(off)) - return FALSE; - - return md_scan_for_html_closer(ctx, _T(">"), 1, - lines, n_lines, off, max_end, p_end, &ctx->html_decl_horizon); -} - -static int -md_is_html_cdata(MD_CTX* ctx, const MD_LINE* lines, int n_lines, OFF beg, OFF max_end, OFF* p_end) -{ - static const CHAR open_str[] = _T("<![CDATA["); - static const SZ open_size = SIZEOF_ARRAY(open_str) - 1; - - OFF off = beg; - - if(off + open_size >= lines[0].end) - return FALSE; - if(memcmp(STR(off), open_str, open_size) != 0) - return FALSE; - off += open_size; - - if(lines[n_lines-1].end < max_end) - max_end = lines[n_lines-1].end - 2; - - return md_scan_for_html_closer(ctx, _T("]]>"), 3, - lines, n_lines, off, max_end, p_end, &ctx->html_cdata_horizon); -} - -static int -md_is_html_any(MD_CTX* ctx, const MD_LINE* lines, int n_lines, OFF beg, OFF max_end, OFF* p_end) -{ - MD_ASSERT(CH(beg) == _T('<')); - return (md_is_html_tag(ctx, lines, n_lines, beg, max_end, p_end) || - md_is_html_comment(ctx, lines, n_lines, beg, max_end, p_end) || - md_is_html_processing_instruction(ctx, lines, n_lines, beg, max_end, p_end) || - md_is_html_declaration(ctx, lines, n_lines, beg, max_end, p_end) || - md_is_html_cdata(ctx, lines, n_lines, beg, max_end, p_end)); -} - - -/**************************** - *** Recognizing Entity *** - ****************************/ - -static int -md_is_hex_entity_contents(MD_CTX* ctx, const CHAR* text, OFF beg, OFF max_end, OFF* p_end) -{ - OFF off = beg; - MD_UNUSED(ctx); - - while(off < max_end && ISXDIGIT_(text[off]) && off - beg <= 8) - off++; - - if(1 <= off - beg && off - beg <= 6) { - *p_end = off; - return TRUE; - } else { - return FALSE; - } -} - -static int -md_is_dec_entity_contents(MD_CTX* ctx, const CHAR* text, OFF beg, OFF max_end, OFF* p_end) -{ - OFF off = beg; - MD_UNUSED(ctx); - - while(off < max_end && ISDIGIT_(text[off]) && off - beg <= 8) - off++; - - if(1 <= off - beg && off - beg <= 7) { - *p_end = off; - return TRUE; - } else { - return FALSE; - } -} - -static int -md_is_named_entity_contents(MD_CTX* ctx, const CHAR* text, OFF beg, OFF max_end, OFF* p_end) -{ - OFF off = beg; - MD_UNUSED(ctx); - - if(off < max_end && ISALPHA_(text[off])) - off++; - else - return FALSE; - - while(off < max_end && ISALNUM_(text[off]) && off - beg <= 48) - off++; - - if(2 <= off - beg && off - beg <= 48) { - *p_end = off; - return TRUE; - } else { - return FALSE; - } -} - -static int -md_is_entity_str(MD_CTX* ctx, const CHAR* text, OFF beg, OFF max_end, OFF* p_end) -{ - int is_contents; - OFF off = beg; - - MD_ASSERT(text[off] == _T('&')); - off++; - - if(off+2 < max_end && text[off] == _T('#') && (text[off+1] == _T('x') || text[off+1] == _T('X'))) - is_contents = md_is_hex_entity_contents(ctx, text, off+2, max_end, &off); - else if(off+1 < max_end && text[off] == _T('#')) - is_contents = md_is_dec_entity_contents(ctx, text, off+1, max_end, &off); - else - is_contents = md_is_named_entity_contents(ctx, text, off, max_end, &off); - - if(is_contents && off < max_end && text[off] == _T(';')) { - *p_end = off+1; - return TRUE; - } else { - return FALSE; - } -} - -static inline int -md_is_entity(MD_CTX* ctx, OFF beg, OFF max_end, OFF* p_end) -{ - return md_is_entity_str(ctx, ctx->text, beg, max_end, p_end); -} - - -/****************************** - *** Attribute Management *** - ******************************/ - -typedef struct MD_ATTRIBUTE_BUILD_tag MD_ATTRIBUTE_BUILD; -struct MD_ATTRIBUTE_BUILD_tag { - CHAR* text; - MD_TEXTTYPE* substr_types; - OFF* substr_offsets; - int substr_count; - int substr_alloc; - MD_TEXTTYPE trivial_types[1]; - OFF trivial_offsets[2]; -}; - - -#define MD_BUILD_ATTR_NO_ESCAPES 0x0001 - -static int -md_build_attr_append_substr(MD_CTX* ctx, MD_ATTRIBUTE_BUILD* build, - MD_TEXTTYPE type, OFF off) -{ - if(build->substr_count >= build->substr_alloc) { - MD_TEXTTYPE* new_substr_types; - OFF* new_substr_offsets; - - build->substr_alloc = (build->substr_alloc > 0 - ? build->substr_alloc + build->substr_alloc / 2 - : 8); - new_substr_types = (MD_TEXTTYPE*) realloc(build->substr_types, - build->substr_alloc * sizeof(MD_TEXTTYPE)); - if(new_substr_types == NULL) { - MD_LOG("realloc() failed."); - return -1; - } - /* Note +1 to reserve space for final offset (== raw_size). */ - new_substr_offsets = (OFF*) realloc(build->substr_offsets, - (build->substr_alloc+1) * sizeof(OFF)); - if(new_substr_offsets == NULL) { - MD_LOG("realloc() failed."); - free(new_substr_types); - return -1; - } - - build->substr_types = new_substr_types; - build->substr_offsets = new_substr_offsets; - } - - build->substr_types[build->substr_count] = type; - build->substr_offsets[build->substr_count] = off; - build->substr_count++; - return 0; -} - -static void -md_free_attribute(MD_CTX* ctx, MD_ATTRIBUTE_BUILD* build) -{ - MD_UNUSED(ctx); - - if(build->substr_alloc > 0) { - free(build->text); - free(build->substr_types); - free(build->substr_offsets); - } -} - -static int -md_build_attribute(MD_CTX* ctx, const CHAR* raw_text, SZ raw_size, - unsigned flags, MD_ATTRIBUTE* attr, MD_ATTRIBUTE_BUILD* build) -{ - OFF raw_off, off; - int is_trivial; - int ret = 0; - - memset(build, 0, sizeof(MD_ATTRIBUTE_BUILD)); - - /* If there is no backslash and no ampersand, build trivial attribute - * without any malloc(). */ - is_trivial = TRUE; - for(raw_off = 0; raw_off < raw_size; raw_off++) { - if(ISANYOF3_(raw_text[raw_off], _T('\\'), _T('&'), _T('\0'))) { - is_trivial = FALSE; - break; - } - } - - if(is_trivial) { - build->text = (CHAR*) (raw_size ? raw_text : NULL); - build->substr_types = build->trivial_types; - build->substr_offsets = build->trivial_offsets; - build->substr_count = 1; - build->substr_alloc = 0; - build->trivial_types[0] = MD_TEXT_NORMAL; - build->trivial_offsets[0] = 0; - build->trivial_offsets[1] = raw_size; - off = raw_size; - } else { - build->text = (CHAR*) malloc(raw_size * sizeof(CHAR)); - if(build->text == NULL) { - MD_LOG("malloc() failed."); - goto abort; - } - - raw_off = 0; - off = 0; - - while(raw_off < raw_size) { - if(raw_text[raw_off] == _T('\0')) { - MD_CHECK(md_build_attr_append_substr(ctx, build, MD_TEXT_NULLCHAR, off)); - memcpy(build->text + off, raw_text + raw_off, 1); - off++; - raw_off++; - continue; - } - - if(raw_text[raw_off] == _T('&')) { - OFF ent_end; - - if(md_is_entity_str(ctx, raw_text, raw_off, raw_size, &ent_end)) { - MD_CHECK(md_build_attr_append_substr(ctx, build, MD_TEXT_ENTITY, off)); - memcpy(build->text + off, raw_text + raw_off, ent_end - raw_off); - off += ent_end - raw_off; - raw_off = ent_end; - continue; - } - } - - if(build->substr_count == 0 || build->substr_types[build->substr_count-1] != MD_TEXT_NORMAL) - MD_CHECK(md_build_attr_append_substr(ctx, build, MD_TEXT_NORMAL, off)); - - if(!(flags & MD_BUILD_ATTR_NO_ESCAPES) && - raw_text[raw_off] == _T('\\') && raw_off+1 < raw_size && - (ISPUNCT_(raw_text[raw_off+1]) || ISNEWLINE_(raw_text[raw_off+1]))) - raw_off++; - - build->text[off++] = raw_text[raw_off++]; - } - build->substr_offsets[build->substr_count] = off; - } - - attr->text = build->text; - attr->size = off; - attr->substr_offsets = build->substr_offsets; - attr->substr_types = build->substr_types; - return 0; - -abort: - md_free_attribute(ctx, build); - return -1; -} - - -/********************************************* - *** Dictionary of Reference Definitions *** - *********************************************/ - -#define MD_FNV1A_BASE 2166136261U -#define MD_FNV1A_PRIME 16777619U - -static inline unsigned -md_fnv1a(unsigned base, const void* data, size_t n) -{ - const unsigned char* buf = (const unsigned char*) data; - unsigned hash = base; - size_t i; - - for(i = 0; i < n; i++) { - hash ^= buf[i]; - hash *= MD_FNV1A_PRIME; - } - - return hash; -} - - -struct MD_REF_DEF_tag { - CHAR* label; - CHAR* title; - unsigned hash; - SZ label_size; - SZ title_size; - OFF dest_beg; - OFF dest_end; - unsigned char label_needs_free : 1; - unsigned char title_needs_free : 1; -}; - -/* Label equivalence is quite complicated with regards to whitespace and case - * folding. This complicates computing a hash of it as well as direct comparison - * of two labels. */ - -static unsigned -md_link_label_hash(const CHAR* label, SZ size) -{ - unsigned hash = MD_FNV1A_BASE; - OFF off; - unsigned codepoint; - int is_whitespace = FALSE; - - off = md_skip_unicode_whitespace(label, 0, size); - while(off < size) { - SZ char_size; - - codepoint = md_decode_unicode(label, off, size, &char_size); - is_whitespace = ISUNICODEWHITESPACE_(codepoint) || ISNEWLINE_(label[off]); - - if(is_whitespace) { - codepoint = ' '; - hash = md_fnv1a(hash, &codepoint, sizeof(unsigned)); - off = md_skip_unicode_whitespace(label, off, size); - } else { - MD_UNICODE_FOLD_INFO fold_info; - - md_get_unicode_fold_info(codepoint, &fold_info); - hash = md_fnv1a(hash, fold_info.codepoints, fold_info.n_codepoints * sizeof(unsigned)); - off += char_size; - } - } - - return hash; -} - -static OFF -md_link_label_cmp_load_fold_info(const CHAR* label, OFF off, SZ size, - MD_UNICODE_FOLD_INFO* fold_info) -{ - unsigned codepoint; - SZ char_size; - - if(off >= size) { - /* Treat end of a link label as a whitespace. */ - goto whitespace; - } - - codepoint = md_decode_unicode(label, off, size, &char_size); - off += char_size; - if(ISUNICODEWHITESPACE_(codepoint)) { - /* Treat all whitespace as equivalent */ - goto whitespace; - } - - /* Get real folding info. */ - md_get_unicode_fold_info(codepoint, fold_info); - return off; - -whitespace: - fold_info->codepoints[0] = _T(' '); - fold_info->n_codepoints = 1; - return md_skip_unicode_whitespace(label, off, size); -} - -static int -md_link_label_cmp(const CHAR* a_label, SZ a_size, const CHAR* b_label, SZ b_size) -{ - OFF a_off; - OFF b_off; - MD_UNICODE_FOLD_INFO a_fi = { { 0 }, 0 }; - MD_UNICODE_FOLD_INFO b_fi = { { 0 }, 0 }; - OFF a_fi_off = 0; - OFF b_fi_off = 0; - int cmp; - - a_off = md_skip_unicode_whitespace(a_label, 0, a_size); - b_off = md_skip_unicode_whitespace(b_label, 0, b_size); - while(a_off < a_size || a_fi_off < a_fi.n_codepoints || - b_off < b_size || b_fi_off < b_fi.n_codepoints) - { - /* If needed, load fold info for next char. */ - if(a_fi_off >= a_fi.n_codepoints) { - a_fi_off = 0; - a_off = md_link_label_cmp_load_fold_info(a_label, a_off, a_size, &a_fi); - } - if(b_fi_off >= b_fi.n_codepoints) { - b_fi_off = 0; - b_off = md_link_label_cmp_load_fold_info(b_label, b_off, b_size, &b_fi); - } - - cmp = b_fi.codepoints[b_fi_off] - a_fi.codepoints[a_fi_off]; - if(cmp != 0) - return cmp; - - a_fi_off++; - b_fi_off++; - } - - return 0; -} - -typedef struct MD_REF_DEF_LIST_tag MD_REF_DEF_LIST; -struct MD_REF_DEF_LIST_tag { - int n_ref_defs; - int alloc_ref_defs; - MD_REF_DEF* ref_defs[]; /* Valid items always point into ctx->ref_defs[] */ -}; - -static int -md_ref_def_cmp(const void* a, const void* b) -{ - const MD_REF_DEF* a_ref = *(const MD_REF_DEF**)a; - const MD_REF_DEF* b_ref = *(const MD_REF_DEF**)b; - - if(a_ref->hash < b_ref->hash) - return -1; - else if(a_ref->hash > b_ref->hash) - return +1; - else - return md_link_label_cmp(a_ref->label, a_ref->label_size, b_ref->label, b_ref->label_size); -} - -static int -md_ref_def_cmp_for_sort(const void* a, const void* b) -{ - int cmp; - - cmp = md_ref_def_cmp(a, b); - - /* Ensure stability of the sorting. */ - if(cmp == 0) { - const MD_REF_DEF* a_ref = *(const MD_REF_DEF**)a; - const MD_REF_DEF* b_ref = *(const MD_REF_DEF**)b; - - if(a_ref < b_ref) - cmp = -1; - else if(a_ref > b_ref) - cmp = +1; - else - cmp = 0; - } - - return cmp; -} - -static int -md_build_ref_def_hashtable(MD_CTX* ctx) -{ - int i, j; - - if(ctx->n_ref_defs == 0) - return 0; - - ctx->ref_def_hashtable_size = (ctx->n_ref_defs * 5) / 4; - ctx->ref_def_hashtable = malloc(ctx->ref_def_hashtable_size * sizeof(void*)); - if(ctx->ref_def_hashtable == NULL) { - MD_LOG("malloc() failed."); - goto abort; - } - memset(ctx->ref_def_hashtable, 0, ctx->ref_def_hashtable_size * sizeof(void*)); - - /* Each member of ctx->ref_def_hashtable[] can be: - * -- NULL, - * -- pointer to the MD_REF_DEF in ctx->ref_defs[], or - * -- pointer to a MD_REF_DEF_LIST, which holds multiple pointers to - * such MD_REF_DEFs. - */ - for(i = 0; i < ctx->n_ref_defs; i++) { - MD_REF_DEF* def = &ctx->ref_defs[i]; - void* bucket; - MD_REF_DEF_LIST* list; - - def->hash = md_link_label_hash(def->label, def->label_size); - bucket = ctx->ref_def_hashtable[def->hash % ctx->ref_def_hashtable_size]; - - if(bucket == NULL) { - /* The bucket is empty. Make it just point to the def. */ - ctx->ref_def_hashtable[def->hash % ctx->ref_def_hashtable_size] = def; - continue; - } - - if(ctx->ref_defs <= (MD_REF_DEF*) bucket && (MD_REF_DEF*) bucket < ctx->ref_defs + ctx->n_ref_defs) { - /* The bucket already contains one ref. def. Lets see whether it - * is the same label (ref. def. duplicate) or different one - * (hash conflict). */ - MD_REF_DEF* old_def = (MD_REF_DEF*) bucket; - - if(md_link_label_cmp(def->label, def->label_size, old_def->label, old_def->label_size) == 0) { - /* Duplicate label: Ignore this ref. def. */ - continue; - } - - /* Make the bucket complex, i.e. able to hold more ref. defs. */ - list = (MD_REF_DEF_LIST*) malloc(sizeof(MD_REF_DEF_LIST) + 2 * sizeof(MD_REF_DEF*)); - if(list == NULL) { - MD_LOG("malloc() failed."); - goto abort; - } - list->ref_defs[0] = old_def; - list->ref_defs[1] = def; - list->n_ref_defs = 2; - list->alloc_ref_defs = 2; - ctx->ref_def_hashtable[def->hash % ctx->ref_def_hashtable_size] = list; - continue; - } - - /* Append the def to the complex bucket list. - * - * Note in this case we ignore potential duplicates to avoid expensive - * iterating over the complex bucket. Below, we revisit all the complex - * buckets and handle it more cheaply after the complex bucket contents - * is sorted. */ - list = (MD_REF_DEF_LIST*) bucket; - if(list->n_ref_defs >= list->alloc_ref_defs) { - int alloc_ref_defs = list->alloc_ref_defs + list->alloc_ref_defs / 2; - MD_REF_DEF_LIST* list_tmp = (MD_REF_DEF_LIST*) realloc(list, - sizeof(MD_REF_DEF_LIST) + alloc_ref_defs * sizeof(MD_REF_DEF*)); - if(list_tmp == NULL) { - MD_LOG("realloc() failed."); - goto abort; - } - list = list_tmp; - list->alloc_ref_defs = alloc_ref_defs; - ctx->ref_def_hashtable[def->hash % ctx->ref_def_hashtable_size] = list; - } - - list->ref_defs[list->n_ref_defs] = def; - list->n_ref_defs++; - } - - /* Sort the complex buckets so we can use bsearch() with them. */ - for(i = 0; i < ctx->ref_def_hashtable_size; i++) { - void* bucket = ctx->ref_def_hashtable[i]; - MD_REF_DEF_LIST* list; - - if(bucket == NULL) - continue; - if(ctx->ref_defs <= (MD_REF_DEF*) bucket && (MD_REF_DEF*) bucket < ctx->ref_defs + ctx->n_ref_defs) - continue; - - list = (MD_REF_DEF_LIST*) bucket; - qsort(list->ref_defs, list->n_ref_defs, sizeof(MD_REF_DEF*), md_ref_def_cmp_for_sort); - - /* Disable all duplicates in the complex bucket by forcing all such - * records to point to the 1st such ref. def. I.e. no matter which - * record is found during the lookup, it will always point to the right - * ref. def. in ctx->ref_defs[]. */ - for(j = 1; j < list->n_ref_defs; j++) { - if(md_ref_def_cmp(&list->ref_defs[j-1], &list->ref_defs[j]) == 0) - list->ref_defs[j] = list->ref_defs[j-1]; - } - } - - return 0; - -abort: - return -1; -} - -static void -md_free_ref_def_hashtable(MD_CTX* ctx) -{ - if(ctx->ref_def_hashtable != NULL) { - int i; - - for(i = 0; i < ctx->ref_def_hashtable_size; i++) { - void* bucket = ctx->ref_def_hashtable[i]; - if(bucket == NULL) - continue; - if(ctx->ref_defs <= (MD_REF_DEF*) bucket && (MD_REF_DEF*) bucket < ctx->ref_defs + ctx->n_ref_defs) - continue; - free(bucket); - } - - free(ctx->ref_def_hashtable); - } -} - -static const MD_REF_DEF* -md_lookup_ref_def(MD_CTX* ctx, const CHAR* label, SZ label_size) -{ - unsigned hash; - void* bucket; - - if(ctx->ref_def_hashtable_size == 0) - return NULL; - - hash = md_link_label_hash(label, label_size); - bucket = ctx->ref_def_hashtable[hash % ctx->ref_def_hashtable_size]; - - if(bucket == NULL) { - return NULL; - } else if(ctx->ref_defs <= (MD_REF_DEF*) bucket && (MD_REF_DEF*) bucket < ctx->ref_defs + ctx->n_ref_defs) { - const MD_REF_DEF* def = (MD_REF_DEF*) bucket; - - if(md_link_label_cmp(def->label, def->label_size, label, label_size) == 0) - return def; - else - return NULL; - } else { - MD_REF_DEF_LIST* list = (MD_REF_DEF_LIST*) bucket; - MD_REF_DEF key_buf; - const MD_REF_DEF* key = &key_buf; - const MD_REF_DEF** ret; - - key_buf.label = (CHAR*) label; - key_buf.label_size = label_size; - key_buf.hash = md_link_label_hash(key_buf.label, key_buf.label_size); - - ret = (const MD_REF_DEF**) bsearch(&key, list->ref_defs, - list->n_ref_defs, sizeof(MD_REF_DEF*), md_ref_def_cmp); - if(ret != NULL) - return *ret; - else - return NULL; - } -} - - -/*************************** - *** Recognizing Links *** - ***************************/ - -/* Note this code is partially shared between processing inlines and blocks - * as reference definitions and links share some helper parser functions. - */ - -typedef struct MD_LINK_ATTR_tag MD_LINK_ATTR; -struct MD_LINK_ATTR_tag { - OFF dest_beg; - OFF dest_end; - - CHAR* title; - SZ title_size; - int title_needs_free; -}; - - -static int -md_is_link_label(MD_CTX* ctx, const MD_LINE* lines, int n_lines, OFF beg, - OFF* p_end, int* p_beg_line_index, int* p_end_line_index, - OFF* p_contents_beg, OFF* p_contents_end) -{ - OFF off = beg; - OFF contents_beg = 0; - OFF contents_end = 0; - int line_index = 0; - int len = 0; - - if(CH(off) != _T('[')) - return FALSE; - off++; - - while(1) { - OFF line_end = lines[line_index].end; - - while(off < line_end) { - if(CH(off) == _T('\\') && off+1 < ctx->size && (ISPUNCT(off+1) || ISNEWLINE(off+1))) { - if(contents_end == 0) { - contents_beg = off; - *p_beg_line_index = line_index; - } - contents_end = off + 2; - off += 2; - } else if(CH(off) == _T('[')) { - return FALSE; - } else if(CH(off) == _T(']')) { - if(contents_beg < contents_end) { - /* Success. */ - *p_contents_beg = contents_beg; - *p_contents_end = contents_end; - *p_end = off+1; - *p_end_line_index = line_index; - return TRUE; - } else { - /* Link label must have some non-whitespace contents. */ - return FALSE; - } - } else { - unsigned codepoint; - SZ char_size; - - codepoint = md_decode_unicode(ctx->text, off, ctx->size, &char_size); - if(!ISUNICODEWHITESPACE_(codepoint)) { - if(contents_end == 0) { - contents_beg = off; - *p_beg_line_index = line_index; - } - contents_end = off + char_size; - } - - off += char_size; - } - - len++; - if(len > 999) - return FALSE; - } - - line_index++; - len++; - if(line_index < n_lines) - off = lines[line_index].beg; - else - break; - } - - return FALSE; -} - -static int -md_is_link_destination_A(MD_CTX* ctx, OFF beg, OFF max_end, OFF* p_end, - OFF* p_contents_beg, OFF* p_contents_end) -{ - OFF off = beg; - - if(off >= max_end || CH(off) != _T('<')) - return FALSE; - off++; - - while(off < max_end) { - if(CH(off) == _T('\\') && off+1 < max_end && ISPUNCT(off+1)) { - off += 2; - continue; - } - - if(ISNEWLINE(off) || CH(off) == _T('<')) - return FALSE; - - if(CH(off) == _T('>')) { - /* Success. */ - *p_contents_beg = beg+1; - *p_contents_end = off; - *p_end = off+1; - return TRUE; - } - - off++; - } - - return FALSE; -} - -static int -md_is_link_destination_B(MD_CTX* ctx, OFF beg, OFF max_end, OFF* p_end, - OFF* p_contents_beg, OFF* p_contents_end) -{ - OFF off = beg; - int parenthesis_level = 0; - - while(off < max_end) { - if(CH(off) == _T('\\') && off+1 < max_end && ISPUNCT(off+1)) { - off += 2; - continue; - } - - if(ISWHITESPACE(off) || ISCNTRL(off)) - break; - - /* Link destination may include balanced pairs of unescaped '(' ')'. - * Note we limit the maximal nesting level by 32 to protect us from - * https://github.com/jgm/cmark/issues/214 */ - if(CH(off) == _T('(')) { - parenthesis_level++; - if(parenthesis_level > 32) - return FALSE; - } else if(CH(off) == _T(')')) { - if(parenthesis_level == 0) - break; - parenthesis_level--; - } - - off++; - } - - if(parenthesis_level != 0 || off == beg) - return FALSE; - - /* Success. */ - *p_contents_beg = beg; - *p_contents_end = off; - *p_end = off; - return TRUE; -} - -static inline int -md_is_link_destination(MD_CTX* ctx, OFF beg, OFF max_end, OFF* p_end, - OFF* p_contents_beg, OFF* p_contents_end) -{ - if(CH(beg) == _T('<')) - return md_is_link_destination_A(ctx, beg, max_end, p_end, p_contents_beg, p_contents_end); - else - return md_is_link_destination_B(ctx, beg, max_end, p_end, p_contents_beg, p_contents_end); -} - -static int -md_is_link_title(MD_CTX* ctx, const MD_LINE* lines, int n_lines, OFF beg, - OFF* p_end, int* p_beg_line_index, int* p_end_line_index, - OFF* p_contents_beg, OFF* p_contents_end) -{ - OFF off = beg; - CHAR closer_char; - int line_index = 0; - - /* White space with up to one line break. */ - while(off < lines[line_index].end && ISWHITESPACE(off)) - off++; - if(off >= lines[line_index].end) { - line_index++; - if(line_index >= n_lines) - return FALSE; - off = lines[line_index].beg; - } - if(off == beg) - return FALSE; - - *p_beg_line_index = line_index; - - /* First char determines how to detect end of it. */ - switch(CH(off)) { - case _T('"'): closer_char = _T('"'); break; - case _T('\''): closer_char = _T('\''); break; - case _T('('): closer_char = _T(')'); break; - default: return FALSE; - } - off++; - - *p_contents_beg = off; - - while(line_index < n_lines) { - OFF line_end = lines[line_index].end; - - while(off < line_end) { - if(CH(off) == _T('\\') && off+1 < ctx->size && (ISPUNCT(off+1) || ISNEWLINE(off+1))) { - off++; - } else if(CH(off) == closer_char) { - /* Success. */ - *p_contents_end = off; - *p_end = off+1; - *p_end_line_index = line_index; - return TRUE; - } else if(closer_char == _T(')') && CH(off) == _T('(')) { - /* ()-style title cannot contain (unescaped '(')) */ - return FALSE; - } - - off++; - } - - line_index++; - } - - return FALSE; -} - -/* Returns 0 if it is not a reference definition. - * - * Returns N > 0 if it is a reference definition. N then corresponds to the - * number of lines forming it). In this case the definition is stored for - * resolving any links referring to it. - * - * Returns -1 in case of an error (out of memory). - */ -static int -md_is_link_reference_definition(MD_CTX* ctx, const MD_LINE* lines, int n_lines) -{ - OFF label_contents_beg; - OFF label_contents_end; - int label_contents_line_index = -1; - int label_is_multiline = FALSE; - OFF dest_contents_beg; - OFF dest_contents_end; - OFF title_contents_beg; - OFF title_contents_end; - int title_contents_line_index; - int title_is_multiline = FALSE; - OFF off; - int line_index = 0; - int tmp_line_index; - MD_REF_DEF* def = NULL; - int ret = 0; - - /* Link label. */ - if(!md_is_link_label(ctx, lines, n_lines, lines[0].beg, - &off, &label_contents_line_index, &line_index, - &label_contents_beg, &label_contents_end)) - return FALSE; - label_is_multiline = (label_contents_line_index != line_index); - - /* Colon. */ - if(off >= lines[line_index].end || CH(off) != _T(':')) - return FALSE; - off++; - - /* Optional white space with up to one line break. */ - while(off < lines[line_index].end && ISWHITESPACE(off)) - off++; - if(off >= lines[line_index].end) { - line_index++; - if(line_index >= n_lines) - return FALSE; - off = lines[line_index].beg; - } - - /* Link destination. */ - if(!md_is_link_destination(ctx, off, lines[line_index].end, - &off, &dest_contents_beg, &dest_contents_end)) - return FALSE; - - /* (Optional) title. Note we interpret it as an title only if nothing - * more follows on its last line. */ - if(md_is_link_title(ctx, lines + line_index, n_lines - line_index, off, - &off, &title_contents_line_index, &tmp_line_index, - &title_contents_beg, &title_contents_end) - && off >= lines[line_index + tmp_line_index].end) - { - title_is_multiline = (tmp_line_index != title_contents_line_index); - title_contents_line_index += line_index; - line_index += tmp_line_index; - } else { - /* Not a title. */ - title_is_multiline = FALSE; - title_contents_beg = off; - title_contents_end = off; - title_contents_line_index = 0; - } - - /* Nothing more can follow on the last line. */ - if(off < lines[line_index].end) - return FALSE; - - /* So, it _is_ a reference definition. Remember it. */ - if(ctx->n_ref_defs >= ctx->alloc_ref_defs) { - MD_REF_DEF* new_defs; - - ctx->alloc_ref_defs = (ctx->alloc_ref_defs > 0 - ? ctx->alloc_ref_defs + ctx->alloc_ref_defs / 2 - : 16); - new_defs = (MD_REF_DEF*) realloc(ctx->ref_defs, ctx->alloc_ref_defs * sizeof(MD_REF_DEF)); - if(new_defs == NULL) { - MD_LOG("realloc() failed."); - goto abort; - } - - ctx->ref_defs = new_defs; - } - def = &ctx->ref_defs[ctx->n_ref_defs]; - memset(def, 0, sizeof(MD_REF_DEF)); - - if(label_is_multiline) { - MD_CHECK(md_merge_lines_alloc(ctx, label_contents_beg, label_contents_end, - lines + label_contents_line_index, n_lines - label_contents_line_index, - _T(' '), &def->label, &def->label_size)); - def->label_needs_free = TRUE; - } else { - def->label = (CHAR*) STR(label_contents_beg); - def->label_size = label_contents_end - label_contents_beg; - } - - if(title_is_multiline) { - MD_CHECK(md_merge_lines_alloc(ctx, title_contents_beg, title_contents_end, - lines + title_contents_line_index, n_lines - title_contents_line_index, - _T('\n'), &def->title, &def->title_size)); - def->title_needs_free = TRUE; - } else { - def->title = (CHAR*) STR(title_contents_beg); - def->title_size = title_contents_end - title_contents_beg; - } - - def->dest_beg = dest_contents_beg; - def->dest_end = dest_contents_end; - - /* Success. */ - ctx->n_ref_defs++; - return line_index + 1; - -abort: - /* Failure. */ - if(def != NULL && def->label_needs_free) - free(def->label); - if(def != NULL && def->title_needs_free) - free(def->title); - return ret; -} - -static int -md_is_link_reference(MD_CTX* ctx, const MD_LINE* lines, int n_lines, - OFF beg, OFF end, MD_LINK_ATTR* attr) -{ - const MD_REF_DEF* def; - const MD_LINE* beg_line; - const MD_LINE* end_line; - CHAR* label; - SZ label_size; - int ret; - - MD_ASSERT(CH(beg) == _T('[') || CH(beg) == _T('!')); - MD_ASSERT(CH(end-1) == _T(']')); - - beg += (CH(beg) == _T('!') ? 2 : 1); - end--; - - /* Find lines corresponding to the beg and end positions. */ - MD_ASSERT(lines[0].beg <= beg); - beg_line = lines; - while(beg >= beg_line->end) - beg_line++; - - MD_ASSERT(end <= lines[n_lines-1].end); - end_line = beg_line; - while(end >= end_line->end) - end_line++; - - if(beg_line != end_line) { - MD_CHECK(md_merge_lines_alloc(ctx, beg, end, beg_line, - (int)(n_lines - (beg_line - lines)), _T(' '), &label, &label_size)); - } else { - label = (CHAR*) STR(beg); - label_size = end - beg; - } - - def = md_lookup_ref_def(ctx, label, label_size); - if(def != NULL) { - attr->dest_beg = def->dest_beg; - attr->dest_end = def->dest_end; - attr->title = def->title; - attr->title_size = def->title_size; - attr->title_needs_free = FALSE; - } - - if(beg_line != end_line) - free(label); - - ret = (def != NULL); - -abort: - return ret; -} - -static int -md_is_inline_link_spec(MD_CTX* ctx, const MD_LINE* lines, int n_lines, - OFF beg, OFF* p_end, MD_LINK_ATTR* attr) -{ - int line_index = 0; - int tmp_line_index; - OFF title_contents_beg; - OFF title_contents_end; - int title_contents_line_index; - int title_is_multiline; - OFF off = beg; - int ret = FALSE; - - while(off >= lines[line_index].end) - line_index++; - - MD_ASSERT(CH(off) == _T('(')); - off++; - - /* Optional white space with up to one line break. */ - while(off < lines[line_index].end && ISWHITESPACE(off)) - off++; - if(off >= lines[line_index].end && ISNEWLINE(off)) { - line_index++; - if(line_index >= n_lines) - return FALSE; - off = lines[line_index].beg; - } - - /* Link destination may be omitted, but only when not also having a title. */ - if(off < ctx->size && CH(off) == _T(')')) { - attr->dest_beg = off; - attr->dest_end = off; - attr->title = NULL; - attr->title_size = 0; - attr->title_needs_free = FALSE; - off++; - *p_end = off; - return TRUE; - } - - /* Link destination. */ - if(!md_is_link_destination(ctx, off, lines[line_index].end, - &off, &attr->dest_beg, &attr->dest_end)) - return FALSE; - - /* (Optional) title. */ - if(md_is_link_title(ctx, lines + line_index, n_lines - line_index, off, - &off, &title_contents_line_index, &tmp_line_index, - &title_contents_beg, &title_contents_end)) - { - title_is_multiline = (tmp_line_index != title_contents_line_index); - title_contents_line_index += line_index; - line_index += tmp_line_index; - } else { - /* Not a title. */ - title_is_multiline = FALSE; - title_contents_beg = off; - title_contents_end = off; - title_contents_line_index = 0; - } - - /* Optional whitespace followed with final ')'. */ - while(off < lines[line_index].end && ISWHITESPACE(off)) - off++; - if(off >= lines[line_index].end && ISNEWLINE(off)) { - line_index++; - if(line_index >= n_lines) - return FALSE; - off = lines[line_index].beg; - } - if(CH(off) != _T(')')) - goto abort; - off++; - - if(title_contents_beg >= title_contents_end) { - attr->title = NULL; - attr->title_size = 0; - attr->title_needs_free = FALSE; - } else if(!title_is_multiline) { - attr->title = (CHAR*) STR(title_contents_beg); - attr->title_size = title_contents_end - title_contents_beg; - attr->title_needs_free = FALSE; - } else { - MD_CHECK(md_merge_lines_alloc(ctx, title_contents_beg, title_contents_end, - lines + title_contents_line_index, n_lines - title_contents_line_index, - _T('\n'), &attr->title, &attr->title_size)); - attr->title_needs_free = TRUE; - } - - *p_end = off; - ret = TRUE; - -abort: - return ret; -} - -static void -md_free_ref_defs(MD_CTX* ctx) -{ - int i; - - for(i = 0; i < ctx->n_ref_defs; i++) { - MD_REF_DEF* def = &ctx->ref_defs[i]; - - if(def->label_needs_free) - free(def->label); - if(def->title_needs_free) - free(def->title); - } - - free(ctx->ref_defs); -} - - -/****************************************** - *** Processing Inlines (a.k.a Spans) *** - ******************************************/ - -/* We process inlines in few phases: - * - * (1) We go through the block text and collect all significant characters - * which may start/end a span or some other significant position into - * ctx->marks[]. Core of this is what md_collect_marks() does. - * - * We also do some very brief preliminary context-less analysis, whether - * it might be opener or closer (e.g. of an emphasis span). - * - * This speeds the other steps as we do not need to re-iterate over all - * characters anymore. - * - * (2) We analyze each potential mark types, in order by their precedence. - * - * In each md_analyze_XXX() function, we re-iterate list of the marks, - * skipping already resolved regions (in preceding precedences) and try to - * resolve them. - * - * (2.1) For trivial marks, which are single (e.g. HTML entity), we just mark - * them as resolved. - * - * (2.2) For range-type marks, we analyze whether the mark could be closer - * and, if yes, whether there is some preceding opener it could satisfy. - * - * If not we check whether it could be really an opener and if yes, we - * remember it so subsequent closers may resolve it. - * - * (3) Finally, when all marks were analyzed, we render the block contents - * by calling MD_RENDERER::text() callback, interrupting by ::enter_span() - * or ::close_span() whenever we reach a resolved mark. - */ - - -/* The mark structure. - * - * '\\': Maybe escape sequence. - * '\0': NULL char. - * '*': Maybe (strong) emphasis start/end. - * '_': Maybe (strong) emphasis start/end. - * '~': Maybe strikethrough start/end (needs MD_FLAG_STRIKETHROUGH). - * '`': Maybe code span start/end. - * '&': Maybe start of entity. - * ';': Maybe end of entity. - * '<': Maybe start of raw HTML or autolink. - * '>': Maybe end of raw HTML or autolink. - * '[': Maybe start of link label or link text. - * '!': Equivalent of '[' for image. - * ']': Maybe end of link label or link text. - * '@': Maybe permissive e-mail auto-link (needs MD_FLAG_PERMISSIVEEMAILAUTOLINKS). - * ':': Maybe permissive URL auto-link (needs MD_FLAG_PERMISSIVEURLAUTOLINKS). - * '.': Maybe permissive WWW auto-link (needs MD_FLAG_PERMISSIVEWWWAUTOLINKS). - * 'D': Dummy mark, it reserves a space for splitting a previous mark - * (e.g. emphasis) or to make more space for storing some special data - * related to the preceding mark (e.g. link). - * - * Note that not all instances of these chars in the text imply creation of the - * structure. Only those which have (or may have, after we see more context) - * the special meaning. - * - * (Keep this struct as small as possible to fit as much of them into CPU - * cache line.) - */ -struct MD_MARK_tag { - OFF beg; - OFF end; - - /* For unresolved openers, 'prev' and 'next' form the chain of open openers - * of given type 'ch'. - * - * During resolving, we disconnect from the chain and point to the - * corresponding counterpart so opener points to its closer and vice versa. - */ - int prev; - int next; - CHAR ch; - unsigned char flags; -}; - -/* Mark flags (these apply to ALL mark types). */ -#define MD_MARK_POTENTIAL_OPENER 0x01 /* Maybe opener. */ -#define MD_MARK_POTENTIAL_CLOSER 0x02 /* Maybe closer. */ -#define MD_MARK_OPENER 0x04 /* Definitely opener. */ -#define MD_MARK_CLOSER 0x08 /* Definitely closer. */ -#define MD_MARK_RESOLVED 0x10 /* Resolved in any definite way. */ - -/* Mark flags specific for various mark types (so they can share bits). */ -#define MD_MARK_EMPH_INTRAWORD 0x20 /* Helper for the "rule of 3". */ -#define MD_MARK_EMPH_MOD3_0 0x40 -#define MD_MARK_EMPH_MOD3_1 0x80 -#define MD_MARK_EMPH_MOD3_2 (0x40 | 0x80) -#define MD_MARK_EMPH_MOD3_MASK (0x40 | 0x80) -#define MD_MARK_AUTOLINK 0x20 /* Distinguisher for '<', '>'. */ -#define MD_MARK_VALIDPERMISSIVEAUTOLINK 0x20 /* For permissive autolinks. */ - -static MD_MARKCHAIN* -md_asterisk_chain(MD_CTX* ctx, unsigned flags) -{ - switch(flags & (MD_MARK_EMPH_INTRAWORD | MD_MARK_EMPH_MOD3_MASK)) { - case MD_MARK_EMPH_INTRAWORD | MD_MARK_EMPH_MOD3_0: return &ASTERISK_OPENERS_intraword_mod3_0; - case MD_MARK_EMPH_INTRAWORD | MD_MARK_EMPH_MOD3_1: return &ASTERISK_OPENERS_intraword_mod3_1; - case MD_MARK_EMPH_INTRAWORD | MD_MARK_EMPH_MOD3_2: return &ASTERISK_OPENERS_intraword_mod3_2; - case MD_MARK_EMPH_MOD3_0: return &ASTERISK_OPENERS_extraword_mod3_0; - case MD_MARK_EMPH_MOD3_1: return &ASTERISK_OPENERS_extraword_mod3_1; - case MD_MARK_EMPH_MOD3_2: return &ASTERISK_OPENERS_extraword_mod3_2; - default: MD_UNREACHABLE(); - } - return NULL; -} - -static MD_MARKCHAIN* -md_mark_chain(MD_CTX* ctx, int mark_index) -{ - MD_MARK* mark = &ctx->marks[mark_index]; - - switch(mark->ch) { - case _T('*'): return md_asterisk_chain(ctx, mark->flags); - case _T('_'): return &UNDERSCORE_OPENERS; - case _T('~'): return (mark->end - mark->beg == 1) ? &TILDE_OPENERS_1 : &TILDE_OPENERS_2; - case _T('['): return &BRACKET_OPENERS; - case _T('|'): return &TABLECELLBOUNDARIES; - default: return NULL; - } -} - -static MD_MARK* -md_push_mark(MD_CTX* ctx) -{ - if(ctx->n_marks >= ctx->alloc_marks) { - MD_MARK* new_marks; - - ctx->alloc_marks = (ctx->alloc_marks > 0 - ? ctx->alloc_marks + ctx->alloc_marks / 2 - : 64); - new_marks = realloc(ctx->marks, ctx->alloc_marks * sizeof(MD_MARK)); - if(new_marks == NULL) { - MD_LOG("realloc() failed."); - return NULL; - } - - ctx->marks = new_marks; - } - - return &ctx->marks[ctx->n_marks++]; -} - -#define PUSH_MARK_() \ - do { \ - mark = md_push_mark(ctx); \ - if(mark == NULL) { \ - ret = -1; \ - goto abort; \ - } \ - } while(0) - -#define PUSH_MARK(ch_, beg_, end_, flags_) \ - do { \ - PUSH_MARK_(); \ - mark->beg = (beg_); \ - mark->end = (end_); \ - mark->prev = -1; \ - mark->next = -1; \ - mark->ch = (char)(ch_); \ - mark->flags = (flags_); \ - } while(0) - - -static void -md_mark_chain_append(MD_CTX* ctx, MD_MARKCHAIN* chain, int mark_index) -{ - if(chain->tail >= 0) - ctx->marks[chain->tail].next = mark_index; - else - chain->head = mark_index; - - ctx->marks[mark_index].prev = chain->tail; - ctx->marks[mark_index].next = -1; - chain->tail = mark_index; -} - -/* Sometimes, we need to store a pointer into the mark. It is quite rare - * so we do not bother to make MD_MARK use union, and it can only happen - * for dummy marks. */ -static inline void -md_mark_store_ptr(MD_CTX* ctx, int mark_index, void* ptr) -{ - MD_MARK* mark = &ctx->marks[mark_index]; - MD_ASSERT(mark->ch == 'D'); - - /* Check only members beg and end are misused for this. */ - MD_ASSERT(sizeof(void*) <= 2 * sizeof(OFF)); - memcpy(mark, &ptr, sizeof(void*)); -} - -static inline void* -md_mark_get_ptr(MD_CTX* ctx, int mark_index) -{ - void* ptr; - MD_MARK* mark = &ctx->marks[mark_index]; - MD_ASSERT(mark->ch == 'D'); - memcpy(&ptr, mark, sizeof(void*)); - return ptr; -} - -static void -md_resolve_range(MD_CTX* ctx, MD_MARKCHAIN* chain, int opener_index, int closer_index) -{ - MD_MARK* opener = &ctx->marks[opener_index]; - MD_MARK* closer = &ctx->marks[closer_index]; - - /* Remove opener from the list of openers. */ - if(chain != NULL) { - if(opener->prev >= 0) - ctx->marks[opener->prev].next = opener->next; - else - chain->head = opener->next; - - if(opener->next >= 0) - ctx->marks[opener->next].prev = opener->prev; - else - chain->tail = opener->prev; - } - - /* Interconnect opener and closer and mark both as resolved. */ - opener->next = closer_index; - opener->flags |= MD_MARK_OPENER | MD_MARK_RESOLVED; - closer->prev = opener_index; - closer->flags |= MD_MARK_CLOSER | MD_MARK_RESOLVED; -} - - -#define MD_ROLLBACK_ALL 0 -#define MD_ROLLBACK_CROSSING 1 - -/* In the range ctx->marks[opener_index] ... [closer_index], undo some or all - * resolvings accordingly to these rules: - * - * (1) All openers BEFORE the range corresponding to any closer inside the - * range are un-resolved and they are re-added to their respective chains - * of unresolved openers. This ensures we can reuse the opener for closers - * AFTER the range. - * - * (2) If 'how' is MD_ROLLBACK_ALL, then ALL resolved marks inside the range - * are discarded. - * - * (3) If 'how' is MD_ROLLBACK_CROSSING, only closers with openers handled - * in (1) are discarded. I.e. pairs of openers and closers which are both - * inside the range are retained as well as any unpaired marks. - */ -static void -md_rollback(MD_CTX* ctx, int opener_index, int closer_index, int how) -{ - int i; - int mark_index; - - /* Cut all unresolved openers at the mark index. */ - for(i = OPENERS_CHAIN_FIRST; i < OPENERS_CHAIN_LAST+1; i++) { - MD_MARKCHAIN* chain = &ctx->mark_chains[i]; - - while(chain->tail >= opener_index) - chain->tail = ctx->marks[chain->tail].prev; - - if(chain->tail >= 0) - ctx->marks[chain->tail].next = -1; - else - chain->head = -1; - } - - /* Go backwards so that unresolved openers are re-added into their - * respective chains, in the right order. */ - mark_index = closer_index - 1; - while(mark_index > opener_index) { - MD_MARK* mark = &ctx->marks[mark_index]; - int mark_flags = mark->flags; - int discard_flag = (how == MD_ROLLBACK_ALL); - - if(mark->flags & MD_MARK_CLOSER) { - int mark_opener_index = mark->prev; - - /* Undo opener BEFORE the range. */ - if(mark_opener_index < opener_index) { - MD_MARK* mark_opener = &ctx->marks[mark_opener_index]; - MD_MARKCHAIN* chain; - - mark_opener->flags &= ~(MD_MARK_OPENER | MD_MARK_CLOSER | MD_MARK_RESOLVED); - chain = md_mark_chain(ctx, opener_index); - if(chain != NULL) { - md_mark_chain_append(ctx, chain, mark_opener_index); - discard_flag = 1; - } - } - } - - /* And reset our flags. */ - if(discard_flag) - mark->flags &= ~(MD_MARK_OPENER | MD_MARK_CLOSER | MD_MARK_RESOLVED); - - /* Jump as far as we can over unresolved or non-interesting marks. */ - switch(how) { - case MD_ROLLBACK_CROSSING: - if((mark_flags & MD_MARK_CLOSER) && mark->prev > opener_index) { - /* If we are closer with opener INSIDE the range, there may - * not be any other crosser inside the subrange. */ - mark_index = mark->prev; - break; - } - MD_FALLTHROUGH(); - default: - mark_index--; - break; - } - } -} - -static void -md_build_mark_char_map(MD_CTX* ctx) -{ - memset(ctx->mark_char_map, 0, sizeof(ctx->mark_char_map)); - - ctx->mark_char_map['\\'] = 1; - ctx->mark_char_map['*'] = 1; - ctx->mark_char_map['_'] = 1; - ctx->mark_char_map['`'] = 1; - ctx->mark_char_map['&'] = 1; - ctx->mark_char_map[';'] = 1; - ctx->mark_char_map['<'] = 1; - ctx->mark_char_map['>'] = 1; - ctx->mark_char_map['['] = 1; - ctx->mark_char_map['!'] = 1; - ctx->mark_char_map[']'] = 1; - ctx->mark_char_map['\0'] = 1; - - if(ctx->parser.flags & MD_FLAG_STRIKETHROUGH) - ctx->mark_char_map['~'] = 1; - - if(ctx->parser.flags & MD_FLAG_LATEXMATHSPANS) - ctx->mark_char_map['$'] = 1; - - if(ctx->parser.flags & MD_FLAG_PERMISSIVEEMAILAUTOLINKS) - ctx->mark_char_map['@'] = 1; - - if(ctx->parser.flags & MD_FLAG_PERMISSIVEURLAUTOLINKS) - ctx->mark_char_map[':'] = 1; - - if(ctx->parser.flags & MD_FLAG_PERMISSIVEWWWAUTOLINKS) - ctx->mark_char_map['.'] = 1; - - if((ctx->parser.flags & MD_FLAG_TABLES) || (ctx->parser.flags & MD_FLAG_WIKILINKS)) - ctx->mark_char_map['|'] = 1; - - if(ctx->parser.flags & MD_FLAG_COLLAPSEWHITESPACE) { - int i; - - for(i = 0; i < (int) sizeof(ctx->mark_char_map); i++) { - if(ISWHITESPACE_(i)) - ctx->mark_char_map[i] = 1; - } - } -} - -/* We limit code span marks to lower than 32 backticks. This solves the - * pathologic case of too many openers, each of different length: Their - * resolving would be then O(n^2). */ -#define CODESPAN_MARK_MAXLEN 32 - -static int -md_is_code_span(MD_CTX* ctx, const MD_LINE* lines, int n_lines, OFF beg, - OFF* p_opener_beg, OFF* p_opener_end, - OFF* p_closer_beg, OFF* p_closer_end, - OFF last_potential_closers[CODESPAN_MARK_MAXLEN], - int* p_reached_paragraph_end) -{ - OFF opener_beg = beg; - OFF opener_end; - OFF closer_beg; - OFF closer_end; - SZ mark_len; - OFF line_end; - int has_space_after_opener = FALSE; - int has_eol_after_opener = FALSE; - int has_space_before_closer = FALSE; - int has_eol_before_closer = FALSE; - int has_only_space = TRUE; - int line_index = 0; - - line_end = lines[0].end; - opener_end = opener_beg; - while(opener_end < line_end && CH(opener_end) == _T('`')) - opener_end++; - has_space_after_opener = (opener_end < line_end && CH(opener_end) == _T(' ')); - has_eol_after_opener = (opener_end == line_end); - - /* The caller needs to know end of the opening mark even if we fail. */ - *p_opener_end = opener_end; - - mark_len = opener_end - opener_beg; - if(mark_len > CODESPAN_MARK_MAXLEN) - return FALSE; - - /* Check whether we already know there is no closer of this length. - * If so, re-scan does no sense. This fixes issue #59. */ - if(last_potential_closers[mark_len-1] >= lines[n_lines-1].end || - (*p_reached_paragraph_end && last_potential_closers[mark_len-1] < opener_end)) - return FALSE; - - closer_beg = opener_end; - closer_end = opener_end; - - /* Find closer mark. */ - while(TRUE) { - while(closer_beg < line_end && CH(closer_beg) != _T('`')) { - if(CH(closer_beg) != _T(' ')) - has_only_space = FALSE; - closer_beg++; - } - closer_end = closer_beg; - while(closer_end < line_end && CH(closer_end) == _T('`')) - closer_end++; - - if(closer_end - closer_beg == mark_len) { - /* Success. */ - has_space_before_closer = (closer_beg > lines[line_index].beg && CH(closer_beg-1) == _T(' ')); - has_eol_before_closer = (closer_beg == lines[line_index].beg); - break; - } - - if(closer_end - closer_beg > 0) { - /* We have found a back-tick which is not part of the closer. */ - has_only_space = FALSE; - - /* But if we eventually fail, remember it as a potential closer - * of its own length for future attempts. This mitigates needs for - * rescans. */ - if(closer_end - closer_beg < CODESPAN_MARK_MAXLEN) { - if(closer_beg > last_potential_closers[closer_end - closer_beg - 1]) - last_potential_closers[closer_end - closer_beg - 1] = closer_beg; - } - } - - if(closer_end >= line_end) { - line_index++; - if(line_index >= n_lines) { - /* Reached end of the paragraph and still nothing. */ - *p_reached_paragraph_end = TRUE; - return FALSE; - } - /* Try on the next line. */ - line_end = lines[line_index].end; - closer_beg = lines[line_index].beg; - } else { - closer_beg = closer_end; - } - } - - /* If there is a space or a new line both after and before the opener - * (and if the code span is not made of spaces only), consume one initial - * and one trailing space as part of the marks. */ - if(!has_only_space && - (has_space_after_opener || has_eol_after_opener) && - (has_space_before_closer || has_eol_before_closer)) - { - if(has_space_after_opener) - opener_end++; - else - opener_end = lines[1].beg; - - if(has_space_before_closer) - closer_beg--; - else { - closer_beg = lines[line_index-1].end; - /* We need to eat the preceding "\r\n" but not any line trailing - * spaces. */ - while(closer_beg < ctx->size && ISBLANK(closer_beg)) - closer_beg++; - } - } - - *p_opener_beg = opener_beg; - *p_opener_end = opener_end; - *p_closer_beg = closer_beg; - *p_closer_end = closer_end; - return TRUE; -} - -static int -md_is_autolink_uri(MD_CTX* ctx, OFF beg, OFF max_end, OFF* p_end) -{ - OFF off = beg+1; - - MD_ASSERT(CH(beg) == _T('<')); - - /* Check for scheme. */ - if(off >= max_end || !ISASCII(off)) - return FALSE; - off++; - while(1) { - if(off >= max_end) - return FALSE; - if(off - beg > 32) - return FALSE; - if(CH(off) == _T(':') && off - beg >= 3) - break; - if(!ISALNUM(off) && CH(off) != _T('+') && CH(off) != _T('-') && CH(off) != _T('.')) - return FALSE; - off++; - } - - /* Check the path after the scheme. */ - while(off < max_end && CH(off) != _T('>')) { - if(ISWHITESPACE(off) || ISCNTRL(off) || CH(off) == _T('<')) - return FALSE; - off++; - } - - if(off >= max_end) - return FALSE; - - MD_ASSERT(CH(off) == _T('>')); - *p_end = off+1; - return TRUE; -} - -static int -md_is_autolink_email(MD_CTX* ctx, OFF beg, OFF max_end, OFF* p_end) -{ - OFF off = beg + 1; - int label_len; - - MD_ASSERT(CH(beg) == _T('<')); - - /* The code should correspond to this regexp: - /^[a-zA-Z0-9.!#$%&'*+\/=?^_`{|}~-]+ - @[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])? - (?:\.[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?)*$/ - */ - - /* Username (before '@'). */ - while(off < max_end && (ISALNUM(off) || ISANYOF(off, _T(".!#$%&'*+/=?^_`{|}~-")))) - off++; - if(off <= beg+1) - return FALSE; - - /* '@' */ - if(off >= max_end || CH(off) != _T('@')) - return FALSE; - off++; - - /* Labels delimited with '.'; each label is sequence of 1 - 63 alnum - * characters or '-', but '-' is not allowed as first or last char. */ - label_len = 0; - while(off < max_end) { - if(ISALNUM(off)) - label_len++; - else if(CH(off) == _T('-') && label_len > 0) - label_len++; - else if(CH(off) == _T('.') && label_len > 0 && CH(off-1) != _T('-')) - label_len = 0; - else - break; - - if(label_len > 63) - return FALSE; - - off++; - } - - if(label_len <= 0 || off >= max_end || CH(off) != _T('>') || CH(off-1) == _T('-')) - return FALSE; - - *p_end = off+1; - return TRUE; -} - -static int -md_is_autolink(MD_CTX* ctx, OFF beg, OFF max_end, OFF* p_end, int* p_missing_mailto) -{ - if(md_is_autolink_uri(ctx, beg, max_end, p_end)) { - *p_missing_mailto = FALSE; - return TRUE; - } - - if(md_is_autolink_email(ctx, beg, max_end, p_end)) { - *p_missing_mailto = TRUE; - return TRUE; - } - - return FALSE; -} - -static int -md_collect_marks(MD_CTX* ctx, const MD_LINE* lines, int n_lines, int table_mode) -{ - int i; - int ret = 0; - MD_MARK* mark; - OFF codespan_last_potential_closers[CODESPAN_MARK_MAXLEN] = { 0 }; - int codespan_scanned_till_paragraph_end = FALSE; - - for(i = 0; i < n_lines; i++) { - const MD_LINE* line = &lines[i]; - OFF off = line->beg; - OFF line_end = line->end; - - while(TRUE) { - CHAR ch; - -#ifdef MD4C_USE_UTF16 - /* For UTF-16, mark_char_map[] covers only ASCII. */ - #define IS_MARK_CHAR(off) ((CH(off) < SIZEOF_ARRAY(ctx->mark_char_map)) && \ - (ctx->mark_char_map[(unsigned char) CH(off)])) -#else - /* For 8-bit encodings, mark_char_map[] covers all 256 elements. */ - #define IS_MARK_CHAR(off) (ctx->mark_char_map[(unsigned char) CH(off)]) -#endif - - /* Optimization: Use some loop unrolling. */ - while(off + 3 < line_end && !IS_MARK_CHAR(off+0) && !IS_MARK_CHAR(off+1) - && !IS_MARK_CHAR(off+2) && !IS_MARK_CHAR(off+3)) - off += 4; - while(off < line_end && !IS_MARK_CHAR(off+0)) - off++; - - if(off >= line_end) - break; - - ch = CH(off); - - /* A backslash escape. - * It can go beyond line->end as it may involve escaped new - * line to form a hard break. */ - if(ch == _T('\\') && off+1 < ctx->size && (ISPUNCT(off+1) || ISNEWLINE(off+1))) { - /* Hard-break cannot be on the last line of the block. */ - if(!ISNEWLINE(off+1) || i+1 < n_lines) - PUSH_MARK(ch, off, off+2, MD_MARK_RESOLVED); - off += 2; - continue; - } - - /* A potential (string) emphasis start/end. */ - if(ch == _T('*') || ch == _T('_')) { - OFF tmp = off+1; - int left_level; /* What precedes: 0 = whitespace; 1 = punctuation; 2 = other char. */ - int right_level; /* What follows: 0 = whitespace; 1 = punctuation; 2 = other char. */ - - while(tmp < line_end && CH(tmp) == ch) - tmp++; - - if(off == line->beg || ISUNICODEWHITESPACEBEFORE(off)) - left_level = 0; - else if(ISUNICODEPUNCTBEFORE(off)) - left_level = 1; - else - left_level = 2; - - if(tmp == line_end || ISUNICODEWHITESPACE(tmp)) - right_level = 0; - else if(ISUNICODEPUNCT(tmp)) - right_level = 1; - else - right_level = 2; - - /* Intra-word underscore doesn't have special meaning. */ - if(ch == _T('_') && left_level == 2 && right_level == 2) { - left_level = 0; - right_level = 0; - } - - if(left_level != 0 || right_level != 0) { - unsigned flags = 0; - - if(left_level > 0 && left_level >= right_level) - flags |= MD_MARK_POTENTIAL_CLOSER; - if(right_level > 0 && right_level >= left_level) - flags |= MD_MARK_POTENTIAL_OPENER; - if(left_level == 2 && right_level == 2) - flags |= MD_MARK_EMPH_INTRAWORD; - - /* For "the rule of three" we need to remember the original - * size of the mark (modulo three), before we potentially - * split the mark when being later resolved partially by some - * shorter closer. */ - switch((tmp - off) % 3) { - case 0: flags |= MD_MARK_EMPH_MOD3_0; break; - case 1: flags |= MD_MARK_EMPH_MOD3_1; break; - case 2: flags |= MD_MARK_EMPH_MOD3_2; break; - } - - PUSH_MARK(ch, off, tmp, flags); - - /* During resolving, multiple asterisks may have to be - * split into independent span start/ends. Consider e.g. - * "**foo* bar*". Therefore we push also some empty dummy - * marks to have enough space for that. */ - off++; - while(off < tmp) { - PUSH_MARK('D', off, off, 0); - off++; - } - continue; - } - - off = tmp; - continue; - } - - /* A potential code span start/end. */ - if(ch == _T('`')) { - OFF opener_beg, opener_end; - OFF closer_beg, closer_end; - int is_code_span; - - is_code_span = md_is_code_span(ctx, lines + i, n_lines - i, off, - &opener_beg, &opener_end, &closer_beg, &closer_end, - codespan_last_potential_closers, - &codespan_scanned_till_paragraph_end); - if(is_code_span) { - PUSH_MARK(_T('`'), opener_beg, opener_end, MD_MARK_OPENER | MD_MARK_RESOLVED); - PUSH_MARK(_T('`'), closer_beg, closer_end, MD_MARK_CLOSER | MD_MARK_RESOLVED); - ctx->marks[ctx->n_marks-2].next = ctx->n_marks-1; - ctx->marks[ctx->n_marks-1].prev = ctx->n_marks-2; - - off = closer_end; - - /* Advance the current line accordingly. */ - while(off > line_end) { - i++; - line++; - line_end = line->end; - } - continue; - } - - off = opener_end; - continue; - } - - /* A potential entity start. */ - if(ch == _T('&')) { - PUSH_MARK(ch, off, off+1, MD_MARK_POTENTIAL_OPENER); - off++; - continue; - } - - /* A potential entity end. */ - if(ch == _T(';')) { - /* We surely cannot be entity unless the previous mark is '&'. */ - if(ctx->n_marks > 0 && ctx->marks[ctx->n_marks-1].ch == _T('&')) - PUSH_MARK(ch, off, off+1, MD_MARK_POTENTIAL_CLOSER); - - off++; - continue; - } - - /* A potential autolink or raw HTML start/end. */ - if(ch == _T('<')) { - int is_autolink; - OFF autolink_end; - int missing_mailto; - - if(!(ctx->parser.flags & MD_FLAG_NOHTMLSPANS)) { - int is_html; - OFF html_end; - - /* Given the nature of the raw HTML, we have to recognize - * it here. Doing so later in md_analyze_lt_gt() could - * open can of worms of quadratic complexity. */ - is_html = md_is_html_any(ctx, lines + i, n_lines - i, off, - lines[n_lines-1].end, &html_end); - if(is_html) { - PUSH_MARK(_T('<'), off, off, MD_MARK_OPENER | MD_MARK_RESOLVED); - PUSH_MARK(_T('>'), html_end, html_end, MD_MARK_CLOSER | MD_MARK_RESOLVED); - ctx->marks[ctx->n_marks-2].next = ctx->n_marks-1; - ctx->marks[ctx->n_marks-1].prev = ctx->n_marks-2; - off = html_end; - - /* Advance the current line accordingly. */ - while(off > line_end) { - i++; - line++; - line_end = line->end; - } - continue; - } - } - - is_autolink = md_is_autolink(ctx, off, lines[n_lines-1].end, - &autolink_end, &missing_mailto); - if(is_autolink) { - PUSH_MARK((missing_mailto ? _T('@') : _T('<')), off, off+1, - MD_MARK_OPENER | MD_MARK_RESOLVED | MD_MARK_AUTOLINK); - PUSH_MARK(_T('>'), autolink_end-1, autolink_end, - MD_MARK_CLOSER | MD_MARK_RESOLVED | MD_MARK_AUTOLINK); - ctx->marks[ctx->n_marks-2].next = ctx->n_marks-1; - ctx->marks[ctx->n_marks-1].prev = ctx->n_marks-2; - off = autolink_end; - continue; - } - - off++; - continue; - } - - /* A potential link or its part. */ - if(ch == _T('[') || (ch == _T('!') && off+1 < line_end && CH(off+1) == _T('['))) { - OFF tmp = (ch == _T('[') ? off+1 : off+2); - PUSH_MARK(ch, off, tmp, MD_MARK_POTENTIAL_OPENER); - off = tmp; - /* Two dummies to make enough place for data we need if it is - * a link. */ - PUSH_MARK('D', off, off, 0); - PUSH_MARK('D', off, off, 0); - continue; - } - if(ch == _T(']')) { - PUSH_MARK(ch, off, off+1, MD_MARK_POTENTIAL_CLOSER); - off++; - continue; - } - - /* A potential permissive e-mail autolink. */ - if(ch == _T('@')) { - if(line->beg + 1 <= off && ISALNUM(off-1) && - off + 3 < line->end && ISALNUM(off+1)) - { - PUSH_MARK(ch, off, off+1, MD_MARK_POTENTIAL_OPENER); - /* Push a dummy as a reserve for a closer. */ - PUSH_MARK('D', off, off, 0); - } - - off++; - continue; - } - - /* A potential permissive URL autolink. */ - if(ch == _T(':')) { - static struct { - const CHAR* scheme; - SZ scheme_size; - const CHAR* suffix; - SZ suffix_size; - } scheme_map[] = { - /* In the order from the most frequently used, arguably. */ - { _T("http"), 4, _T("//"), 2 }, - { _T("https"), 5, _T("//"), 2 }, - { _T("ftp"), 3, _T("//"), 2 } - }; - int scheme_index; - - for(scheme_index = 0; scheme_index < (int) SIZEOF_ARRAY(scheme_map); scheme_index++) { - const CHAR* scheme = scheme_map[scheme_index].scheme; - const SZ scheme_size = scheme_map[scheme_index].scheme_size; - const CHAR* suffix = scheme_map[scheme_index].suffix; - const SZ suffix_size = scheme_map[scheme_index].suffix_size; - - if(line->beg + scheme_size <= off && md_ascii_eq(STR(off-scheme_size), scheme, scheme_size) && - (line->beg + scheme_size == off || ISWHITESPACE(off-scheme_size-1) || ISANYOF(off-scheme_size-1, _T("*_~(["))) && - off + 1 + suffix_size < line->end && md_ascii_eq(STR(off+1), suffix, suffix_size)) - { - PUSH_MARK(ch, off-scheme_size, off+1+suffix_size, MD_MARK_POTENTIAL_OPENER); - /* Push a dummy as a reserve for a closer. */ - PUSH_MARK('D', off, off, 0); - off += 1 + suffix_size; - break; - } - } - - off++; - continue; - } - - /* A potential permissive WWW autolink. */ - if(ch == _T('.')) { - if(line->beg + 3 <= off && md_ascii_eq(STR(off-3), _T("www"), 3) && - (line->beg + 3 == off || ISWHITESPACE(off-4) || ISANYOF(off-4, _T("*_~(["))) && - off + 1 < line_end) - { - PUSH_MARK(ch, off-3, off+1, MD_MARK_POTENTIAL_OPENER); - /* Push a dummy as a reserve for a closer. */ - PUSH_MARK('D', off, off, 0); - off++; - continue; - } - - off++; - continue; - } - - /* A potential table cell boundary or wiki link label delimiter. */ - if((table_mode || ctx->parser.flags & MD_FLAG_WIKILINKS) && ch == _T('|')) { - PUSH_MARK(ch, off, off+1, 0); - off++; - continue; - } - - /* A potential strikethrough start/end. */ - if(ch == _T('~')) { - OFF tmp = off+1; - - while(tmp < line_end && CH(tmp) == _T('~')) - tmp++; - - if(tmp - off < 3) { - unsigned flags = 0; - - if(tmp < line_end && !ISUNICODEWHITESPACE(tmp)) - flags |= MD_MARK_POTENTIAL_OPENER; - if(off > line->beg && !ISUNICODEWHITESPACEBEFORE(off)) - flags |= MD_MARK_POTENTIAL_CLOSER; - if(flags != 0) - PUSH_MARK(ch, off, tmp, flags); - } - - off = tmp; - continue; - } - - /* A potential equation start/end */ - if(ch == _T('$')) { - /* We can have at most two consecutive $ signs, - * where two dollar signs signify a display equation. */ - OFF tmp = off+1; - - while(tmp < line_end && CH(tmp) == _T('$')) - tmp++; - - if (tmp - off <= 2) - PUSH_MARK(ch, off, tmp, MD_MARK_POTENTIAL_OPENER | MD_MARK_POTENTIAL_CLOSER); - off = tmp; - continue; - } - - /* Turn non-trivial whitespace into single space. */ - if(ISWHITESPACE_(ch)) { - OFF tmp = off+1; - - while(tmp < line_end && ISWHITESPACE(tmp)) - tmp++; - - if(tmp - off > 1 || ch != _T(' ')) - PUSH_MARK(ch, off, tmp, MD_MARK_RESOLVED); - - off = tmp; - continue; - } - - /* NULL character. */ - if(ch == _T('\0')) { - PUSH_MARK(ch, off, off+1, MD_MARK_RESOLVED); - off++; - continue; - } - - off++; - } - } - - /* Add a dummy mark at the end of the mark vector to simplify - * process_inlines(). */ - PUSH_MARK(127, ctx->size, ctx->size, MD_MARK_RESOLVED); - -abort: - return ret; -} - -static void -md_analyze_bracket(MD_CTX* ctx, int mark_index) -{ - /* We cannot really resolve links here as for that we would need - * more context. E.g. a following pair of brackets (reference link), - * or enclosing pair of brackets (if the inner is the link, the outer - * one cannot be.) - * - * Therefore we here only construct a list of resolved '[' ']' pairs - * ordered by position of the closer. This allows ur to analyze what is - * or is not link in the right order, from inside to outside in case - * of nested brackets. - * - * The resolving itself is deferred into md_resolve_links(). - */ - - MD_MARK* mark = &ctx->marks[mark_index]; - - if(mark->flags & MD_MARK_POTENTIAL_OPENER) { - md_mark_chain_append(ctx, &BRACKET_OPENERS, mark_index); - return; - } - - if(BRACKET_OPENERS.tail >= 0) { - /* Pop the opener from the chain. */ - int opener_index = BRACKET_OPENERS.tail; - MD_MARK* opener = &ctx->marks[opener_index]; - if(opener->prev >= 0) - ctx->marks[opener->prev].next = -1; - else - BRACKET_OPENERS.head = -1; - BRACKET_OPENERS.tail = opener->prev; - - /* Interconnect the opener and closer. */ - opener->next = mark_index; - mark->prev = opener_index; - - /* Add the pair into chain of potential links for md_resolve_links(). - * Note we misuse opener->prev for this as opener->next points to its - * closer. */ - if(ctx->unresolved_link_tail >= 0) - ctx->marks[ctx->unresolved_link_tail].prev = opener_index; - else - ctx->unresolved_link_head = opener_index; - ctx->unresolved_link_tail = opener_index; - opener->prev = -1; - } -} - -/* Forward declaration. */ -static void md_analyze_link_contents(MD_CTX* ctx, const MD_LINE* lines, int n_lines, - int mark_beg, int mark_end); - -static int -md_resolve_links(MD_CTX* ctx, const MD_LINE* lines, int n_lines) -{ - int opener_index = ctx->unresolved_link_head; - OFF last_link_beg = 0; - OFF last_link_end = 0; - OFF last_img_beg = 0; - OFF last_img_end = 0; - - while(opener_index >= 0) { - MD_MARK* opener = &ctx->marks[opener_index]; - int closer_index = opener->next; - MD_MARK* closer = &ctx->marks[closer_index]; - int next_index = opener->prev; - MD_MARK* next_opener; - MD_MARK* next_closer; - MD_LINK_ATTR attr; - int is_link = FALSE; - - if(next_index >= 0) { - next_opener = &ctx->marks[next_index]; - next_closer = &ctx->marks[next_opener->next]; - } else { - next_opener = NULL; - next_closer = NULL; - } - - /* If nested ("[ [ ] ]"), we need to make sure that: - * - The outer does not end inside of (...) belonging to the inner. - * - The outer cannot be link if the inner is link (i.e. not image). - * - * (Note we here analyze from inner to outer as the marks are ordered - * by closer->beg.) - */ - if((opener->beg < last_link_beg && closer->end < last_link_end) || - (opener->beg < last_img_beg && closer->end < last_img_end) || - (opener->beg < last_link_end && opener->ch == '[')) - { - opener_index = next_index; - continue; - } - - /* Recognize and resolve wiki links. - * Wiki-links maybe '[[destination]]' or '[[destination|label]]'. - */ - if ((ctx->parser.flags & MD_FLAG_WIKILINKS) && - (opener->end - opener->beg == 1) && /* not image */ - next_opener != NULL && /* double '[' opener */ - next_opener->ch == '[' && - (next_opener->beg == opener->beg - 1) && - (next_opener->end - next_opener->beg == 1) && - next_closer != NULL && /* double ']' closer */ - next_closer->ch == ']' && - (next_closer->beg == closer->beg + 1) && - (next_closer->end - next_closer->beg == 1)) - { - MD_MARK* delim = NULL; - int delim_index; - OFF dest_beg, dest_end; - - is_link = TRUE; - - /* We don't allow destination to be longer than 100 characters. - * Lets scan to see whether there is '|'. (If not then the whole - * wiki-link has to be below the 100 characters.) */ - delim_index = opener_index + 1; - while(delim_index < closer_index) { - MD_MARK* m = &ctx->marks[delim_index]; - if(m->ch == '|') { - delim = m; - break; - } - if(m->ch != 'D' && m->beg - opener->end > 100) - break; - delim_index++; - } - dest_beg = opener->end; - dest_end = (delim != NULL) ? delim->beg : closer->beg; - if(dest_end - dest_beg == 0 || dest_end - dest_beg > 100) - is_link = FALSE; - - /* There may not be any new line in the destination. */ - if(is_link) { - OFF off; - for(off = dest_beg; off < dest_end; off++) { - if(ISNEWLINE(off)) { - is_link = FALSE; - break; - } - } - } - - if(is_link) { - if(delim != NULL) { - if(delim->end < closer->beg) { - opener->end = delim->beg; - } else { - /* The pipe is just before the closer: [[foo|]] */ - closer->beg = delim->beg; - delim = NULL; - } - } - - opener->beg = next_opener->beg; - opener->next = closer_index; - opener->flags |= MD_MARK_OPENER | MD_MARK_RESOLVED; - - closer->end = next_closer->end; - closer->prev = opener_index; - closer->flags |= MD_MARK_CLOSER | MD_MARK_RESOLVED; - - last_link_beg = opener->beg; - last_link_end = closer->end; - - if(delim != NULL) { - delim->flags |= MD_MARK_RESOLVED; - md_rollback(ctx, opener_index, delim_index, MD_ROLLBACK_ALL); - md_analyze_link_contents(ctx, lines, n_lines, opener_index+1, closer_index); - } else { - md_rollback(ctx, opener_index, closer_index, MD_ROLLBACK_ALL); - } - - opener_index = next_opener->prev; - continue; - } - } - - if(next_opener != NULL && next_opener->beg == closer->end) { - if(next_closer->beg > closer->end + 1) { - /* Might be full reference link. */ - is_link = md_is_link_reference(ctx, lines, n_lines, next_opener->beg, next_closer->end, &attr); - } else { - /* Might be shortcut reference link. */ - is_link = md_is_link_reference(ctx, lines, n_lines, opener->beg, closer->end, &attr); - } - - if(is_link < 0) - return -1; - - if(is_link) { - /* Eat the 2nd "[...]". */ - closer->end = next_closer->end; - - /* Do not analyze the label as a standalone link in the next - * iteration. */ - next_index = ctx->marks[next_index].prev; - } - } else { - if(closer->end < ctx->size && CH(closer->end) == _T('(')) { - /* Might be inline link. */ - OFF inline_link_end = UINT_MAX; - - is_link = md_is_inline_link_spec(ctx, lines, n_lines, closer->end, &inline_link_end, &attr); - if(is_link < 0) - return -1; - - /* Check the closing ')' is not inside an already resolved range - * (i.e. a range with a higher priority), e.g. a code span. */ - if(is_link) { - int i = closer_index + 1; - - while(i < ctx->n_marks) { - MD_MARK* mark = &ctx->marks[i]; - - if(mark->beg >= inline_link_end) - break; - if((mark->flags & (MD_MARK_OPENER | MD_MARK_RESOLVED)) == (MD_MARK_OPENER | MD_MARK_RESOLVED)) { - if(ctx->marks[mark->next].beg >= inline_link_end) { - /* Cancel the link status. */ - if(attr.title_needs_free) - free(attr.title); - is_link = FALSE; - break; - } - - i = mark->next + 1; - } else { - i++; - } - } - } - - if(is_link) { - /* Eat the "(...)" */ - closer->end = inline_link_end; - } - } - - if(!is_link) { - /* Might be collapsed reference link. */ - is_link = md_is_link_reference(ctx, lines, n_lines, opener->beg, closer->end, &attr); - if(is_link < 0) - return -1; - } - } - - if(is_link) { - /* Resolve the brackets as a link. */ - opener->flags |= MD_MARK_OPENER | MD_MARK_RESOLVED; - closer->flags |= MD_MARK_CLOSER | MD_MARK_RESOLVED; - - /* If it is a link, we store the destination and title in the two - * dummy marks after the opener. */ - MD_ASSERT(ctx->marks[opener_index+1].ch == 'D'); - ctx->marks[opener_index+1].beg = attr.dest_beg; - ctx->marks[opener_index+1].end = attr.dest_end; - - MD_ASSERT(ctx->marks[opener_index+2].ch == 'D'); - md_mark_store_ptr(ctx, opener_index+2, attr.title); - /* The title might or might not have been allocated for us. */ - if(attr.title_needs_free) - md_mark_chain_append(ctx, &PTR_CHAIN, opener_index+2); - ctx->marks[opener_index+2].prev = attr.title_size; - - if(opener->ch == '[') { - last_link_beg = opener->beg; - last_link_end = closer->end; - } else { - last_img_beg = opener->beg; - last_img_end = closer->end; - } - - md_analyze_link_contents(ctx, lines, n_lines, opener_index+1, closer_index); - - /* If the link text is formed by nothing but permissive autolink, - * suppress the autolink. - * See https://github.com/mity/md4c/issues/152 for more info. */ - if(ctx->parser.flags & MD_FLAG_PERMISSIVEAUTOLINKS) { - MD_MARK* first_nested; - MD_MARK* last_nested; - - first_nested = opener + 1; - while(first_nested->ch == _T('D') && first_nested < closer) - first_nested++; - - last_nested = closer - 1; - while(first_nested->ch == _T('D') && last_nested > opener) - last_nested--; - - if((first_nested->flags & MD_MARK_RESOLVED) && - first_nested->beg == opener->end && - ISANYOF_(first_nested->ch, _T("@:.")) && - first_nested->next == (last_nested - ctx->marks) && - last_nested->end == closer->beg) - { - first_nested->ch = _T('D'); - first_nested->flags &= ~MD_MARK_RESOLVED; - last_nested->ch = _T('D'); - last_nested->flags &= ~MD_MARK_RESOLVED; - } - } - } - - opener_index = next_index; - } - - return 0; -} - -/* Analyze whether the mark '&' starts a HTML entity. - * If so, update its flags as well as flags of corresponding closer ';'. */ -static void -md_analyze_entity(MD_CTX* ctx, int mark_index) -{ - MD_MARK* opener = &ctx->marks[mark_index]; - MD_MARK* closer; - OFF off; - - /* Cannot be entity if there is no closer as the next mark. - * (Any other mark between would mean strange character which cannot be - * part of the entity. - * - * So we can do all the work on '&' and do not call this later for the - * closing mark ';'. - */ - if(mark_index + 1 >= ctx->n_marks) - return; - closer = &ctx->marks[mark_index+1]; - if(closer->ch != ';') - return; - - if(md_is_entity(ctx, opener->beg, closer->end, &off)) { - MD_ASSERT(off == closer->end); - - md_resolve_range(ctx, NULL, mark_index, mark_index+1); - opener->end = closer->end; - } -} - -static void -md_analyze_table_cell_boundary(MD_CTX* ctx, int mark_index) -{ - MD_MARK* mark = &ctx->marks[mark_index]; - mark->flags |= MD_MARK_RESOLVED; - - md_mark_chain_append(ctx, &TABLECELLBOUNDARIES, mark_index); - ctx->n_table_cell_boundaries++; -} - -/* Split a longer mark into two. The new mark takes the given count of - * characters. May only be called if an adequate number of dummy 'D' marks - * follows. - */ -static int -md_split_emph_mark(MD_CTX* ctx, int mark_index, SZ n) -{ - MD_MARK* mark = &ctx->marks[mark_index]; - int new_mark_index = mark_index + (mark->end - mark->beg - n); - MD_MARK* dummy = &ctx->marks[new_mark_index]; - - MD_ASSERT(mark->end - mark->beg > n); - MD_ASSERT(dummy->ch == 'D'); - - memcpy(dummy, mark, sizeof(MD_MARK)); - mark->end -= n; - dummy->beg = mark->end; - - return new_mark_index; -} - -static void -md_analyze_emph(MD_CTX* ctx, int mark_index) -{ - MD_MARK* mark = &ctx->marks[mark_index]; - MD_MARKCHAIN* chain = md_mark_chain(ctx, mark_index); - - /* If we can be a closer, try to resolve with the preceding opener. */ - if(mark->flags & MD_MARK_POTENTIAL_CLOSER) { - MD_MARK* opener = NULL; - int opener_index = 0; - - if(mark->ch == _T('*')) { - MD_MARKCHAIN* opener_chains[6]; - int i, n_opener_chains; - unsigned flags = mark->flags; - - /* Apply the "rule of three". */ - n_opener_chains = 0; - opener_chains[n_opener_chains++] = &ASTERISK_OPENERS_intraword_mod3_0; - if((flags & MD_MARK_EMPH_MOD3_MASK) != MD_MARK_EMPH_MOD3_2) - opener_chains[n_opener_chains++] = &ASTERISK_OPENERS_intraword_mod3_1; - if((flags & MD_MARK_EMPH_MOD3_MASK) != MD_MARK_EMPH_MOD3_1) - opener_chains[n_opener_chains++] = &ASTERISK_OPENERS_intraword_mod3_2; - opener_chains[n_opener_chains++] = &ASTERISK_OPENERS_extraword_mod3_0; - if(!(flags & MD_MARK_EMPH_INTRAWORD) || (flags & MD_MARK_EMPH_MOD3_MASK) != MD_MARK_EMPH_MOD3_2) - opener_chains[n_opener_chains++] = &ASTERISK_OPENERS_extraword_mod3_1; - if(!(flags & MD_MARK_EMPH_INTRAWORD) || (flags & MD_MARK_EMPH_MOD3_MASK) != MD_MARK_EMPH_MOD3_1) - opener_chains[n_opener_chains++] = &ASTERISK_OPENERS_extraword_mod3_2; - - /* Opener is the most recent mark from the allowed chains. */ - for(i = 0; i < n_opener_chains; i++) { - if(opener_chains[i]->tail >= 0) { - int tmp_index = opener_chains[i]->tail; - MD_MARK* tmp_mark = &ctx->marks[tmp_index]; - if(opener == NULL || tmp_mark->end > opener->end) { - opener_index = tmp_index; - opener = tmp_mark; - } - } - } - } else { - /* Simple emph. mark */ - if(chain->tail >= 0) { - opener_index = chain->tail; - opener = &ctx->marks[opener_index]; - } - } - - /* Resolve, if we have found matching opener. */ - if(opener != NULL) { - SZ opener_size = opener->end - opener->beg; - SZ closer_size = mark->end - mark->beg; - MD_MARKCHAIN* opener_chain = md_mark_chain(ctx, opener_index); - - if(opener_size > closer_size) { - opener_index = md_split_emph_mark(ctx, opener_index, closer_size); - md_mark_chain_append(ctx, opener_chain, opener_index); - } else if(opener_size < closer_size) { - md_split_emph_mark(ctx, mark_index, closer_size - opener_size); - } - - md_rollback(ctx, opener_index, mark_index, MD_ROLLBACK_CROSSING); - md_resolve_range(ctx, opener_chain, opener_index, mark_index); - return; - } - } - - /* If we could not resolve as closer, we may be yet be an opener. */ - if(mark->flags & MD_MARK_POTENTIAL_OPENER) - md_mark_chain_append(ctx, chain, mark_index); -} - -static void -md_analyze_tilde(MD_CTX* ctx, int mark_index) -{ - MD_MARK* mark = &ctx->marks[mark_index]; - MD_MARKCHAIN* chain = md_mark_chain(ctx, mark_index); - - /* We attempt to be Github Flavored Markdown compatible here. GFM accepts - * only tildes sequences of length 1 and 2, and the length of the opener - * and closer has to match. */ - - if((mark->flags & MD_MARK_POTENTIAL_CLOSER) && chain->head >= 0) { - int opener_index = chain->head; - - md_rollback(ctx, opener_index, mark_index, MD_ROLLBACK_CROSSING); - md_resolve_range(ctx, chain, opener_index, mark_index); - return; - } - - if(mark->flags & MD_MARK_POTENTIAL_OPENER) - md_mark_chain_append(ctx, chain, mark_index); -} - -static void -md_analyze_dollar(MD_CTX* ctx, int mark_index) -{ - /* This should mimic the way inline equations work in LaTeX, so there - * can only ever be one item in the chain (i.e. the dollars can't be - * nested). This is basically the same as the md_analyze_tilde function, - * except that we require matching openers and closers to be of the same - * length. - * - * E.g.: $abc$$def$$ => abc (display equation) def (end equation) */ - if(DOLLAR_OPENERS.head >= 0) { - /* If the potential closer has a non-matching number of $, discard */ - MD_MARK* open = &ctx->marks[DOLLAR_OPENERS.head]; - MD_MARK* close = &ctx->marks[mark_index]; - - int opener_index = DOLLAR_OPENERS.head; - md_rollback(ctx, opener_index, mark_index, MD_ROLLBACK_ALL); - if (open->end - open->beg == close->end - close->beg) { - /* We are the matching closer */ - md_resolve_range(ctx, &DOLLAR_OPENERS, opener_index, mark_index); - } else { - /* We don't match the opener, so discard old opener and insert as opener */ - md_mark_chain_append(ctx, &DOLLAR_OPENERS, mark_index); - } - } else { - /* No unmatched openers, so we are opener */ - md_mark_chain_append(ctx, &DOLLAR_OPENERS, mark_index); - } -} - -static void -md_analyze_permissive_url_autolink(MD_CTX* ctx, int mark_index) -{ - MD_MARK* opener = &ctx->marks[mark_index]; - int closer_index = mark_index + 1; - MD_MARK* closer = &ctx->marks[closer_index]; - MD_MARK* next_resolved_mark; - OFF off = opener->end; - int n_dots = FALSE; - int has_underscore_in_last_seg = FALSE; - int has_underscore_in_next_to_last_seg = FALSE; - int n_opened_parenthesis = 0; - int n_excess_parenthesis = 0; - - /* Check for domain. */ - while(off < ctx->size) { - if(ISALNUM(off) || CH(off) == _T('-')) { - off++; - } else if(CH(off) == _T('.')) { - /* We must see at least one period. */ - n_dots++; - has_underscore_in_next_to_last_seg = has_underscore_in_last_seg; - has_underscore_in_last_seg = FALSE; - off++; - } else if(CH(off) == _T('_')) { - /* No underscore may be present in the last two domain segments. */ - has_underscore_in_last_seg = TRUE; - off++; - } else { - break; - } - } - if(off > opener->end && CH(off-1) == _T('.')) { - off--; - n_dots--; - } - if(off <= opener->end || n_dots == 0 || has_underscore_in_next_to_last_seg || has_underscore_in_last_seg) - return; - - /* Check for path. */ - next_resolved_mark = closer + 1; - while(next_resolved_mark->ch == 'D' || !(next_resolved_mark->flags & MD_MARK_RESOLVED)) - next_resolved_mark++; - while(off < next_resolved_mark->beg && CH(off) != _T('<') && !ISWHITESPACE(off) && !ISNEWLINE(off)) { - /* Parenthesis must be balanced. */ - if(CH(off) == _T('(')) { - n_opened_parenthesis++; - } else if(CH(off) == _T(')')) { - if(n_opened_parenthesis > 0) - n_opened_parenthesis--; - else - n_excess_parenthesis++; - } - - off++; - } - - /* Trim a trailing punctuation from the end. */ - while(TRUE) { - if(ISANYOF(off-1, _T("?!.,:*_~"))) { - off--; - } else if(CH(off-1) == ')' && n_excess_parenthesis > 0) { - /* Unmatched ')' can be in an interior of the path but not at the - * of it, so the auto-link may be safely nested in a parenthesis - * pair. */ - off--; - n_excess_parenthesis--; - } else { - break; - } - } - - /* Ok. Lets call it an auto-link. Adapt opener and create closer to zero - * length so all the contents becomes the link text. */ - MD_ASSERT(closer->ch == 'D'); - opener->end = opener->beg; - closer->ch = opener->ch; - closer->beg = off; - closer->end = off; - md_resolve_range(ctx, NULL, mark_index, closer_index); -} - -/* The permissive autolinks do not have to be enclosed in '<' '>' but we - * instead impose stricter rules what is understood as an e-mail address - * here. Actually any non-alphanumeric characters with exception of '.' - * are prohibited both in username and after '@'. */ -static void -md_analyze_permissive_email_autolink(MD_CTX* ctx, int mark_index) -{ - MD_MARK* opener = &ctx->marks[mark_index]; - int closer_index; - MD_MARK* closer; - OFF beg = opener->beg; - OFF end = opener->end; - int dot_count = 0; - - MD_ASSERT(CH(beg) == _T('@')); - - /* Scan for name before '@'. */ - while(beg > 0 && (ISALNUM(beg-1) || ISANYOF(beg-1, _T(".-_+")))) - beg--; - - /* Scan for domain after '@'. */ - while(end < ctx->size && (ISALNUM(end) || ISANYOF(end, _T(".-_")))) { - if(CH(end) == _T('.')) - dot_count++; - end++; - } - if(CH(end-1) == _T('.')) { /* Final '.' not part of it. */ - dot_count--; - end--; - } - else if(ISANYOF2(end-1, _T('-'), _T('_'))) /* These are forbidden at the end. */ - return; - if(CH(end-1) == _T('@') || dot_count == 0) - return; - - /* Ok. Lets call it auto-link. Adapt opener and create closer to zero - * length so all the contents becomes the link text. */ - closer_index = mark_index + 1; - closer = &ctx->marks[closer_index]; - MD_ASSERT(closer->ch == 'D'); - - opener->beg = beg; - opener->end = beg; - closer->ch = opener->ch; - closer->beg = end; - closer->end = end; - md_resolve_range(ctx, NULL, mark_index, closer_index); -} - -static inline void -md_analyze_marks(MD_CTX* ctx, const MD_LINE* lines, int n_lines, - int mark_beg, int mark_end, const CHAR* mark_chars) -{ - int i = mark_beg; - MD_UNUSED(lines); - MD_UNUSED(n_lines); - - while(i < mark_end) { - MD_MARK* mark = &ctx->marks[i]; - - /* Skip resolved spans. */ - if(mark->flags & MD_MARK_RESOLVED) { - if(mark->flags & MD_MARK_OPENER) { - MD_ASSERT(i < mark->next); - i = mark->next + 1; - } else { - i++; - } - continue; - } - - /* Skip marks we do not want to deal with. */ - if(!ISANYOF_(mark->ch, mark_chars)) { - i++; - continue; - } - - /* Analyze the mark. */ - switch(mark->ch) { - case '[': /* Pass through. */ - case '!': /* Pass through. */ - case ']': md_analyze_bracket(ctx, i); break; - case '&': md_analyze_entity(ctx, i); break; - case '|': md_analyze_table_cell_boundary(ctx, i); break; - case '_': /* Pass through. */ - case '*': md_analyze_emph(ctx, i); break; - case '~': md_analyze_tilde(ctx, i); break; - case '$': md_analyze_dollar(ctx, i); break; - case '.': /* Pass through. */ - case ':': md_analyze_permissive_url_autolink(ctx, i); break; - case '@': md_analyze_permissive_email_autolink(ctx, i); break; - } - - i++; - } -} - -/* Analyze marks (build ctx->marks). */ -static int -md_analyze_inlines(MD_CTX* ctx, const MD_LINE* lines, int n_lines, int table_mode) -{ - int ret; - - /* Reset the previously collected stack of marks. */ - ctx->n_marks = 0; - - /* Collect all marks. */ - MD_CHECK(md_collect_marks(ctx, lines, n_lines, table_mode)); - - /* We analyze marks in few groups to handle their precedence. */ - /* (1) Entities; code spans; autolinks; raw HTML. */ - md_analyze_marks(ctx, lines, n_lines, 0, ctx->n_marks, _T("&")); - - /* (2) Links. */ - md_analyze_marks(ctx, lines, n_lines, 0, ctx->n_marks, _T("[]!")); - MD_CHECK(md_resolve_links(ctx, lines, n_lines)); - BRACKET_OPENERS.head = -1; - BRACKET_OPENERS.tail = -1; - ctx->unresolved_link_head = -1; - ctx->unresolved_link_tail = -1; - - if(table_mode) { - /* (3) Analyze table cell boundaries. - * Note we reset TABLECELLBOUNDARIES chain prior to the call md_analyze_marks(), - * not after, because caller may need it. */ - MD_ASSERT(n_lines == 1); - TABLECELLBOUNDARIES.head = -1; - TABLECELLBOUNDARIES.tail = -1; - ctx->n_table_cell_boundaries = 0; - md_analyze_marks(ctx, lines, n_lines, 0, ctx->n_marks, _T("|")); - return ret; - } - - /* (4) Emphasis and strong emphasis; permissive autolinks. */ - md_analyze_link_contents(ctx, lines, n_lines, 0, ctx->n_marks); - -abort: - return ret; -} - -static void -md_analyze_link_contents(MD_CTX* ctx, const MD_LINE* lines, int n_lines, - int mark_beg, int mark_end) -{ - int i; - - md_analyze_marks(ctx, lines, n_lines, mark_beg, mark_end, _T("*_~$@:.")); - - for(i = OPENERS_CHAIN_FIRST; i <= OPENERS_CHAIN_LAST; i++) { - ctx->mark_chains[i].head = -1; - ctx->mark_chains[i].tail = -1; - } -} - -static int -md_enter_leave_span_a(MD_CTX* ctx, int enter, MD_SPANTYPE type, - const CHAR* dest, SZ dest_size, int prohibit_escapes_in_dest, - const CHAR* title, SZ title_size) -{ - MD_ATTRIBUTE_BUILD href_build = { 0 }; - MD_ATTRIBUTE_BUILD title_build = { 0 }; - MD_SPAN_A_DETAIL det; - int ret = 0; - - /* Note we here rely on fact that MD_SPAN_A_DETAIL and - * MD_SPAN_IMG_DETAIL are binary-compatible. */ - memset(&det, 0, sizeof(MD_SPAN_A_DETAIL)); - MD_CHECK(md_build_attribute(ctx, dest, dest_size, - (prohibit_escapes_in_dest ? MD_BUILD_ATTR_NO_ESCAPES : 0), - &det.href, &href_build)); - MD_CHECK(md_build_attribute(ctx, title, title_size, 0, &det.title, &title_build)); - - if(enter) - MD_ENTER_SPAN(type, &det); - else - MD_LEAVE_SPAN(type, &det); - -abort: - md_free_attribute(ctx, &href_build); - md_free_attribute(ctx, &title_build); - return ret; -} - -static int -md_enter_leave_span_wikilink(MD_CTX* ctx, int enter, const CHAR* target, SZ target_size) -{ - MD_ATTRIBUTE_BUILD target_build = { 0 }; - MD_SPAN_WIKILINK_DETAIL det; - int ret = 0; - - memset(&det, 0, sizeof(MD_SPAN_WIKILINK_DETAIL)); - MD_CHECK(md_build_attribute(ctx, target, target_size, 0, &det.target, &target_build)); - - if (enter) - MD_ENTER_SPAN(MD_SPAN_WIKILINK, &det); - else - MD_LEAVE_SPAN(MD_SPAN_WIKILINK, &det); - -abort: - md_free_attribute(ctx, &target_build); - return ret; -} - - -/* Render the output, accordingly to the analyzed ctx->marks. */ -static int -md_process_inlines(MD_CTX* ctx, const MD_LINE* lines, int n_lines) -{ - MD_TEXTTYPE text_type; - const MD_LINE* line = lines; - MD_MARK* prev_mark = NULL; - MD_MARK* mark; - OFF off = lines[0].beg; - OFF end = lines[n_lines-1].end; - int enforce_hardbreak = 0; - int ret = 0; - - /* Find first resolved mark. Note there is always at least one resolved - * mark, the dummy last one after the end of the latest line we actually - * never really reach. This saves us of a lot of special checks and cases - * in this function. */ - mark = ctx->marks; - while(!(mark->flags & MD_MARK_RESOLVED)) - mark++; - - text_type = MD_TEXT_NORMAL; - - while(1) { - /* Process the text up to the next mark or end-of-line. */ - OFF tmp = (line->end < mark->beg ? line->end : mark->beg); - if(tmp > off) { - MD_TEXT(text_type, STR(off), tmp - off); - off = tmp; - } - - /* If reached the mark, process it and move to next one. */ - if(off >= mark->beg) { - switch(mark->ch) { - case '\\': /* Backslash escape. */ - if(ISNEWLINE(mark->beg+1)) - enforce_hardbreak = 1; - else - MD_TEXT(text_type, STR(mark->beg+1), 1); - break; - - case ' ': /* Non-trivial space. */ - MD_TEXT(text_type, _T(" "), 1); - break; - - case '`': /* Code span. */ - if(mark->flags & MD_MARK_OPENER) { - MD_ENTER_SPAN(MD_SPAN_CODE, NULL); - text_type = MD_TEXT_CODE; - } else { - MD_LEAVE_SPAN(MD_SPAN_CODE, NULL); - text_type = MD_TEXT_NORMAL; - } - break; - - case '_': /* Underline (or emphasis if we fall through). */ - if(ctx->parser.flags & MD_FLAG_UNDERLINE) { - if(mark->flags & MD_MARK_OPENER) { - while(off < mark->end) { - MD_ENTER_SPAN(MD_SPAN_U, NULL); - off++; - } - } else { - while(off < mark->end) { - MD_LEAVE_SPAN(MD_SPAN_U, NULL); - off++; - } - } - break; - } - MD_FALLTHROUGH(); - - case '*': /* Emphasis, strong emphasis. */ - if(mark->flags & MD_MARK_OPENER) { - if((mark->end - off) % 2) { - MD_ENTER_SPAN(MD_SPAN_EM, NULL); - off++; - } - while(off + 1 < mark->end) { - MD_ENTER_SPAN(MD_SPAN_STRONG, NULL); - off += 2; - } - } else { - while(off + 1 < mark->end) { - MD_LEAVE_SPAN(MD_SPAN_STRONG, NULL); - off += 2; - } - if((mark->end - off) % 2) { - MD_LEAVE_SPAN(MD_SPAN_EM, NULL); - off++; - } - } - break; - - case '~': - if(mark->flags & MD_MARK_OPENER) - MD_ENTER_SPAN(MD_SPAN_DEL, NULL); - else - MD_LEAVE_SPAN(MD_SPAN_DEL, NULL); - break; - - case '$': - if(mark->flags & MD_MARK_OPENER) { - MD_ENTER_SPAN((mark->end - off) % 2 ? MD_SPAN_LATEXMATH : MD_SPAN_LATEXMATH_DISPLAY, NULL); - text_type = MD_TEXT_LATEXMATH; - } else { - MD_LEAVE_SPAN((mark->end - off) % 2 ? MD_SPAN_LATEXMATH : MD_SPAN_LATEXMATH_DISPLAY, NULL); - text_type = MD_TEXT_NORMAL; - } - break; - - case '[': /* Link, wiki link, image. */ - case '!': - case ']': - { - const MD_MARK* opener = (mark->ch != ']' ? mark : &ctx->marks[mark->prev]); - const MD_MARK* closer = &ctx->marks[opener->next]; - const MD_MARK* dest_mark; - const MD_MARK* title_mark; - - if ((opener->ch == '[' && closer->ch == ']') && - opener->end - opener->beg >= 2 && - closer->end - closer->beg >= 2) - { - int has_label = (opener->end - opener->beg > 2); - SZ target_sz; - - if(has_label) - target_sz = opener->end - (opener->beg+2); - else - target_sz = closer->beg - opener->end; - - MD_CHECK(md_enter_leave_span_wikilink(ctx, (mark->ch != ']'), - has_label ? STR(opener->beg+2) : STR(opener->end), - target_sz)); - - break; - } - - dest_mark = opener+1; - MD_ASSERT(dest_mark->ch == 'D'); - title_mark = opener+2; - MD_ASSERT(title_mark->ch == 'D'); - - MD_CHECK(md_enter_leave_span_a(ctx, (mark->ch != ']'), - (opener->ch == '!' ? MD_SPAN_IMG : MD_SPAN_A), - STR(dest_mark->beg), dest_mark->end - dest_mark->beg, FALSE, - md_mark_get_ptr(ctx, (int)(title_mark - ctx->marks)), - title_mark->prev)); - - /* link/image closer may span multiple lines. */ - if(mark->ch == ']') { - while(mark->end > line->end) - line++; - } - - break; - } - - case '<': - case '>': /* Autolink or raw HTML. */ - if(!(mark->flags & MD_MARK_AUTOLINK)) { - /* Raw HTML. */ - if(mark->flags & MD_MARK_OPENER) - text_type = MD_TEXT_HTML; - else - text_type = MD_TEXT_NORMAL; - break; - } - /* Pass through, if auto-link. */ - MD_FALLTHROUGH(); - - case '@': /* Permissive e-mail autolink. */ - case ':': /* Permissive URL autolink. */ - case '.': /* Permissive WWW autolink. */ - { - MD_MARK* opener = ((mark->flags & MD_MARK_OPENER) ? mark : &ctx->marks[mark->prev]); - MD_MARK* closer = &ctx->marks[opener->next]; - const CHAR* dest = STR(opener->end); - SZ dest_size = closer->beg - opener->end; - - /* For permissive auto-links we do not know closer mark - * position at the time of md_collect_marks(), therefore - * it can be out-of-order in ctx->marks[]. - * - * With this flag, we make sure that we output the closer - * only if we processed the opener. */ - if(mark->flags & MD_MARK_OPENER) - closer->flags |= MD_MARK_VALIDPERMISSIVEAUTOLINK; - - if(opener->ch == '@' || opener->ch == '.') { - dest_size += 7; - MD_TEMP_BUFFER(dest_size * sizeof(CHAR)); - memcpy(ctx->buffer, - (opener->ch == '@' ? _T("mailto:") : _T("http://")), - 7 * sizeof(CHAR)); - memcpy(ctx->buffer + 7, dest, (dest_size-7) * sizeof(CHAR)); - dest = ctx->buffer; - } - - if(closer->flags & MD_MARK_VALIDPERMISSIVEAUTOLINK) - MD_CHECK(md_enter_leave_span_a(ctx, (mark->flags & MD_MARK_OPENER), - MD_SPAN_A, dest, dest_size, TRUE, NULL, 0)); - break; - } - - case '&': /* Entity. */ - MD_TEXT(MD_TEXT_ENTITY, STR(mark->beg), mark->end - mark->beg); - break; - - case '\0': - MD_TEXT(MD_TEXT_NULLCHAR, _T(""), 1); - break; - - case 127: - goto abort; - } - - off = mark->end; - - /* Move to next resolved mark. */ - prev_mark = mark; - mark++; - while(!(mark->flags & MD_MARK_RESOLVED) || mark->beg < off) - mark++; - } - - /* If reached end of line, move to next one. */ - if(off >= line->end) { - /* If it is the last line, we are done. */ - if(off >= end) - break; - - if(text_type == MD_TEXT_CODE || text_type == MD_TEXT_LATEXMATH) { - OFF tmp; - - MD_ASSERT(prev_mark != NULL); - MD_ASSERT(ISANYOF2_(prev_mark->ch, '`', '$') && (prev_mark->flags & MD_MARK_OPENER)); - MD_ASSERT(ISANYOF2_(mark->ch, '`', '$') && (mark->flags & MD_MARK_CLOSER)); - - /* Inside a code span, trailing line whitespace has to be - * outputted. */ - tmp = off; - while(off < ctx->size && ISBLANK(off)) - off++; - if(off > tmp) - MD_TEXT(text_type, STR(tmp), off-tmp); - - /* and new lines are transformed into single spaces. */ - if(prev_mark->end < off && off < mark->beg) - MD_TEXT(text_type, _T(" "), 1); - } else if(text_type == MD_TEXT_HTML) { - /* Inside raw HTML, we output the new line verbatim, including - * any trailing spaces. */ - OFF tmp = off; - - while(tmp < end && ISBLANK(tmp)) - tmp++; - if(tmp > off) - MD_TEXT(MD_TEXT_HTML, STR(off), tmp - off); - MD_TEXT(MD_TEXT_HTML, _T("\n"), 1); - } else { - /* Output soft or hard line break. */ - MD_TEXTTYPE break_type = MD_TEXT_SOFTBR; - - if(text_type == MD_TEXT_NORMAL) { - if(enforce_hardbreak) - break_type = MD_TEXT_BR; - else if((CH(line->end) == _T(' ') && CH(line->end+1) == _T(' '))) - break_type = MD_TEXT_BR; - } - - MD_TEXT(break_type, _T("\n"), 1); - } - - /* Move to the next line. */ - line++; - off = line->beg; - - enforce_hardbreak = 0; - } - } - -abort: - return ret; -} - - -/*************************** - *** Processing Tables *** - ***************************/ - -static void -md_analyze_table_alignment(MD_CTX* ctx, OFF beg, OFF end, MD_ALIGN* align, int n_align) -{ - static const MD_ALIGN align_map[] = { MD_ALIGN_DEFAULT, MD_ALIGN_LEFT, MD_ALIGN_RIGHT, MD_ALIGN_CENTER }; - OFF off = beg; - - while(n_align > 0) { - int index = 0; /* index into align_map[] */ - - while(CH(off) != _T('-')) - off++; - if(off > beg && CH(off-1) == _T(':')) - index |= 1; - while(off < end && CH(off) == _T('-')) - off++; - if(off < end && CH(off) == _T(':')) - index |= 2; - - *align = align_map[index]; - align++; - n_align--; - } - -} - -/* Forward declaration. */ -static int md_process_normal_block_contents(MD_CTX* ctx, const MD_LINE* lines, int n_lines); - -static int -md_process_table_cell(MD_CTX* ctx, MD_BLOCKTYPE cell_type, MD_ALIGN align, OFF beg, OFF end) -{ - MD_LINE line; - MD_BLOCK_TD_DETAIL det; - int ret = 0; - - while(beg < end && ISWHITESPACE(beg)) - beg++; - while(end > beg && ISWHITESPACE(end-1)) - end--; - - det.align = align; - line.beg = beg; - line.end = end; - - MD_ENTER_BLOCK(cell_type, &det); - MD_CHECK(md_process_normal_block_contents(ctx, &line, 1)); - MD_LEAVE_BLOCK(cell_type, &det); - -abort: - return ret; -} - -static int -md_process_table_row(MD_CTX* ctx, MD_BLOCKTYPE cell_type, OFF beg, OFF end, - const MD_ALIGN* align, int col_count) -{ - MD_LINE line; - OFF* pipe_offs = NULL; - int i, j, k, n; - int ret = 0; - - line.beg = beg; - line.end = end; - - /* Break the line into table cells by identifying pipe characters who - * form the cell boundary. */ - MD_CHECK(md_analyze_inlines(ctx, &line, 1, TRUE)); - - /* We have to remember the cell boundaries in local buffer because - * ctx->marks[] shall be reused during cell contents processing. */ - n = ctx->n_table_cell_boundaries + 2; - pipe_offs = (OFF*) malloc(n * sizeof(OFF)); - if(pipe_offs == NULL) { - MD_LOG("malloc() failed."); - ret = -1; - goto abort; - } - j = 0; - pipe_offs[j++] = beg; - for(i = TABLECELLBOUNDARIES.head; i >= 0; i = ctx->marks[i].next) { - MD_MARK* mark = &ctx->marks[i]; - pipe_offs[j++] = mark->end; - } - pipe_offs[j++] = end+1; - - /* Process cells. */ - MD_ENTER_BLOCK(MD_BLOCK_TR, NULL); - k = 0; - for(i = 0; i < j-1 && k < col_count; i++) { - if(pipe_offs[i] < pipe_offs[i+1]-1) - MD_CHECK(md_process_table_cell(ctx, cell_type, align[k++], pipe_offs[i], pipe_offs[i+1]-1)); - } - /* Make sure we call enough table cells even if the current table contains - * too few of them. */ - while(k < col_count) - MD_CHECK(md_process_table_cell(ctx, cell_type, align[k++], 0, 0)); - MD_LEAVE_BLOCK(MD_BLOCK_TR, NULL); - -abort: - free(pipe_offs); - - /* Free any temporary memory blocks stored within some dummy marks. */ - for(i = PTR_CHAIN.head; i >= 0; i = ctx->marks[i].next) - free(md_mark_get_ptr(ctx, i)); - PTR_CHAIN.head = -1; - PTR_CHAIN.tail = -1; - - return ret; -} - -static int -md_process_table_block_contents(MD_CTX* ctx, int col_count, const MD_LINE* lines, int n_lines) -{ - MD_ALIGN* align; - int i; - int ret = 0; - - /* At least two lines have to be present: The column headers and the line - * with the underlines. */ - MD_ASSERT(n_lines >= 2); - - align = malloc(col_count * sizeof(MD_ALIGN)); - if(align == NULL) { - MD_LOG("malloc() failed."); - ret = -1; - goto abort; - } - - md_analyze_table_alignment(ctx, lines[1].beg, lines[1].end, align, col_count); - - MD_ENTER_BLOCK(MD_BLOCK_THEAD, NULL); - MD_CHECK(md_process_table_row(ctx, MD_BLOCK_TH, - lines[0].beg, lines[0].end, align, col_count)); - MD_LEAVE_BLOCK(MD_BLOCK_THEAD, NULL); - - if(n_lines > 2) { - MD_ENTER_BLOCK(MD_BLOCK_TBODY, NULL); - for(i = 2; i < n_lines; i++) { - MD_CHECK(md_process_table_row(ctx, MD_BLOCK_TD, - lines[i].beg, lines[i].end, align, col_count)); - } - MD_LEAVE_BLOCK(MD_BLOCK_TBODY, NULL); - } - -abort: - free(align); - return ret; -} - - -/************************** - *** Processing Block *** - **************************/ - -#define MD_BLOCK_CONTAINER_OPENER 0x01 -#define MD_BLOCK_CONTAINER_CLOSER 0x02 -#define MD_BLOCK_CONTAINER (MD_BLOCK_CONTAINER_OPENER | MD_BLOCK_CONTAINER_CLOSER) -#define MD_BLOCK_LOOSE_LIST 0x04 -#define MD_BLOCK_SETEXT_HEADER 0x08 - -struct MD_BLOCK_tag { - MD_BLOCKTYPE type : 8; - unsigned flags : 8; - - /* MD_BLOCK_H: Header level (1 - 6) - * MD_BLOCK_CODE: Non-zero if fenced, zero if indented. - * MD_BLOCK_LI: Task mark character (0 if not task list item, 'x', 'X' or ' '). - * MD_BLOCK_TABLE: Column count (as determined by the table underline). - */ - unsigned data : 16; - - /* Leaf blocks: Count of lines (MD_LINE or MD_VERBATIMLINE) on the block. - * MD_BLOCK_LI: Task mark offset in the input doc. - * MD_BLOCK_OL: Start item number. - */ - unsigned n_lines; -}; - -struct MD_CONTAINER_tag { - CHAR ch; - unsigned is_loose : 8; - unsigned is_task : 8; - unsigned start; - unsigned mark_indent; - unsigned contents_indent; - OFF block_byte_off; - OFF task_mark_off; -}; - - -static int -md_process_normal_block_contents(MD_CTX* ctx, const MD_LINE* lines, int n_lines) -{ - int i; - int ret; - - MD_CHECK(md_analyze_inlines(ctx, lines, n_lines, FALSE)); - MD_CHECK(md_process_inlines(ctx, lines, n_lines)); - -abort: - /* Free any temporary memory blocks stored within some dummy marks. */ - for(i = PTR_CHAIN.head; i >= 0; i = ctx->marks[i].next) - free(md_mark_get_ptr(ctx, i)); - PTR_CHAIN.head = -1; - PTR_CHAIN.tail = -1; - - return ret; -} - -static int -md_process_verbatim_block_contents(MD_CTX* ctx, MD_TEXTTYPE text_type, const MD_VERBATIMLINE* lines, int n_lines) -{ - static const CHAR indent_chunk_str[] = _T(" "); - static const SZ indent_chunk_size = SIZEOF_ARRAY(indent_chunk_str) - 1; - - int i; - int ret = 0; - - for(i = 0; i < n_lines; i++) { - const MD_VERBATIMLINE* line = &lines[i]; - int indent = line->indent; - - MD_ASSERT(indent >= 0); - - /* Output code indentation. */ - while(indent > (int) indent_chunk_size) { - MD_TEXT(text_type, indent_chunk_str, indent_chunk_size); - indent -= indent_chunk_size; - } - if(indent > 0) - MD_TEXT(text_type, indent_chunk_str, indent); - - /* Output the code line itself. */ - MD_TEXT_INSECURE(text_type, STR(line->beg), line->end - line->beg); - - /* Enforce end-of-line. */ - MD_TEXT(text_type, _T("\n"), 1); - } - -abort: - return ret; -} - -static int -md_process_code_block_contents(MD_CTX* ctx, int is_fenced, const MD_VERBATIMLINE* lines, int n_lines) -{ - if(is_fenced) { - /* Skip the first line in case of fenced code: It is the fence. - * (Only the starting fence is present due to logic in md_analyze_line().) */ - lines++; - n_lines--; - } else { - /* Ignore blank lines at start/end of indented code block. */ - while(n_lines > 0 && lines[0].beg == lines[0].end) { - lines++; - n_lines--; - } - while(n_lines > 0 && lines[n_lines-1].beg == lines[n_lines-1].end) { - n_lines--; - } - } - - if(n_lines == 0) - return 0; - - return md_process_verbatim_block_contents(ctx, MD_TEXT_CODE, lines, n_lines); -} - -static int -md_setup_fenced_code_detail(MD_CTX* ctx, const MD_BLOCK* block, MD_BLOCK_CODE_DETAIL* det, - MD_ATTRIBUTE_BUILD* info_build, MD_ATTRIBUTE_BUILD* lang_build) -{ - const MD_VERBATIMLINE* fence_line = (const MD_VERBATIMLINE*)(block + 1); - OFF beg = fence_line->beg; - OFF end = fence_line->end; - OFF lang_end; - CHAR fence_ch = CH(fence_line->beg); - int ret = 0; - - /* Skip the fence itself. */ - while(beg < ctx->size && CH(beg) == fence_ch) - beg++; - /* Trim initial spaces. */ - while(beg < ctx->size && CH(beg) == _T(' ')) - beg++; - - /* Trim trailing spaces. */ - while(end > beg && CH(end-1) == _T(' ')) - end--; - - /* Build info string attribute. */ - MD_CHECK(md_build_attribute(ctx, STR(beg), end - beg, 0, &det->info, info_build)); - - /* Build info string attribute. */ - lang_end = beg; - while(lang_end < end && !ISWHITESPACE(lang_end)) - lang_end++; - MD_CHECK(md_build_attribute(ctx, STR(beg), lang_end - beg, 0, &det->lang, lang_build)); - - det->fence_char = fence_ch; - -abort: - return ret; -} - -static int -md_process_leaf_block(MD_CTX* ctx, const MD_BLOCK* block) -{ - union { - MD_BLOCK_H_DETAIL header; - MD_BLOCK_CODE_DETAIL code; - MD_BLOCK_TABLE_DETAIL table; - } det; - MD_ATTRIBUTE_BUILD info_build; - MD_ATTRIBUTE_BUILD lang_build; - int is_in_tight_list; - int clean_fence_code_detail = FALSE; - int ret = 0; - - memset(&det, 0, sizeof(det)); - - if(ctx->n_containers == 0) - is_in_tight_list = FALSE; - else - is_in_tight_list = !ctx->containers[ctx->n_containers-1].is_loose; - - switch(block->type) { - case MD_BLOCK_H: - det.header.level = block->data; - break; - - case MD_BLOCK_CODE: - /* For fenced code block, we may need to set the info string. */ - if(block->data != 0) { - memset(&det.code, 0, sizeof(MD_BLOCK_CODE_DETAIL)); - clean_fence_code_detail = TRUE; - MD_CHECK(md_setup_fenced_code_detail(ctx, block, &det.code, &info_build, &lang_build)); - } - break; - - case MD_BLOCK_TABLE: - det.table.col_count = block->data; - det.table.head_row_count = 1; - det.table.body_row_count = block->n_lines - 2; - break; - - default: - /* Noop. */ - break; - } - - if(!is_in_tight_list || block->type != MD_BLOCK_P) - MD_ENTER_BLOCK(block->type, (void*) &det); - - /* Process the block contents accordingly to is type. */ - switch(block->type) { - case MD_BLOCK_HR: - /* noop */ - break; - - case MD_BLOCK_CODE: - MD_CHECK(md_process_code_block_contents(ctx, (block->data != 0), - (const MD_VERBATIMLINE*)(block + 1), block->n_lines)); - break; - - case MD_BLOCK_HTML: - MD_CHECK(md_process_verbatim_block_contents(ctx, MD_TEXT_HTML, - (const MD_VERBATIMLINE*)(block + 1), block->n_lines)); - break; - - case MD_BLOCK_TABLE: - MD_CHECK(md_process_table_block_contents(ctx, block->data, - (const MD_LINE*)(block + 1), block->n_lines)); - break; - - default: - MD_CHECK(md_process_normal_block_contents(ctx, - (const MD_LINE*)(block + 1), block->n_lines)); - break; - } - - if(!is_in_tight_list || block->type != MD_BLOCK_P) - MD_LEAVE_BLOCK(block->type, (void*) &det); - -abort: - if(clean_fence_code_detail) { - md_free_attribute(ctx, &info_build); - md_free_attribute(ctx, &lang_build); - } - return ret; -} - -static int -md_process_all_blocks(MD_CTX* ctx) -{ - int byte_off = 0; - int ret = 0; - - /* ctx->containers now is not needed for detection of lists and list items - * so we reuse it for tracking what lists are loose or tight. We rely - * on the fact the vector is large enough to hold the deepest nesting - * level of lists. */ - ctx->n_containers = 0; - - while(byte_off < ctx->n_block_bytes) { - MD_BLOCK* block = (MD_BLOCK*)((char*)ctx->block_bytes + byte_off); - union { - MD_BLOCK_UL_DETAIL ul; - MD_BLOCK_OL_DETAIL ol; - MD_BLOCK_LI_DETAIL li; - } det; - - switch(block->type) { - case MD_BLOCK_UL: - det.ul.is_tight = (block->flags & MD_BLOCK_LOOSE_LIST) ? FALSE : TRUE; - det.ul.mark = (CHAR) block->data; - break; - - case MD_BLOCK_OL: - det.ol.start = block->n_lines; - det.ol.is_tight = (block->flags & MD_BLOCK_LOOSE_LIST) ? FALSE : TRUE; - det.ol.mark_delimiter = (CHAR) block->data; - break; - - case MD_BLOCK_LI: - det.li.is_task = (block->data != 0); - det.li.task_mark = (CHAR) block->data; - det.li.task_mark_offset = (OFF) block->n_lines; - break; - - default: - /* noop */ - break; - } - - if(block->flags & MD_BLOCK_CONTAINER) { - if(block->flags & MD_BLOCK_CONTAINER_CLOSER) { - MD_LEAVE_BLOCK(block->type, &det); - - if(block->type == MD_BLOCK_UL || block->type == MD_BLOCK_OL || block->type == MD_BLOCK_QUOTE) - ctx->n_containers--; - } - - if(block->flags & MD_BLOCK_CONTAINER_OPENER) { - MD_ENTER_BLOCK(block->type, &det); - - if(block->type == MD_BLOCK_UL || block->type == MD_BLOCK_OL) { - ctx->containers[ctx->n_containers].is_loose = (block->flags & MD_BLOCK_LOOSE_LIST); - ctx->n_containers++; - } else if(block->type == MD_BLOCK_QUOTE) { - /* This causes that any text in a block quote, even if - * nested inside a tight list item, is wrapped with - * <p>...</p>. */ - ctx->containers[ctx->n_containers].is_loose = TRUE; - ctx->n_containers++; - } - } - } else { - MD_CHECK(md_process_leaf_block(ctx, block)); - - if(block->type == MD_BLOCK_CODE || block->type == MD_BLOCK_HTML) - byte_off += block->n_lines * sizeof(MD_VERBATIMLINE); - else - byte_off += block->n_lines * sizeof(MD_LINE); - } - - byte_off += sizeof(MD_BLOCK); - } - - ctx->n_block_bytes = 0; - -abort: - return ret; -} - - -/************************************ - *** Grouping Lines into Blocks *** - ************************************/ - -static void* -md_push_block_bytes(MD_CTX* ctx, int n_bytes) -{ - void* ptr; - - if(ctx->n_block_bytes + n_bytes > ctx->alloc_block_bytes) { - void* new_block_bytes; - - ctx->alloc_block_bytes = (ctx->alloc_block_bytes > 0 - ? ctx->alloc_block_bytes + ctx->alloc_block_bytes / 2 - : 512); - new_block_bytes = realloc(ctx->block_bytes, ctx->alloc_block_bytes); - if(new_block_bytes == NULL) { - MD_LOG("realloc() failed."); - return NULL; - } - - /* Fix the ->current_block after the reallocation. */ - if(ctx->current_block != NULL) { - OFF off_current_block = (OFF) ((char*) ctx->current_block - (char*) ctx->block_bytes); - ctx->current_block = (MD_BLOCK*) ((char*) new_block_bytes + off_current_block); - } - - ctx->block_bytes = new_block_bytes; - } - - ptr = (char*)ctx->block_bytes + ctx->n_block_bytes; - ctx->n_block_bytes += n_bytes; - return ptr; -} - -static int -md_start_new_block(MD_CTX* ctx, const MD_LINE_ANALYSIS* line) -{ - MD_BLOCK* block; - - MD_ASSERT(ctx->current_block == NULL); - - block = (MD_BLOCK*) md_push_block_bytes(ctx, sizeof(MD_BLOCK)); - if(block == NULL) - return -1; - - switch(line->type) { - case MD_LINE_HR: - block->type = MD_BLOCK_HR; - break; - - case MD_LINE_ATXHEADER: - case MD_LINE_SETEXTHEADER: - block->type = MD_BLOCK_H; - break; - - case MD_LINE_FENCEDCODE: - case MD_LINE_INDENTEDCODE: - block->type = MD_BLOCK_CODE; - break; - - case MD_LINE_TEXT: - block->type = MD_BLOCK_P; - break; - - case MD_LINE_HTML: - block->type = MD_BLOCK_HTML; - break; - - case MD_LINE_BLANK: - case MD_LINE_SETEXTUNDERLINE: - case MD_LINE_TABLEUNDERLINE: - default: - MD_UNREACHABLE(); - break; - } - - block->flags = 0; - block->data = line->data; - block->n_lines = 0; - - ctx->current_block = block; - return 0; -} - -/* Eat from start of current (textual) block any reference definitions and - * remember them so we can resolve any links referring to them. - * - * (Reference definitions can only be at start of it as they cannot break - * a paragraph.) - */ -static int -md_consume_link_reference_definitions(MD_CTX* ctx) -{ - MD_LINE* lines = (MD_LINE*) (ctx->current_block + 1); - int n_lines = ctx->current_block->n_lines; - int n = 0; - - /* Compute how many lines at the start of the block form one or more - * reference definitions. */ - while(n < n_lines) { - int n_link_ref_lines; - - n_link_ref_lines = md_is_link_reference_definition(ctx, - lines + n, n_lines - n); - /* Not a reference definition? */ - if(n_link_ref_lines == 0) - break; - - /* We fail if it is the ref. def. but it could not be stored due - * a memory allocation error. */ - if(n_link_ref_lines < 0) - return -1; - - n += n_link_ref_lines; - } - - /* If there was at least one reference definition, we need to remove - * its lines from the block, or perhaps even the whole block. */ - if(n > 0) { - if(n == n_lines) { - /* Remove complete block. */ - ctx->n_block_bytes -= n * sizeof(MD_LINE); - ctx->n_block_bytes -= sizeof(MD_BLOCK); - ctx->current_block = NULL; - } else { - /* Remove just some initial lines from the block. */ - memmove(lines, lines + n, (n_lines - n) * sizeof(MD_LINE)); - ctx->current_block->n_lines -= n; - ctx->n_block_bytes -= n * sizeof(MD_LINE); - } - } - - return 0; -} - -static int -md_end_current_block(MD_CTX* ctx) -{ - int ret = 0; - - if(ctx->current_block == NULL) - return ret; - - /* Check whether there is a reference definition. (We do this here instead - * of in md_analyze_line() because reference definition can take multiple - * lines.) */ - if(ctx->current_block->type == MD_BLOCK_P || - (ctx->current_block->type == MD_BLOCK_H && (ctx->current_block->flags & MD_BLOCK_SETEXT_HEADER))) - { - MD_LINE* lines = (MD_LINE*) (ctx->current_block + 1); - if(CH(lines[0].beg) == _T('[')) { - MD_CHECK(md_consume_link_reference_definitions(ctx)); - if(ctx->current_block == NULL) - return ret; - } - } - - if(ctx->current_block->type == MD_BLOCK_H && (ctx->current_block->flags & MD_BLOCK_SETEXT_HEADER)) { - int n_lines = ctx->current_block->n_lines; - - if(n_lines > 1) { - /* Get rid of the underline. */ - ctx->current_block->n_lines--; - ctx->n_block_bytes -= sizeof(MD_LINE); - } else { - /* Only the underline has left after eating the ref. defs. - * Keep the line as beginning of a new ordinary paragraph. */ - ctx->current_block->type = MD_BLOCK_P; - return 0; - } - } - - /* Mark we are not building any block anymore. */ - ctx->current_block = NULL; - -abort: - return ret; -} - -static int -md_add_line_into_current_block(MD_CTX* ctx, const MD_LINE_ANALYSIS* analysis) -{ - MD_ASSERT(ctx->current_block != NULL); - - if(ctx->current_block->type == MD_BLOCK_CODE || ctx->current_block->type == MD_BLOCK_HTML) { - MD_VERBATIMLINE* line; - - line = (MD_VERBATIMLINE*) md_push_block_bytes(ctx, sizeof(MD_VERBATIMLINE)); - if(line == NULL) - return -1; - - line->indent = analysis->indent; - line->beg = analysis->beg; - line->end = analysis->end; - } else { - MD_LINE* line; - - line = (MD_LINE*) md_push_block_bytes(ctx, sizeof(MD_LINE)); - if(line == NULL) - return -1; - - line->beg = analysis->beg; - line->end = analysis->end; - } - ctx->current_block->n_lines++; - - return 0; -} - -static int -md_push_container_bytes(MD_CTX* ctx, MD_BLOCKTYPE type, unsigned start, - unsigned data, unsigned flags) -{ - MD_BLOCK* block; - int ret = 0; - - MD_CHECK(md_end_current_block(ctx)); - - block = (MD_BLOCK*) md_push_block_bytes(ctx, sizeof(MD_BLOCK)); - if(block == NULL) - return -1; - - block->type = type; - block->flags = flags; - block->data = data; - block->n_lines = start; - -abort: - return ret; -} - - - -/*********************** - *** Line Analysis *** - ***********************/ - -static int -md_is_hr_line(MD_CTX* ctx, OFF beg, OFF* p_end, OFF* p_killer) -{ - OFF off = beg + 1; - int n = 1; - - while(off < ctx->size && (CH(off) == CH(beg) || CH(off) == _T(' ') || CH(off) == _T('\t'))) { - if(CH(off) == CH(beg)) - n++; - off++; - } - - if(n < 3) { - *p_killer = off; - return FALSE; - } - - /* Nothing else can be present on the line. */ - if(off < ctx->size && !ISNEWLINE(off)) { - *p_killer = off; - return FALSE; - } - - *p_end = off; - return TRUE; -} - -static int -md_is_atxheader_line(MD_CTX* ctx, OFF beg, OFF* p_beg, OFF* p_end, unsigned* p_level) -{ - int n; - OFF off = beg + 1; - - while(off < ctx->size && CH(off) == _T('#') && off - beg < 7) - off++; - n = off - beg; - - if(n > 6) - return FALSE; - *p_level = n; - - if(!(ctx->parser.flags & MD_FLAG_PERMISSIVEATXHEADERS) && off < ctx->size && - CH(off) != _T(' ') && CH(off) != _T('\t') && !ISNEWLINE(off)) - return FALSE; - - while(off < ctx->size && CH(off) == _T(' ')) - off++; - *p_beg = off; - *p_end = off; - return TRUE; -} - -static int -md_is_setext_underline(MD_CTX* ctx, OFF beg, OFF* p_end, unsigned* p_level) -{ - OFF off = beg + 1; - - while(off < ctx->size && CH(off) == CH(beg)) - off++; - - /* Optionally, space(s) can follow. */ - while(off < ctx->size && CH(off) == _T(' ')) - off++; - - /* But nothing more is allowed on the line. */ - if(off < ctx->size && !ISNEWLINE(off)) - return FALSE; - - *p_level = (CH(beg) == _T('=') ? 1 : 2); - *p_end = off; - return TRUE; -} - -static int -md_is_table_underline(MD_CTX* ctx, OFF beg, OFF* p_end, unsigned* p_col_count) -{ - OFF off = beg; - int found_pipe = FALSE; - unsigned col_count = 0; - - if(off < ctx->size && CH(off) == _T('|')) { - found_pipe = TRUE; - off++; - while(off < ctx->size && ISWHITESPACE(off)) - off++; - } - - while(1) { - OFF cell_beg; - int delimited = FALSE; - - /* Cell underline ("-----", ":----", "----:" or ":----:") */ - cell_beg = off; - if(off < ctx->size && CH(off) == _T(':')) - off++; - while(off < ctx->size && CH(off) == _T('-')) - off++; - if(off < ctx->size && CH(off) == _T(':')) - off++; - if(off - cell_beg < 3) - return FALSE; - - col_count++; - - /* Pipe delimiter (optional at the end of line). */ - while(off < ctx->size && ISWHITESPACE(off)) - off++; - if(off < ctx->size && CH(off) == _T('|')) { - delimited = TRUE; - found_pipe = TRUE; - off++; - while(off < ctx->size && ISWHITESPACE(off)) - off++; - } - - /* Success, if we reach end of line. */ - if(off >= ctx->size || ISNEWLINE(off)) - break; - - if(!delimited) - return FALSE; - } - - if(!found_pipe) - return FALSE; - - *p_end = off; - *p_col_count = col_count; - return TRUE; -} - -static int -md_is_opening_code_fence(MD_CTX* ctx, OFF beg, OFF* p_end) -{ - OFF off = beg; - - while(off < ctx->size && CH(off) == CH(beg)) - off++; - - /* Fence must have at least three characters. */ - if(off - beg < 3) - return FALSE; - - ctx->code_fence_length = off - beg; - - /* Optionally, space(s) can follow. */ - while(off < ctx->size && CH(off) == _T(' ')) - off++; - - /* Optionally, an info string can follow. */ - while(off < ctx->size && !ISNEWLINE(off)) { - /* Backtick-based fence must not contain '`' in the info string. */ - if(CH(beg) == _T('`') && CH(off) == _T('`')) - return FALSE; - off++; - } - - *p_end = off; - return TRUE; -} - -static int -md_is_closing_code_fence(MD_CTX* ctx, CHAR ch, OFF beg, OFF* p_end) -{ - OFF off = beg; - int ret = FALSE; - - /* Closing fence must have at least the same length and use same char as - * opening one. */ - while(off < ctx->size && CH(off) == ch) - off++; - if(off - beg < ctx->code_fence_length) - goto out; - - /* Optionally, space(s) can follow */ - while(off < ctx->size && CH(off) == _T(' ')) - off++; - - /* But nothing more is allowed on the line. */ - if(off < ctx->size && !ISNEWLINE(off)) - goto out; - - ret = TRUE; - -out: - /* Note we set *p_end even on failure: If we are not closing fence, caller - * would eat the line anyway without any parsing. */ - *p_end = off; - return ret; -} - -/* Returns type of the raw HTML block, or FALSE if it is not HTML block. - * (Refer to CommonMark specification for details about the types.) - */ -static int -md_is_html_block_start_condition(MD_CTX* ctx, OFF beg) -{ - typedef struct TAG_tag TAG; - struct TAG_tag { - const CHAR* name; - unsigned len : 8; - }; - - /* Type 6 is started by a long list of allowed tags. We use two-level - * tree to speed-up the search. */ -#ifdef X - #undef X -#endif -#define X(name) { _T(name), (sizeof(name)-1) / sizeof(CHAR) } -#define Xend { NULL, 0 } - static const TAG t1[] = { X("pre"), X("script"), X("style"), X("textarea"), Xend }; - - static const TAG a6[] = { X("address"), X("article"), X("aside"), Xend }; - static const TAG b6[] = { X("base"), X("basefont"), X("blockquote"), X("body"), Xend }; - static const TAG c6[] = { X("caption"), X("center"), X("col"), X("colgroup"), Xend }; - static const TAG d6[] = { X("dd"), X("details"), X("dialog"), X("dir"), - X("div"), X("dl"), X("dt"), Xend }; - static const TAG f6[] = { X("fieldset"), X("figcaption"), X("figure"), X("footer"), - X("form"), X("frame"), X("frameset"), Xend }; - static const TAG h6[] = { X("h1"), X("head"), X("header"), X("hr"), X("html"), Xend }; - static const TAG i6[] = { X("iframe"), Xend }; - static const TAG l6[] = { X("legend"), X("li"), X("link"), Xend }; - static const TAG m6[] = { X("main"), X("menu"), X("menuitem"), Xend }; - static const TAG n6[] = { X("nav"), X("noframes"), Xend }; - static const TAG o6[] = { X("ol"), X("optgroup"), X("option"), Xend }; - static const TAG p6[] = { X("p"), X("param"), Xend }; - static const TAG s6[] = { X("section"), X("source"), X("summary"), Xend }; - static const TAG t6[] = { X("table"), X("tbody"), X("td"), X("tfoot"), X("th"), - X("thead"), X("title"), X("tr"), X("track"), Xend }; - static const TAG u6[] = { X("ul"), Xend }; - static const TAG xx[] = { Xend }; -#undef X - - static const TAG* map6[26] = { - a6, b6, c6, d6, xx, f6, xx, h6, i6, xx, xx, l6, m6, - n6, o6, p6, xx, xx, s6, t6, u6, xx, xx, xx, xx, xx - }; - OFF off = beg + 1; - int i; - - /* Check for type 1: <script, <pre, or <style */ - for(i = 0; t1[i].name != NULL; i++) { - if(off + t1[i].len <= ctx->size) { - if(md_ascii_case_eq(STR(off), t1[i].name, t1[i].len)) - return 1; - } - } - - /* Check for type 2: <!-- */ - if(off + 3 < ctx->size && CH(off) == _T('!') && CH(off+1) == _T('-') && CH(off+2) == _T('-')) - return 2; - - /* Check for type 3: <? */ - if(off < ctx->size && CH(off) == _T('?')) - return 3; - - /* Check for type 4 or 5: <! */ - if(off < ctx->size && CH(off) == _T('!')) { - /* Check for type 4: <! followed by uppercase letter. */ - if(off + 1 < ctx->size && ISASCII(off+1)) - return 4; - - /* Check for type 5: <![CDATA[ */ - if(off + 8 < ctx->size) { - if(md_ascii_eq(STR(off), _T("![CDATA["), 8)) - return 5; - } - } - - /* Check for type 6: Many possible starting tags listed above. */ - if(off + 1 < ctx->size && (ISALPHA(off) || (CH(off) == _T('/') && ISALPHA(off+1)))) { - int slot; - const TAG* tags; - - if(CH(off) == _T('/')) - off++; - - slot = (ISUPPER(off) ? CH(off) - 'A' : CH(off) - 'a'); - tags = map6[slot]; - - for(i = 0; tags[i].name != NULL; i++) { - if(off + tags[i].len <= ctx->size) { - if(md_ascii_case_eq(STR(off), tags[i].name, tags[i].len)) { - OFF tmp = off + tags[i].len; - if(tmp >= ctx->size) - return 6; - if(ISBLANK(tmp) || ISNEWLINE(tmp) || CH(tmp) == _T('>')) - return 6; - if(tmp+1 < ctx->size && CH(tmp) == _T('/') && CH(tmp+1) == _T('>')) - return 6; - break; - } - } - } - } - - /* Check for type 7: any COMPLETE other opening or closing tag. */ - if(off + 1 < ctx->size) { - OFF end; - - if(md_is_html_tag(ctx, NULL, 0, beg, ctx->size, &end)) { - /* Only optional whitespace and new line may follow. */ - while(end < ctx->size && ISWHITESPACE(end)) - end++; - if(end >= ctx->size || ISNEWLINE(end)) - return 7; - } - } - - return FALSE; -} - -/* Case sensitive check whether there is a substring 'what' between 'beg' - * and end of line. */ -static int -md_line_contains(MD_CTX* ctx, OFF beg, const CHAR* what, SZ what_len, OFF* p_end) -{ - OFF i; - for(i = beg; i + what_len < ctx->size; i++) { - if(ISNEWLINE(i)) - break; - if(memcmp(STR(i), what, what_len * sizeof(CHAR)) == 0) { - *p_end = i + what_len; - return TRUE; - } - } - - *p_end = i; - return FALSE; -} - -/* Returns type of HTML block end condition or FALSE if not an end condition. - * - * Note it fills p_end even when it is not end condition as the caller - * does not need to analyze contents of a raw HTML block. - */ -static int -md_is_html_block_end_condition(MD_CTX* ctx, OFF beg, OFF* p_end) -{ - switch(ctx->html_block_type) { - case 1: - { - OFF off = beg; - - while(off < ctx->size && !ISNEWLINE(off)) { - if(CH(off) == _T('<')) { - if(md_ascii_case_eq(STR(off), _T("</script>"), 9)) { - *p_end = off + 9; - return TRUE; - } - - if(md_ascii_case_eq(STR(off), _T("</style>"), 8)) { - *p_end = off + 8; - return TRUE; - } - - if(md_ascii_case_eq(STR(off), _T("</pre>"), 6)) { - *p_end = off + 6; - return TRUE; - } - } - - off++; - } - *p_end = off; - return FALSE; - } - - case 2: - return (md_line_contains(ctx, beg, _T("-->"), 3, p_end) ? 2 : FALSE); - - case 3: - return (md_line_contains(ctx, beg, _T("?>"), 2, p_end) ? 3 : FALSE); - - case 4: - return (md_line_contains(ctx, beg, _T(">"), 1, p_end) ? 4 : FALSE); - - case 5: - return (md_line_contains(ctx, beg, _T("]]>"), 3, p_end) ? 5 : FALSE); - - case 6: /* Pass through */ - case 7: - *p_end = beg; - return (ISNEWLINE(beg) ? ctx->html_block_type : FALSE); - - default: - MD_UNREACHABLE(); - } - return FALSE; -} - - -static int -md_is_container_compatible(const MD_CONTAINER* pivot, const MD_CONTAINER* container) -{ - /* Block quote has no "items" like lists. */ - if(container->ch == _T('>')) - return FALSE; - - if(container->ch != pivot->ch) - return FALSE; - if(container->mark_indent > pivot->contents_indent) - return FALSE; - - return TRUE; -} - -static int -md_push_container(MD_CTX* ctx, const MD_CONTAINER* container) -{ - if(ctx->n_containers >= ctx->alloc_containers) { - MD_CONTAINER* new_containers; - - ctx->alloc_containers = (ctx->alloc_containers > 0 - ? ctx->alloc_containers + ctx->alloc_containers / 2 - : 16); - new_containers = realloc(ctx->containers, ctx->alloc_containers * sizeof(MD_CONTAINER)); - if(new_containers == NULL) { - MD_LOG("realloc() failed."); - return -1; - } - - ctx->containers = new_containers; - } - - memcpy(&ctx->containers[ctx->n_containers++], container, sizeof(MD_CONTAINER)); - return 0; -} - -static int -md_enter_child_containers(MD_CTX* ctx, int n_children) -{ - int i; - int ret = 0; - - for(i = ctx->n_containers - n_children; i < ctx->n_containers; i++) { - MD_CONTAINER* c = &ctx->containers[i]; - int is_ordered_list = FALSE; - - switch(c->ch) { - case _T(')'): - case _T('.'): - is_ordered_list = TRUE; - MD_FALLTHROUGH(); - - case _T('-'): - case _T('+'): - case _T('*'): - /* Remember offset in ctx->block_bytes so we can revisit the - * block if we detect it is a loose list. */ - md_end_current_block(ctx); - c->block_byte_off = ctx->n_block_bytes; - - MD_CHECK(md_push_container_bytes(ctx, - (is_ordered_list ? MD_BLOCK_OL : MD_BLOCK_UL), - c->start, c->ch, MD_BLOCK_CONTAINER_OPENER)); - MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI, - c->task_mark_off, - (c->is_task ? CH(c->task_mark_off) : 0), - MD_BLOCK_CONTAINER_OPENER)); - break; - - case _T('>'): - MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_QUOTE, 0, 0, MD_BLOCK_CONTAINER_OPENER)); - break; - - default: - MD_UNREACHABLE(); - break; - } - } - -abort: - return ret; -} - -static int -md_leave_child_containers(MD_CTX* ctx, int n_keep) -{ - int ret = 0; - - while(ctx->n_containers > n_keep) { - MD_CONTAINER* c = &ctx->containers[ctx->n_containers-1]; - int is_ordered_list = FALSE; - - switch(c->ch) { - case _T(')'): - case _T('.'): - is_ordered_list = TRUE; - MD_FALLTHROUGH(); - - case _T('-'): - case _T('+'): - case _T('*'): - MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI, - c->task_mark_off, (c->is_task ? CH(c->task_mark_off) : 0), - MD_BLOCK_CONTAINER_CLOSER)); - MD_CHECK(md_push_container_bytes(ctx, - (is_ordered_list ? MD_BLOCK_OL : MD_BLOCK_UL), 0, - c->ch, MD_BLOCK_CONTAINER_CLOSER)); - break; - - case _T('>'): - MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_QUOTE, 0, - 0, MD_BLOCK_CONTAINER_CLOSER)); - break; - - default: - MD_UNREACHABLE(); - break; - } - - ctx->n_containers--; - } - -abort: - return ret; -} - -static int -md_is_container_mark(MD_CTX* ctx, unsigned indent, OFF beg, OFF* p_end, MD_CONTAINER* p_container) -{ - OFF off = beg; - OFF max_end; - - if(off >= ctx->size || indent >= ctx->code_indent_offset) - return FALSE; - - /* Check for block quote mark. */ - if(CH(off) == _T('>')) { - off++; - p_container->ch = _T('>'); - p_container->is_loose = FALSE; - p_container->is_task = FALSE; - p_container->mark_indent = indent; - p_container->contents_indent = indent + 1; - *p_end = off; - return TRUE; - } - - /* Check for list item bullet mark. */ - if(ISANYOF(off, _T("-+*")) && (off+1 >= ctx->size || ISBLANK(off+1) || ISNEWLINE(off+1))) { - p_container->ch = CH(off); - p_container->is_loose = FALSE; - p_container->is_task = FALSE; - p_container->mark_indent = indent; - p_container->contents_indent = indent + 1; - *p_end = off+1; - return TRUE; - } - - /* Check for ordered list item marks. */ - max_end = off + 9; - if(max_end > ctx->size) - max_end = ctx->size; - p_container->start = 0; - while(off < max_end && ISDIGIT(off)) { - p_container->start = p_container->start * 10 + CH(off) - _T('0'); - off++; - } - if(off > beg && - (CH(off) == _T('.') || CH(off) == _T(')')) && - (off+1 >= ctx->size || ISBLANK(off+1) || ISNEWLINE(off+1))) - { - p_container->ch = CH(off); - p_container->is_loose = FALSE; - p_container->is_task = FALSE; - p_container->mark_indent = indent; - p_container->contents_indent = indent + off - beg + 1; - *p_end = off+1; - return TRUE; - } - - return FALSE; -} - -static unsigned -md_line_indentation(MD_CTX* ctx, unsigned total_indent, OFF beg, OFF* p_end) -{ - OFF off = beg; - unsigned indent = total_indent; - - while(off < ctx->size && ISBLANK(off)) { - if(CH(off) == _T('\t')) - indent = (indent + 4) & ~3; - else - indent++; - off++; - } - - *p_end = off; - return indent - total_indent; -} - -static const MD_LINE_ANALYSIS md_dummy_blank_line = { MD_LINE_BLANK, 0, 0, 0, 0 }; - -/* Analyze type of the line and find some its properties. This serves as a - * main input for determining type and boundaries of a block. */ -static int -md_analyze_line(MD_CTX* ctx, OFF beg, OFF* p_end, - const MD_LINE_ANALYSIS* pivot_line, MD_LINE_ANALYSIS* line) -{ - unsigned total_indent = 0; - int n_parents = 0; - int n_brothers = 0; - int n_children = 0; - MD_CONTAINER container = { 0 }; - int prev_line_has_list_loosening_effect = ctx->last_line_has_list_loosening_effect; - OFF off = beg; - OFF hr_killer = 0; - int ret = 0; - - line->indent = md_line_indentation(ctx, total_indent, off, &off); - total_indent += line->indent; - line->beg = off; - - /* Given the indentation and block quote marks '>', determine how many of - * the current containers are our parents. */ - while(n_parents < ctx->n_containers) { - MD_CONTAINER* c = &ctx->containers[n_parents]; - - if(c->ch == _T('>') && line->indent < ctx->code_indent_offset && - off < ctx->size && CH(off) == _T('>')) - { - /* Block quote mark. */ - off++; - total_indent++; - line->indent = md_line_indentation(ctx, total_indent, off, &off); - total_indent += line->indent; - - /* The optional 1st space after '>' is part of the block quote mark. */ - if(line->indent > 0) - line->indent--; - - line->beg = off; - - } else if(c->ch != _T('>') && line->indent >= c->contents_indent) { - /* List. */ - line->indent -= c->contents_indent; - } else { - break; - } - - n_parents++; - } - - if(off >= ctx->size || ISNEWLINE(off)) { - /* Blank line does not need any real indentation to be nested inside - * a list. */ - if(n_brothers + n_children == 0) { - while(n_parents < ctx->n_containers && ctx->containers[n_parents].ch != _T('>')) - n_parents++; - } - } - - while(TRUE) { - /* Check whether we are fenced code continuation. */ - if(pivot_line->type == MD_LINE_FENCEDCODE) { - line->beg = off; - - /* We are another MD_LINE_FENCEDCODE unless we are closing fence - * which we transform into MD_LINE_BLANK. */ - if(line->indent < ctx->code_indent_offset) { - if(md_is_closing_code_fence(ctx, CH(pivot_line->beg), off, &off)) { - line->type = MD_LINE_BLANK; - ctx->last_line_has_list_loosening_effect = FALSE; - break; - } - } - - /* Change indentation accordingly to the initial code fence. */ - if(n_parents == ctx->n_containers) { - if(line->indent > pivot_line->indent) - line->indent -= pivot_line->indent; - else - line->indent = 0; - - line->type = MD_LINE_FENCEDCODE; - break; - } - } - - /* Check whether we are HTML block continuation. */ - if(pivot_line->type == MD_LINE_HTML && ctx->html_block_type > 0) { - if(n_parents < ctx->n_containers) { - /* HTML block is implicitly ended if the enclosing container - * block ends. */ - ctx->html_block_type = 0; - } else { - int html_block_type; - - html_block_type = md_is_html_block_end_condition(ctx, off, &off); - if(html_block_type > 0) { - MD_ASSERT(html_block_type == ctx->html_block_type); - - /* Make sure this is the last line of the block. */ - ctx->html_block_type = 0; - - /* Some end conditions serve as blank lines at the same time. */ - if(html_block_type == 6 || html_block_type == 7) { - line->type = MD_LINE_BLANK; - line->indent = 0; - break; - } - } - - line->type = MD_LINE_HTML; - n_parents = ctx->n_containers; - break; - } - } - - /* Check for blank line. */ - if(off >= ctx->size || ISNEWLINE(off)) { - if(pivot_line->type == MD_LINE_INDENTEDCODE && n_parents == ctx->n_containers) { - line->type = MD_LINE_INDENTEDCODE; - if(line->indent > ctx->code_indent_offset) - line->indent -= ctx->code_indent_offset; - else - line->indent = 0; - ctx->last_line_has_list_loosening_effect = FALSE; - } else { - line->type = MD_LINE_BLANK; - ctx->last_line_has_list_loosening_effect = (n_parents > 0 && - n_brothers + n_children == 0 && - ctx->containers[n_parents-1].ch != _T('>')); - - #if 1 - /* See https://github.com/mity/md4c/issues/6 - * - * This ugly checking tests we are in (yet empty) list item but - * not its very first line (i.e. not the line with the list - * item mark). - * - * If we are such a blank line, then any following non-blank - * line which would be part of the list item actually has to - * end the list because according to the specification, "a list - * item can begin with at most one blank line." - */ - if(n_parents > 0 && ctx->containers[n_parents-1].ch != _T('>') && - n_brothers + n_children == 0 && ctx->current_block == NULL && - ctx->n_block_bytes > (int) sizeof(MD_BLOCK)) - { - MD_BLOCK* top_block = (MD_BLOCK*) ((char*)ctx->block_bytes + ctx->n_block_bytes - sizeof(MD_BLOCK)); - if(top_block->type == MD_BLOCK_LI) - ctx->last_list_item_starts_with_two_blank_lines = TRUE; - } - #endif - } - break; - } else { - #if 1 - /* This is the 2nd half of the hack. If the flag is set (i.e. there - * was a 2nd blank line at the beginning of the list item) and if - * we would otherwise still belong to the list item, we enforce - * the end of the list. */ - ctx->last_line_has_list_loosening_effect = FALSE; - if(ctx->last_list_item_starts_with_two_blank_lines) { - if(n_parents > 0 && ctx->containers[n_parents-1].ch != _T('>') && - n_brothers + n_children == 0 && ctx->current_block == NULL && - ctx->n_block_bytes > (int) sizeof(MD_BLOCK)) - { - MD_BLOCK* top_block = (MD_BLOCK*) ((char*)ctx->block_bytes + ctx->n_block_bytes - sizeof(MD_BLOCK)); - if(top_block->type == MD_BLOCK_LI) - n_parents--; - } - - ctx->last_list_item_starts_with_two_blank_lines = FALSE; - } - #endif - } - - /* Check whether we are Setext underline. */ - if(line->indent < ctx->code_indent_offset && pivot_line->type == MD_LINE_TEXT - && off < ctx->size && ISANYOF2(off, _T('='), _T('-')) - && (n_parents == ctx->n_containers)) - { - unsigned level; - - if(md_is_setext_underline(ctx, off, &off, &level)) { - line->type = MD_LINE_SETEXTUNDERLINE; - line->data = level; - break; - } - } - - /* Check for thematic break line. */ - if(line->indent < ctx->code_indent_offset - && off < ctx->size && off >= hr_killer - && ISANYOF(off, _T("-_*"))) - { - if(md_is_hr_line(ctx, off, &off, &hr_killer)) { - line->type = MD_LINE_HR; - break; - } - } - - /* Check for "brother" container. I.e. whether we are another list item - * in already started list. */ - if(n_parents < ctx->n_containers && n_brothers + n_children == 0) { - OFF tmp; - - if(md_is_container_mark(ctx, line->indent, off, &tmp, &container) && - md_is_container_compatible(&ctx->containers[n_parents], &container)) - { - pivot_line = &md_dummy_blank_line; - - off = tmp; - - total_indent += container.contents_indent - container.mark_indent; - line->indent = md_line_indentation(ctx, total_indent, off, &off); - total_indent += line->indent; - line->beg = off; - - /* Some of the following whitespace actually still belongs to the mark. */ - if(off >= ctx->size || ISNEWLINE(off)) { - container.contents_indent++; - } else if(line->indent <= ctx->code_indent_offset) { - container.contents_indent += line->indent; - line->indent = 0; - } else { - container.contents_indent += 1; - line->indent--; - } - - ctx->containers[n_parents].mark_indent = container.mark_indent; - ctx->containers[n_parents].contents_indent = container.contents_indent; - - n_brothers++; - continue; - } - } - - /* Check for indented code. - * Note indented code block cannot interrupt a paragraph. */ - if(line->indent >= ctx->code_indent_offset && - (pivot_line->type == MD_LINE_BLANK || pivot_line->type == MD_LINE_INDENTEDCODE)) - { - line->type = MD_LINE_INDENTEDCODE; - MD_ASSERT(line->indent >= ctx->code_indent_offset); - line->indent -= ctx->code_indent_offset; - line->data = 0; - break; - } - - /* Check for start of a new container block. */ - if(line->indent < ctx->code_indent_offset && - md_is_container_mark(ctx, line->indent, off, &off, &container)) - { - if(pivot_line->type == MD_LINE_TEXT && n_parents == ctx->n_containers && - (off >= ctx->size || ISNEWLINE(off)) && container.ch != _T('>')) - { - /* Noop. List mark followed by a blank line cannot interrupt a paragraph. */ - } else if(pivot_line->type == MD_LINE_TEXT && n_parents == ctx->n_containers && - ISANYOF2_(container.ch, _T('.'), _T(')')) && container.start != 1) - { - /* Noop. Ordered list cannot interrupt a paragraph unless the start index is 1. */ - } else { - total_indent += container.contents_indent - container.mark_indent; - line->indent = md_line_indentation(ctx, total_indent, off, &off); - total_indent += line->indent; - - line->beg = off; - line->data = container.ch; - - /* Some of the following whitespace actually still belongs to the mark. */ - if(off >= ctx->size || ISNEWLINE(off)) { - container.contents_indent++; - } else if(line->indent <= ctx->code_indent_offset) { - container.contents_indent += line->indent; - line->indent = 0; - } else { - container.contents_indent += 1; - line->indent--; - } - - if(n_brothers + n_children == 0) - pivot_line = &md_dummy_blank_line; - - if(n_children == 0) - MD_CHECK(md_leave_child_containers(ctx, n_parents + n_brothers)); - - n_children++; - MD_CHECK(md_push_container(ctx, &container)); - continue; - } - } - - /* Check whether we are table continuation. */ - if(pivot_line->type == MD_LINE_TABLE && n_parents == ctx->n_containers) { - line->type = MD_LINE_TABLE; - break; - } - - /* Check for ATX header. */ - if(line->indent < ctx->code_indent_offset && - off < ctx->size && CH(off) == _T('#')) - { - unsigned level; - - if(md_is_atxheader_line(ctx, off, &line->beg, &off, &level)) { - line->type = MD_LINE_ATXHEADER; - line->data = level; - break; - } - } - - /* Check whether we are starting code fence. */ - if(off < ctx->size && ISANYOF2(off, _T('`'), _T('~'))) { - if(md_is_opening_code_fence(ctx, off, &off)) { - line->type = MD_LINE_FENCEDCODE; - line->data = 1; - break; - } - } - - /* Check for start of raw HTML block. */ - if(off < ctx->size && CH(off) == _T('<') - && !(ctx->parser.flags & MD_FLAG_NOHTMLBLOCKS)) - { - ctx->html_block_type = md_is_html_block_start_condition(ctx, off); - - /* HTML block type 7 cannot interrupt paragraph. */ - if(ctx->html_block_type == 7 && pivot_line->type == MD_LINE_TEXT) - ctx->html_block_type = 0; - - if(ctx->html_block_type > 0) { - /* The line itself also may immediately close the block. */ - if(md_is_html_block_end_condition(ctx, off, &off) == ctx->html_block_type) { - /* Make sure this is the last line of the block. */ - ctx->html_block_type = 0; - } - - line->type = MD_LINE_HTML; - break; - } - } - - /* Check for table underline. */ - if((ctx->parser.flags & MD_FLAG_TABLES) && pivot_line->type == MD_LINE_TEXT - && off < ctx->size && ISANYOF3(off, _T('|'), _T('-'), _T(':')) - && n_parents == ctx->n_containers) - { - unsigned col_count; - - if(ctx->current_block != NULL && ctx->current_block->n_lines == 1 && - md_is_table_underline(ctx, off, &off, &col_count)) - { - line->data = col_count; - line->type = MD_LINE_TABLEUNDERLINE; - break; - } - } - - /* By default, we are normal text line. */ - line->type = MD_LINE_TEXT; - if(pivot_line->type == MD_LINE_TEXT && n_brothers + n_children == 0) { - /* Lazy continuation. */ - n_parents = ctx->n_containers; - } - - /* Check for task mark. */ - if((ctx->parser.flags & MD_FLAG_TASKLISTS) && n_brothers + n_children > 0 && - ISANYOF_(ctx->containers[ctx->n_containers-1].ch, _T("-+*.)"))) - { - OFF tmp = off; - - while(tmp < ctx->size && tmp < off + 3 && ISBLANK(tmp)) - tmp++; - if(tmp + 2 < ctx->size && CH(tmp) == _T('[') && - ISANYOF(tmp+1, _T("xX ")) && CH(tmp+2) == _T(']') && - (tmp + 3 == ctx->size || ISBLANK(tmp+3) || ISNEWLINE(tmp+3))) - { - MD_CONTAINER* task_container = (n_children > 0 ? &ctx->containers[ctx->n_containers-1] : &container); - task_container->is_task = TRUE; - task_container->task_mark_off = tmp + 1; - off = tmp + 3; - while(ISWHITESPACE(off)) - off++; - line->beg = off; - } - } - - break; - } - - /* Scan for end of the line. - * - * Note this is quite a bottleneck of the parsing as we here iterate almost - * over compete document. - */ -#if defined __linux__ && !defined MD4C_USE_UTF16 - /* Recent glibc versions have superbly optimized strcspn(), even using - * vectorization if available. */ - if(ctx->doc_ends_with_newline && off < ctx->size) { - while(TRUE) { - off += (OFF) strcspn(STR(off), "\r\n"); - - /* strcspn() can stop on zero terminator; but that can appear - * anywhere in the Markfown input... */ - if(CH(off) == _T('\0')) - off++; - else - break; - } - } else -#endif - { - /* Optimization: Use some loop unrolling. */ - while(off + 3 < ctx->size && !ISNEWLINE(off+0) && !ISNEWLINE(off+1) - && !ISNEWLINE(off+2) && !ISNEWLINE(off+3)) - off += 4; - while(off < ctx->size && !ISNEWLINE(off)) - off++; - } - - /* Set end of the line. */ - line->end = off; - - /* But for ATX header, we should exclude the optional trailing mark. */ - if(line->type == MD_LINE_ATXHEADER) { - OFF tmp = line->end; - while(tmp > line->beg && CH(tmp-1) == _T(' ')) - tmp--; - while(tmp > line->beg && CH(tmp-1) == _T('#')) - tmp--; - if(tmp == line->beg || CH(tmp-1) == _T(' ') || (ctx->parser.flags & MD_FLAG_PERMISSIVEATXHEADERS)) - line->end = tmp; - } - - /* Trim trailing spaces. */ - if(line->type != MD_LINE_INDENTEDCODE && line->type != MD_LINE_FENCEDCODE) { - while(line->end > line->beg && CH(line->end-1) == _T(' ')) - line->end--; - } - - /* Eat also the new line. */ - if(off < ctx->size && CH(off) == _T('\r')) - off++; - if(off < ctx->size && CH(off) == _T('\n')) - off++; - - *p_end = off; - - /* If we belong to a list after seeing a blank line, the list is loose. */ - if(prev_line_has_list_loosening_effect && line->type != MD_LINE_BLANK && n_parents + n_brothers > 0) { - MD_CONTAINER* c = &ctx->containers[n_parents + n_brothers - 1]; - if(c->ch != _T('>')) { - MD_BLOCK* block = (MD_BLOCK*) (((char*)ctx->block_bytes) + c->block_byte_off); - block->flags |= MD_BLOCK_LOOSE_LIST; - } - } - - /* Leave any containers we are not part of anymore. */ - if(n_children == 0 && n_parents + n_brothers < ctx->n_containers) - MD_CHECK(md_leave_child_containers(ctx, n_parents + n_brothers)); - - /* Enter any container we found a mark for. */ - if(n_brothers > 0) { - MD_ASSERT(n_brothers == 1); - MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI, - ctx->containers[n_parents].task_mark_off, - (ctx->containers[n_parents].is_task ? CH(ctx->containers[n_parents].task_mark_off) : 0), - MD_BLOCK_CONTAINER_CLOSER)); - MD_CHECK(md_push_container_bytes(ctx, MD_BLOCK_LI, - container.task_mark_off, - (container.is_task ? CH(container.task_mark_off) : 0), - MD_BLOCK_CONTAINER_OPENER)); - ctx->containers[n_parents].is_task = container.is_task; - ctx->containers[n_parents].task_mark_off = container.task_mark_off; - } - - if(n_children > 0) - MD_CHECK(md_enter_child_containers(ctx, n_children)); - -abort: - return ret; -} - -static int -md_process_line(MD_CTX* ctx, const MD_LINE_ANALYSIS** p_pivot_line, MD_LINE_ANALYSIS* line) -{ - const MD_LINE_ANALYSIS* pivot_line = *p_pivot_line; - int ret = 0; - - /* Blank line ends current leaf block. */ - if(line->type == MD_LINE_BLANK) { - MD_CHECK(md_end_current_block(ctx)); - *p_pivot_line = &md_dummy_blank_line; - return 0; - } - - /* Some line types form block on their own. */ - if(line->type == MD_LINE_HR || line->type == MD_LINE_ATXHEADER) { - MD_CHECK(md_end_current_block(ctx)); - - /* Add our single-line block. */ - MD_CHECK(md_start_new_block(ctx, line)); - MD_CHECK(md_add_line_into_current_block(ctx, line)); - MD_CHECK(md_end_current_block(ctx)); - *p_pivot_line = &md_dummy_blank_line; - return 0; - } - - /* MD_LINE_SETEXTUNDERLINE changes meaning of the current block and ends it. */ - if(line->type == MD_LINE_SETEXTUNDERLINE) { - MD_ASSERT(ctx->current_block != NULL); - ctx->current_block->type = MD_BLOCK_H; - ctx->current_block->data = line->data; - ctx->current_block->flags |= MD_BLOCK_SETEXT_HEADER; - MD_CHECK(md_add_line_into_current_block(ctx, line)); - MD_CHECK(md_end_current_block(ctx)); - if(ctx->current_block == NULL) { - *p_pivot_line = &md_dummy_blank_line; - } else { - /* This happens if we have consumed all the body as link ref. defs. - * and downgraded the underline into start of a new paragraph block. */ - line->type = MD_LINE_TEXT; - *p_pivot_line = line; - } - return 0; - } - - /* MD_LINE_TABLEUNDERLINE changes meaning of the current block. */ - if(line->type == MD_LINE_TABLEUNDERLINE) { - MD_ASSERT(ctx->current_block != NULL); - MD_ASSERT(ctx->current_block->n_lines == 1); - ctx->current_block->type = MD_BLOCK_TABLE; - ctx->current_block->data = line->data; - MD_ASSERT(pivot_line != &md_dummy_blank_line); - ((MD_LINE_ANALYSIS*)pivot_line)->type = MD_LINE_TABLE; - MD_CHECK(md_add_line_into_current_block(ctx, line)); - return 0; - } - - /* The current block also ends if the line has different type. */ - if(line->type != pivot_line->type) - MD_CHECK(md_end_current_block(ctx)); - - /* The current line may start a new block. */ - if(ctx->current_block == NULL) { - MD_CHECK(md_start_new_block(ctx, line)); - *p_pivot_line = line; - } - - /* In all other cases the line is just a continuation of the current block. */ - MD_CHECK(md_add_line_into_current_block(ctx, line)); - -abort: - return ret; -} - -static int -md_process_doc(MD_CTX *ctx) -{ - const MD_LINE_ANALYSIS* pivot_line = &md_dummy_blank_line; - MD_LINE_ANALYSIS line_buf[2]; - MD_LINE_ANALYSIS* line = &line_buf[0]; - OFF off = 0; - int ret = 0; - - MD_ENTER_BLOCK(MD_BLOCK_DOC, NULL); - - while(off < ctx->size) { - if(line == pivot_line) - line = (line == &line_buf[0] ? &line_buf[1] : &line_buf[0]); - - MD_CHECK(md_analyze_line(ctx, off, &off, pivot_line, line)); - MD_CHECK(md_process_line(ctx, &pivot_line, line)); - } - - md_end_current_block(ctx); - - MD_CHECK(md_build_ref_def_hashtable(ctx)); - - /* Process all blocks. */ - MD_CHECK(md_leave_child_containers(ctx, 0)); - MD_CHECK(md_process_all_blocks(ctx)); - - MD_LEAVE_BLOCK(MD_BLOCK_DOC, NULL); - -abort: - -#if 0 - /* Output some memory consumption statistics. */ - { - char buffer[256]; - sprintf(buffer, "Alloced %u bytes for block buffer.", - (unsigned)(ctx->alloc_block_bytes)); - MD_LOG(buffer); - - sprintf(buffer, "Alloced %u bytes for containers buffer.", - (unsigned)(ctx->alloc_containers * sizeof(MD_CONTAINER))); - MD_LOG(buffer); - - sprintf(buffer, "Alloced %u bytes for marks buffer.", - (unsigned)(ctx->alloc_marks * sizeof(MD_MARK))); - MD_LOG(buffer); - - sprintf(buffer, "Alloced %u bytes for aux. buffer.", - (unsigned)(ctx->alloc_buffer * sizeof(MD_CHAR))); - MD_LOG(buffer); - } -#endif - - return ret; -} - - -/******************** - *** Public API *** - ********************/ - -int -md_parse(const MD_CHAR* text, MD_SIZE size, const MD_PARSER* parser, void* userdata) -{ - MD_CTX ctx; - int i; - int ret; - - if(parser->abi_version != 0) { - if(parser->debug_log != NULL) - parser->debug_log("Unsupported abi_version.", userdata); - return -1; - } - - /* Setup context structure. */ - memset(&ctx, 0, sizeof(MD_CTX)); - ctx.text = text; - ctx.size = size; - memcpy(&ctx.parser, parser, sizeof(MD_PARSER)); - ctx.userdata = userdata; - ctx.code_indent_offset = (ctx.parser.flags & MD_FLAG_NOINDENTEDCODEBLOCKS) ? (OFF)(-1) : 4; - md_build_mark_char_map(&ctx); - ctx.doc_ends_with_newline = (size > 0 && ISNEWLINE_(text[size-1])); - - /* Reset all unresolved opener mark chains. */ - for(i = 0; i < (int) SIZEOF_ARRAY(ctx.mark_chains); i++) { - ctx.mark_chains[i].head = -1; - ctx.mark_chains[i].tail = -1; - } - ctx.unresolved_link_head = -1; - ctx.unresolved_link_tail = -1; - - /* All the work. */ - ret = md_process_doc(&ctx); - - /* Clean-up. */ - md_free_ref_defs(&ctx); - md_free_ref_def_hashtable(&ctx); - free(ctx.buffer); - free(ctx.marks); - free(ctx.block_bytes); - free(ctx.containers); - - return ret; -} diff --git a/src/md4c.h b/src/md4c.h deleted file mode 100644 index 95f78f9..0000000 --- a/src/md4c.h +++ /dev/null @@ -1,405 +0,0 @@ -/* - * MD4C: Markdown parser for C - * (http://github.com/mity/md4c) - * - * Copyright (c) 2016-2020 Martin Mitas - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS - * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ - -#ifndef MD4C_H -#define MD4C_H - -#ifdef __cplusplus - extern "C" { -#endif - -#if defined MD4C_USE_UTF16 - /* Magic to support UTF-16. Note that in order to use it, you have to define - * the macro MD4C_USE_UTF16 both when building MD4C as well as when - * including this header in your code. */ - #ifdef _WIN32 - #include <windows.h> - typedef WCHAR MD_CHAR; - #else - #error MD4C_USE_UTF16 is only supported on Windows. - #endif -#else - typedef char MD_CHAR; -#endif - -typedef unsigned MD_SIZE; -typedef unsigned MD_OFFSET; - - -/* Block represents a part of document hierarchy structure like a paragraph - * or list item. - */ -typedef enum MD_BLOCKTYPE { - /* <body>...</body> */ - MD_BLOCK_DOC = 0, - - /* <blockquote>...</blockquote> */ - MD_BLOCK_QUOTE, - - /* <ul>...</ul> - * Detail: Structure MD_BLOCK_UL_DETAIL. */ - MD_BLOCK_UL, - - /* <ol>...</ol> - * Detail: Structure MD_BLOCK_OL_DETAIL. */ - MD_BLOCK_OL, - - /* <li>...</li> - * Detail: Structure MD_BLOCK_LI_DETAIL. */ - MD_BLOCK_LI, - - /* <hr> */ - MD_BLOCK_HR, - - /* <h1>...</h1> (for levels up to 6) - * Detail: Structure MD_BLOCK_H_DETAIL. */ - MD_BLOCK_H, - - /* <pre><code>...</code></pre> - * Note the text lines within code blocks are terminated with '\n' - * instead of explicit MD_TEXT_BR. */ - MD_BLOCK_CODE, - - /* Raw HTML block. This itself does not correspond to any particular HTML - * tag. The contents of it _is_ raw HTML source intended to be put - * in verbatim form to the HTML output. */ - MD_BLOCK_HTML, - - /* <p>...</p> */ - MD_BLOCK_P, - - /* <table>...</table> and its contents. - * Detail: Structure MD_BLOCK_TABLE_DETAIL (for MD_BLOCK_TABLE), - * structure MD_BLOCK_TD_DETAIL (for MD_BLOCK_TH and MD_BLOCK_TD) - * Note all of these are used only if extension MD_FLAG_TABLES is enabled. */ - MD_BLOCK_TABLE, - MD_BLOCK_THEAD, - MD_BLOCK_TBODY, - MD_BLOCK_TR, - MD_BLOCK_TH, - MD_BLOCK_TD -} MD_BLOCKTYPE; - -/* Span represents an in-line piece of a document which should be rendered with - * the same font, color and other attributes. A sequence of spans forms a block - * like paragraph or list item. */ -typedef enum MD_SPANTYPE { - /* <em>...</em> */ - MD_SPAN_EM, - - /* <strong>...</strong> */ - MD_SPAN_STRONG, - - /* <a href="xxx">...</a> - * Detail: Structure MD_SPAN_A_DETAIL. */ - MD_SPAN_A, - - /* <img src="xxx">...</a> - * Detail: Structure MD_SPAN_IMG_DETAIL. - * Note: Image text can contain nested spans and even nested images. - * If rendered into ALT attribute of HTML <IMG> tag, it's responsibility - * of the parser to deal with it. - */ - MD_SPAN_IMG, - - /* <code>...</code> */ - MD_SPAN_CODE, - - /* <del>...</del> - * Note: Recognized only when MD_FLAG_STRIKETHROUGH is enabled. - */ - MD_SPAN_DEL, - - /* For recognizing inline ($) and display ($$) equations - * Note: Recognized only when MD_FLAG_LATEXMATHSPANS is enabled. - */ - MD_SPAN_LATEXMATH, - MD_SPAN_LATEXMATH_DISPLAY, - - /* Wiki links - * Note: Recognized only when MD_FLAG_WIKILINKS is enabled. - */ - MD_SPAN_WIKILINK, - - /* <u>...</u> - * Note: Recognized only when MD_FLAG_UNDERLINE is enabled. */ - MD_SPAN_U -} MD_SPANTYPE; - -/* Text is the actual textual contents of span. */ -typedef enum MD_TEXTTYPE { - /* Normal text. */ - MD_TEXT_NORMAL = 0, - - /* NULL character. CommonMark requires replacing NULL character with - * the replacement char U+FFFD, so this allows caller to do that easily. */ - MD_TEXT_NULLCHAR, - - /* Line breaks. - * Note these are not sent from blocks with verbatim output (MD_BLOCK_CODE - * or MD_BLOCK_HTML). In such cases, '\n' is part of the text itself. */ - MD_TEXT_BR, /* <br> (hard break) */ - MD_TEXT_SOFTBR, /* '\n' in source text where it is not semantically meaningful (soft break) */ - - /* Entity. - * (a) Named entity, e.g. - * (Note MD4C does not have a list of known entities. - * Anything matching the regexp /&[A-Za-z][A-Za-z0-9]{1,47};/ is - * treated as a named entity.) - * (b) Numerical entity, e.g. Ӓ - * (c) Hexadecimal entity, e.g. ካ - * - * As MD4C is mostly encoding agnostic, application gets the verbatim - * entity text into the MD_PARSER::text_callback(). */ - MD_TEXT_ENTITY, - - /* Text in a code block (inside MD_BLOCK_CODE) or inlined code (`code`). - * If it is inside MD_BLOCK_CODE, it includes spaces for indentation and - * '\n' for new lines. MD_TEXT_BR and MD_TEXT_SOFTBR are not sent for this - * kind of text. */ - MD_TEXT_CODE, - - /* Text is a raw HTML. If it is contents of a raw HTML block (i.e. not - * an inline raw HTML), then MD_TEXT_BR and MD_TEXT_SOFTBR are not used. - * The text contains verbatim '\n' for the new lines. */ - MD_TEXT_HTML, - - /* Text is inside an equation. This is processed the same way as inlined code - * spans (`code`). */ - MD_TEXT_LATEXMATH -} MD_TEXTTYPE; - - -/* Alignment enumeration. */ -typedef enum MD_ALIGN { - MD_ALIGN_DEFAULT = 0, /* When unspecified. */ - MD_ALIGN_LEFT, - MD_ALIGN_CENTER, - MD_ALIGN_RIGHT -} MD_ALIGN; - - -/* String attribute. - * - * This wraps strings which are outside of a normal text flow and which are - * propagated within various detailed structures, but which still may contain - * string portions of different types like e.g. entities. - * - * So, for example, lets consider this image: - * - * ![image alt text](http://example.org/image.png 'foo " bar') - * - * The image alt text is propagated as a normal text via the MD_PARSER::text() - * callback. However, the image title ('foo " bar') is propagated as - * MD_ATTRIBUTE in MD_SPAN_IMG_DETAIL::title. - * - * Then the attribute MD_SPAN_IMG_DETAIL::title shall provide the following: - * -- [0]: "foo " (substr_types[0] == MD_TEXT_NORMAL; substr_offsets[0] == 0) - * -- [1]: """ (substr_types[1] == MD_TEXT_ENTITY; substr_offsets[1] == 4) - * -- [2]: " bar" (substr_types[2] == MD_TEXT_NORMAL; substr_offsets[2] == 10) - * -- [3]: (n/a) (n/a ; substr_offsets[3] == 14) - * - * Note that these invariants are always guaranteed: - * -- substr_offsets[0] == 0 - * -- substr_offsets[LAST+1] == size - * -- Currently, only MD_TEXT_NORMAL, MD_TEXT_ENTITY, MD_TEXT_NULLCHAR - * substrings can appear. This could change only of the specification - * changes. - */ -typedef struct MD_ATTRIBUTE { - const MD_CHAR* text; - MD_SIZE size; - const MD_TEXTTYPE* substr_types; - const MD_OFFSET* substr_offsets; -} MD_ATTRIBUTE; - - -/* Detailed info for MD_BLOCK_UL. */ -typedef struct MD_BLOCK_UL_DETAIL { - int is_tight; /* Non-zero if tight list, zero if loose. */ - MD_CHAR mark; /* Item bullet character in MarkDown source of the list, e.g. '-', '+', '*'. */ -} MD_BLOCK_UL_DETAIL; - -/* Detailed info for MD_BLOCK_OL. */ -typedef struct MD_BLOCK_OL_DETAIL { - unsigned start; /* Start index of the ordered list. */ - int is_tight; /* Non-zero if tight list, zero if loose. */ - MD_CHAR mark_delimiter; /* Character delimiting the item marks in MarkDown source, e.g. '.' or ')' */ -} MD_BLOCK_OL_DETAIL; - -/* Detailed info for MD_BLOCK_LI. */ -typedef struct MD_BLOCK_LI_DETAIL { - int is_task; /* Can be non-zero only with MD_FLAG_TASKLISTS */ - MD_CHAR task_mark; /* If is_task, then one of 'x', 'X' or ' '. Undefined otherwise. */ - MD_OFFSET task_mark_offset; /* If is_task, then offset in the input of the char between '[' and ']'. */ -} MD_BLOCK_LI_DETAIL; - -/* Detailed info for MD_BLOCK_H. */ -typedef struct MD_BLOCK_H_DETAIL { - unsigned level; /* Header level (1 - 6) */ -} MD_BLOCK_H_DETAIL; - -/* Detailed info for MD_BLOCK_CODE. */ -typedef struct MD_BLOCK_CODE_DETAIL { - MD_ATTRIBUTE info; - MD_ATTRIBUTE lang; - MD_CHAR fence_char; /* The character used for fenced code block; or zero for indented code block. */ -} MD_BLOCK_CODE_DETAIL; - -/* Detailed info for MD_BLOCK_TABLE. */ -typedef struct MD_BLOCK_TABLE_DETAIL { - unsigned col_count; /* Count of columns in the table. */ - unsigned head_row_count; /* Count of rows in the table header (currently always 1) */ - unsigned body_row_count; /* Count of rows in the table body */ -} MD_BLOCK_TABLE_DETAIL; - -/* Detailed info for MD_BLOCK_TH and MD_BLOCK_TD. */ -typedef struct MD_BLOCK_TD_DETAIL { - MD_ALIGN align; -} MD_BLOCK_TD_DETAIL; - -/* Detailed info for MD_SPAN_A. */ -typedef struct MD_SPAN_A_DETAIL { - MD_ATTRIBUTE href; - MD_ATTRIBUTE title; -} MD_SPAN_A_DETAIL; - -/* Detailed info for MD_SPAN_IMG. */ -typedef struct MD_SPAN_IMG_DETAIL { - MD_ATTRIBUTE src; - MD_ATTRIBUTE title; -} MD_SPAN_IMG_DETAIL; - -/* Detailed info for MD_SPAN_WIKILINK. */ -typedef struct MD_SPAN_WIKILINK { - MD_ATTRIBUTE target; -} MD_SPAN_WIKILINK_DETAIL; - -/* Flags specifying extensions/deviations from CommonMark specification. - * - * By default (when MD_PARSER::flags == 0), we follow CommonMark specification. - * The following flags may allow some extensions or deviations from it. - */ -#define MD_FLAG_COLLAPSEWHITESPACE 0x0001 /* In MD_TEXT_NORMAL, collapse non-trivial whitespace into single ' ' */ -#define MD_FLAG_PERMISSIVEATXHEADERS 0x0002 /* Do not require space in ATX headers ( ###header ) */ -#define MD_FLAG_PERMISSIVEURLAUTOLINKS 0x0004 /* Recognize URLs as autolinks even without '<', '>' */ -#define MD_FLAG_PERMISSIVEEMAILAUTOLINKS 0x0008 /* Recognize e-mails as autolinks even without '<', '>' and 'mailto:' */ -#define MD_FLAG_NOINDENTEDCODEBLOCKS 0x0010 /* Disable indented code blocks. (Only fenced code works.) */ -#define MD_FLAG_NOHTMLBLOCKS 0x0020 /* Disable raw HTML blocks. */ -#define MD_FLAG_NOHTMLSPANS 0x0040 /* Disable raw HTML (inline). */ -#define MD_FLAG_TABLES 0x0100 /* Enable tables extension. */ -#define MD_FLAG_STRIKETHROUGH 0x0200 /* Enable strikethrough extension. */ -#define MD_FLAG_PERMISSIVEWWWAUTOLINKS 0x0400 /* Enable WWW autolinks (even without any scheme prefix, if they begin with 'www.') */ -#define MD_FLAG_TASKLISTS 0x0800 /* Enable task list extension. */ -#define MD_FLAG_LATEXMATHSPANS 0x1000 /* Enable $ and $$ containing LaTeX equations. */ -#define MD_FLAG_WIKILINKS 0x2000 /* Enable wiki links extension. */ -#define MD_FLAG_UNDERLINE 0x4000 /* Enable underline extension (and disables '_' for normal emphasis). */ - -#define MD_FLAG_PERMISSIVEAUTOLINKS (MD_FLAG_PERMISSIVEEMAILAUTOLINKS | MD_FLAG_PERMISSIVEURLAUTOLINKS | MD_FLAG_PERMISSIVEWWWAUTOLINKS) -#define MD_FLAG_NOHTML (MD_FLAG_NOHTMLBLOCKS | MD_FLAG_NOHTMLSPANS) - -/* Convenient sets of flags corresponding to well-known Markdown dialects. - * - * Note we may only support subset of features of the referred dialect. - * The constant just enables those extensions which bring us as close as - * possible given what features we implement. - * - * ABI compatibility note: Meaning of these can change in time as new - * extensions, bringing the dialect closer to the original, are implemented. - */ -#define MD_DIALECT_COMMONMARK 0 -#define MD_DIALECT_GITHUB (MD_FLAG_PERMISSIVEAUTOLINKS | MD_FLAG_TABLES | MD_FLAG_STRIKETHROUGH | MD_FLAG_TASKLISTS) - -/* Parser structure. - */ -typedef struct MD_PARSER { - /* Reserved. Set to zero. - */ - unsigned abi_version; - - /* Dialect options. Bitmask of MD_FLAG_xxxx values. - */ - unsigned flags; - - /* Caller-provided rendering callbacks. - * - * For some block/span types, more detailed information is provided in a - * type-specific structure pointed by the argument 'detail'. - * - * The last argument of all callbacks, 'userdata', is just propagated from - * md_parse() and is available for any use by the application. - * - * Note any strings provided to the callbacks as their arguments or as - * members of any detail structure are generally not zero-terminated. - * Application has to take the respective size information into account. - * - * Any rendering callback may abort further parsing of the document by - * returning non-zero. - */ - int (*enter_block)(MD_BLOCKTYPE /*type*/, void* /*detail*/, void* /*userdata*/); - int (*leave_block)(MD_BLOCKTYPE /*type*/, void* /*detail*/, void* /*userdata*/); - - int (*enter_span)(MD_SPANTYPE /*type*/, void* /*detail*/, void* /*userdata*/); - int (*leave_span)(MD_SPANTYPE /*type*/, void* /*detail*/, void* /*userdata*/); - - int (*text)(MD_TEXTTYPE /*type*/, const MD_CHAR* /*text*/, MD_SIZE /*size*/, void* /*userdata*/); - - /* Debug callback. Optional (may be NULL). - * - * If provided and something goes wrong, this function gets called. - * This is intended for debugging and problem diagnosis for developers; - * it is not intended to provide any errors suitable for displaying to an - * end user. - */ - void (*debug_log)(const char* /*msg*/, void* /*userdata*/); - - /* Reserved. Set to NULL. - */ - void (*syntax)(void); -} MD_PARSER; - - -/* For backward compatibility. Do not use in new code. - */ -typedef MD_PARSER MD_RENDERER; - - -/* Parse the Markdown document stored in the string 'text' of size 'size'. - * The parser provides callbacks to be called during the parsing so the - * caller can render the document on the screen or convert the Markdown - * to another format. - * - * Zero is returned on success. If a runtime error occurs (e.g. a memory - * fails), -1 is returned. If the processing is aborted due any callback - * returning non-zero, the return value of the callback is returned. - */ -int md_parse(const MD_CHAR* text, MD_SIZE size, const MD_PARSER* parser, void* userdata); - - -#ifdef __cplusplus - } /* extern "C" { */ -#endif - -#endif /* MD4C_H */ diff --git a/src/md4c.pc.in b/src/md4c.pc.in deleted file mode 100644 index cd8842d..0000000 --- a/src/md4c.pc.in +++ /dev/null @@ -1,13 +0,0 @@ -prefix=@CMAKE_INSTALL_PREFIX@ -exec_prefix=@CMAKE_INSTALL_PREFIX@ -libdir=${exec_prefix}/@CMAKE_INSTALL_LIBDIR@ -includedir=${prefix}/@CMAKE_INSTALL_INCLUDEDIR@ - -Name: @PROJECT_NAME@ -Description: Markdown parser library with a SAX-like callback-based interface. -Version: @PROJECT_VERSION@ -URL: @PROJECT_URL@ - -Requires: -Libs: -L${libdir} -lmd4c -Cflags: -I${includedir} |