1 /* 2 * MD4C: Markdown parser for C 3 * (http://github.com/mity/md4c) 4 * 5 * Copyright (c) 2016-2020 Martin Mitas 6 * 7 * Permission is hereby granted, free of charge, to any person obtaining a 8 * copy of this software and associated documentation files (the "Software"), 9 * to deal in the Software without restriction, including without limitation 10 * the rights to use, copy, modify, merge, publish, distribute, sublicense, 11 * and/or sell copies of the Software, and to permit persons to whom the 12 * Software is furnished to do so, subject to the following conditions: 13 * 14 * The above copyright notice and this permission notice shall be included in 15 * all copies or substantial portions of the Software. 16 * 17 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS 18 * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 19 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 20 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 21 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING 22 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS 23 * IN THE SOFTWARE. 24 */ 25 26 #ifndef MD4C_H 27 #define MD4C_H 28 29 #ifdef __cplusplus 30 extern "C" { 31 #endif 32 33 #if defined MD4C_USE_UTF16 34 /* Magic to support UTF-16. Note that in order to use it, you have to define 35 * the macro MD4C_USE_UTF16 both when building MD4C as well as when 36 * including this header in your code. */ 37 #ifdef _WIN32 38 #include <windows.h> 39 typedef WCHAR MD_CHAR; 40 #else 41 #error MD4C_USE_UTF16 is only supported on Windows. 42 #endif 43 #else 44 typedef char MD_CHAR; 45 #endif 46 47 typedef unsigned MD_SIZE; 48 typedef unsigned MD_OFFSET; 49 50 51 /* Block represents a part of document hierarchy structure like a paragraph 52 * or list item. 53 */ 54 typedef enum MD_BLOCKTYPE { 55 /* <body>...</body> */ 56 MD_BLOCK_DOC = 0, 57 58 /* <blockquote>...</blockquote> */ 59 MD_BLOCK_QUOTE, 60 61 /* <ul>...</ul> 62 * Detail: Structure MD_BLOCK_UL_DETAIL. */ 63 MD_BLOCK_UL, 64 65 /* <ol>...</ol> 66 * Detail: Structure MD_BLOCK_OL_DETAIL. */ 67 MD_BLOCK_OL, 68 69 /* <li>...</li> 70 * Detail: Structure MD_BLOCK_LI_DETAIL. */ 71 MD_BLOCK_LI, 72 73 /* <hr> */ 74 MD_BLOCK_HR, 75 76 /* <h1>...</h1> (for levels up to 6) 77 * Detail: Structure MD_BLOCK_H_DETAIL. */ 78 MD_BLOCK_H, 79 80 /* <pre><code>...</code></pre> 81 * Note the text lines within code blocks are terminated with '\n' 82 * instead of explicit MD_TEXT_BR. */ 83 MD_BLOCK_CODE, 84 85 /* Raw HTML block. This itself does not correspond to any particular HTML 86 * tag. The contents of it _is_ raw HTML source intended to be put 87 * in verbatim form to the HTML output. */ 88 MD_BLOCK_HTML, 89 90 /* <p>...</p> */ 91 MD_BLOCK_P, 92 93 /* <table>...</table> and its contents. 94 * Detail: Structure MD_BLOCK_TD_DETAIL (used with MD_BLOCK_TH and MD_BLOCK_TD) 95 * Note all of these are used only if extension MD_FLAG_TABLES is enabled. */ 96 MD_BLOCK_TABLE, 97 MD_BLOCK_THEAD, 98 MD_BLOCK_TBODY, 99 MD_BLOCK_TR, 100 MD_BLOCK_TH, 101 MD_BLOCK_TD 102 } MD_BLOCKTYPE; 103 104 /* Span represents an in-line piece of a document which should be rendered with 105 * the same font, color and other attributes. A sequence of spans forms a block 106 * like paragraph or list item. */ 107 typedef enum MD_SPANTYPE { 108 /* <em>...</em> */ 109 MD_SPAN_EM, 110 111 /* <strong>...</strong> */ 112 MD_SPAN_STRONG, 113 114 /* <a href="xxx">...</a> 115 * Detail: Structure MD_SPAN_A_DETAIL. */ 116 MD_SPAN_A, 117 118 /* <img src="xxx">...</a> 119 * Detail: Structure MD_SPAN_IMG_DETAIL. 120 * Note: Image text can contain nested spans and even nested images. 121 * If rendered into ALT attribute of HTML <IMG> tag, it's responsibility 122 * of the parser to deal with it. 123 */ 124 MD_SPAN_IMG, 125 126 /* <code>...</code> */ 127 MD_SPAN_CODE, 128 129 /* <del>...</del> 130 * Note: Recognized only when MD_FLAG_STRIKETHROUGH is enabled. 131 */ 132 MD_SPAN_DEL, 133 134 /* For recognizing inline ($) and display ($$) equations 135 * Note: Recognized only when MD_FLAG_LATEXMATHSPANS is enabled. 136 */ 137 MD_SPAN_LATEXMATH, 138 MD_SPAN_LATEXMATH_DISPLAY, 139 140 /* Wiki links 141 * Note: Recognized only when MD_FLAG_WIKILINKS is enabled. 142 */ 143 MD_SPAN_WIKILINK, 144 145 /* <u>...</u> 146 * Note: Recognized only when MD_FLAG_UNDERLINE is enabled. */ 147 MD_SPAN_U 148 } MD_SPANTYPE; 149 150 /* Text is the actual textual contents of span. */ 151 typedef enum MD_TEXTTYPE { 152 /* Normal text. */ 153 MD_TEXT_NORMAL = 0, 154 155 /* NULL character. CommonMark requires replacing NULL character with 156 * the replacement char U+FFFD, so this allows caller to do that easily. */ 157 MD_TEXT_NULLCHAR, 158 159 /* Line breaks. 160 * Note these are not sent from blocks with verbatim output (MD_BLOCK_CODE 161 * or MD_BLOCK_HTML). In such cases, '\n' is part of the text itself. */ 162 MD_TEXT_BR, /* <br> (hard break) */ 163 MD_TEXT_SOFTBR, /* '\n' in source text where it is not semantically meaningful (soft break) */ 164 165 /* Entity. 166 * (a) Named entity, e.g. 167 * (Note MD4C does not have a list of known entities. 168 * Anything matching the regexp /&[A-Za-z][A-Za-z0-9]{1,47};/ is 169 * treated as a named entity.) 170 * (b) Numerical entity, e.g. Ӓ 171 * (c) Hexadecimal entity, e.g. ካ 172 * 173 * As MD4C is mostly encoding agnostic, application gets the verbatim 174 * entity text into the MD_PARSER::text_callback(). */ 175 MD_TEXT_ENTITY, 176 177 /* Text in a code block (inside MD_BLOCK_CODE) or inlined code (`code`). 178 * If it is inside MD_BLOCK_CODE, it includes spaces for indentation and 179 * '\n' for new lines. MD_TEXT_BR and MD_TEXT_SOFTBR are not sent for this 180 * kind of text. */ 181 MD_TEXT_CODE, 182 183 /* Text is a raw HTML. If it is contents of a raw HTML block (i.e. not 184 * an inline raw HTML), then MD_TEXT_BR and MD_TEXT_SOFTBR are not used. 185 * The text contains verbatim '\n' for the new lines. */ 186 MD_TEXT_HTML, 187 188 /* Text is inside an equation. This is processed the same way as inlined code 189 * spans (`code`). */ 190 MD_TEXT_LATEXMATH 191 } MD_TEXTTYPE; 192 193 194 /* Alignment enumeration. */ 195 typedef enum MD_ALIGN { 196 MD_ALIGN_DEFAULT = 0, /* When unspecified. */ 197 MD_ALIGN_LEFT, 198 MD_ALIGN_CENTER, 199 MD_ALIGN_RIGHT 200 } MD_ALIGN; 201 202 203 /* String attribute. 204 * 205 * This wraps strings which are outside of a normal text flow and which are 206 * propagated within various detailed structures, but which still may contain 207 * string portions of different types like e.g. entities. 208 * 209 * So, for example, lets consider this image: 210 * 211 * ![image alt text](http://example.org/image.png 'foo " bar') 212 * 213 * The image alt text is propagated as a normal text via the MD_PARSER::text() 214 * callback. However, the image title ('foo " bar') is propagated as 215 * MD_ATTRIBUTE in MD_SPAN_IMG_DETAIL::title. 216 * 217 * Then the attribute MD_SPAN_IMG_DETAIL::title shall provide the following: 218 * -- [0]: "foo " (substr_types[0] == MD_TEXT_NORMAL; substr_offsets[0] == 0) 219 * -- [1]: """ (substr_types[1] == MD_TEXT_ENTITY; substr_offsets[1] == 4) 220 * -- [2]: " bar" (substr_types[2] == MD_TEXT_NORMAL; substr_offsets[2] == 10) 221 * -- [3]: (n/a) (n/a ; substr_offsets[3] == 14) 222 * 223 * Note that these invariants are always guaranteed: 224 * -- substr_offsets[0] == 0 225 * -- substr_offsets[LAST+1] == size 226 * -- Currently, only MD_TEXT_NORMAL, MD_TEXT_ENTITY, MD_TEXT_NULLCHAR 227 * substrings can appear. This could change only of the specification 228 * changes. 229 */ 230 typedef struct MD_ATTRIBUTE { 231 const MD_CHAR* text; 232 MD_SIZE size; 233 const MD_TEXTTYPE* substr_types; 234 const MD_OFFSET* substr_offsets; 235 } MD_ATTRIBUTE; 236 237 238 /* Detailed info for MD_BLOCK_UL. */ 239 typedef struct MD_BLOCK_UL_DETAIL { 240 int is_tight; /* Non-zero if tight list, zero if loose. */ 241 MD_CHAR mark; /* Item bullet character in MarkDown source of the list, e.g. '-', '+', '*'. */ 242 } MD_BLOCK_UL_DETAIL; 243 244 /* Detailed info for MD_BLOCK_OL. */ 245 typedef struct MD_BLOCK_OL_DETAIL { 246 unsigned start; /* Start index of the ordered list. */ 247 int is_tight; /* Non-zero if tight list, zero if loose. */ 248 MD_CHAR mark_delimiter; /* Character delimiting the item marks in MarkDown source, e.g. '.' or ')' */ 249 } MD_BLOCK_OL_DETAIL; 250 251 /* Detailed info for MD_BLOCK_LI. */ 252 typedef struct MD_BLOCK_LI_DETAIL { 253 int is_task; /* Can be non-zero only with MD_FLAG_TASKLISTS */ 254 MD_CHAR task_mark; /* If is_task, then one of 'x', 'X' or ' '. Undefined otherwise. */ 255 MD_OFFSET task_mark_offset; /* If is_task, then offset in the input of the char between '[' and ']'. */ 256 } MD_BLOCK_LI_DETAIL; 257 258 /* Detailed info for MD_BLOCK_H. */ 259 typedef struct MD_BLOCK_H_DETAIL { 260 unsigned level; /* Header level (1 - 6) */ 261 } MD_BLOCK_H_DETAIL; 262 263 /* Detailed info for MD_BLOCK_CODE. */ 264 typedef struct MD_BLOCK_CODE_DETAIL { 265 MD_ATTRIBUTE info; 266 MD_ATTRIBUTE lang; 267 MD_CHAR fence_char; /* The character used for fenced code block; or zero for indented code block. */ 268 } MD_BLOCK_CODE_DETAIL; 269 270 /* Detailed info for MD_BLOCK_TH and MD_BLOCK_TD. */ 271 typedef struct MD_BLOCK_TD_DETAIL { 272 MD_ALIGN align; 273 } MD_BLOCK_TD_DETAIL; 274 275 /* Detailed info for MD_SPAN_A. */ 276 typedef struct MD_SPAN_A_DETAIL { 277 MD_ATTRIBUTE href; 278 MD_ATTRIBUTE title; 279 } MD_SPAN_A_DETAIL; 280 281 /* Detailed info for MD_SPAN_IMG. */ 282 typedef struct MD_SPAN_IMG_DETAIL { 283 MD_ATTRIBUTE src; 284 MD_ATTRIBUTE title; 285 } MD_SPAN_IMG_DETAIL; 286 287 /* Detailed info for MD_SPAN_WIKILINK. */ 288 typedef struct MD_SPAN_WIKILINK { 289 MD_ATTRIBUTE target; 290 } MD_SPAN_WIKILINK_DETAIL; 291 292 /* Flags specifying extensions/deviations from CommonMark specification. 293 * 294 * By default (when MD_PARSER::flags == 0), we follow CommonMark specification. 295 * The following flags may allow some extensions or deviations from it. 296 */ 297 #define MD_FLAG_COLLAPSEWHITESPACE 0x0001 /* In MD_TEXT_NORMAL, collapse non-trivial whitespace into single ' ' */ 298 #define MD_FLAG_PERMISSIVEATXHEADERS 0x0002 /* Do not require space in ATX headers ( ###header ) */ 299 #define MD_FLAG_PERMISSIVEURLAUTOLINKS 0x0004 /* Recognize URLs as autolinks even without '<', '>' */ 300 #define MD_FLAG_PERMISSIVEEMAILAUTOLINKS 0x0008 /* Recognize e-mails as autolinks even without '<', '>' and 'mailto:' */ 301 #define MD_FLAG_NOINDENTEDCODEBLOCKS 0x0010 /* Disable indented code blocks. (Only fenced code works.) */ 302 #define MD_FLAG_NOHTMLBLOCKS 0x0020 /* Disable raw HTML blocks. */ 303 #define MD_FLAG_NOHTMLSPANS 0x0040 /* Disable raw HTML (inline). */ 304 #define MD_FLAG_TABLES 0x0100 /* Enable tables extension. */ 305 #define MD_FLAG_STRIKETHROUGH 0x0200 /* Enable strikethrough extension. */ 306 #define MD_FLAG_PERMISSIVEWWWAUTOLINKS 0x0400 /* Enable WWW autolinks (even without any scheme prefix, if they begin with 'www.') */ 307 #define MD_FLAG_TASKLISTS 0x0800 /* Enable task list extension. */ 308 #define MD_FLAG_LATEXMATHSPANS 0x1000 /* Enable $ and $$ containing LaTeX equations. */ 309 #define MD_FLAG_WIKILINKS 0x2000 /* Enable wiki links extension. */ 310 #define MD_FLAG_UNDERLINE 0x4000 /* Enable underline extension (and disables '_' for normal emphasis). */ 311 312 #define MD_FLAG_PERMISSIVEAUTOLINKS (MD_FLAG_PERMISSIVEEMAILAUTOLINKS | MD_FLAG_PERMISSIVEURLAUTOLINKS | MD_FLAG_PERMISSIVEWWWAUTOLINKS) 313 #define MD_FLAG_NOHTML (MD_FLAG_NOHTMLBLOCKS | MD_FLAG_NOHTMLSPANS) 314 315 /* Convenient sets of flags corresponding to well-known Markdown dialects. 316 * 317 * Note we may only support subset of features of the referred dialect. 318 * The constant just enables those extensions which bring us as close as 319 * possible given what features we implement. 320 * 321 * ABI compatibility note: Meaning of these can change in time as new 322 * extensions, bringing the dialect closer to the original, are implemented. 323 */ 324 #define MD_DIALECT_COMMONMARK 0 325 #define MD_DIALECT_GITHUB (MD_FLAG_PERMISSIVEAUTOLINKS | MD_FLAG_TABLES | MD_FLAG_STRIKETHROUGH | MD_FLAG_TASKLISTS) 326 327 /* Parser structure. 328 */ 329 typedef struct MD_PARSER { 330 /* Reserved. Set to zero. 331 */ 332 unsigned abi_version; 333 334 /* Dialect options. Bitmask of MD_FLAG_xxxx values. 335 */ 336 unsigned flags; 337 338 /* Caller-provided rendering callbacks. 339 * 340 * For some block/span types, more detailed information is provided in a 341 * type-specific structure pointed by the argument 'detail'. 342 * 343 * The last argument of all callbacks, 'userdata', is just propagated from 344 * md_parse() and is available for any use by the application. 345 * 346 * Note any strings provided to the callbacks as their arguments or as 347 * members of any detail structure are generally not zero-terminated. 348 * Application has to take the respective size information into account. 349 * 350 * Any rendering callback may abort further parsing of the document by 351 * returning non-zero. 352 */ 353 int (*enter_block)(MD_BLOCKTYPE /*type*/, void* /*detail*/, void* /*userdata*/); 354 int (*leave_block)(MD_BLOCKTYPE /*type*/, void* /*detail*/, void* /*userdata*/); 355 356 int (*enter_span)(MD_SPANTYPE /*type*/, void* /*detail*/, void* /*userdata*/); 357 int (*leave_span)(MD_SPANTYPE /*type*/, void* /*detail*/, void* /*userdata*/); 358 359 int (*text)(MD_TEXTTYPE /*type*/, const MD_CHAR* /*text*/, MD_SIZE /*size*/, void* /*userdata*/); 360 361 /* Debug callback. Optional (may be NULL). 362 * 363 * If provided and something goes wrong, this function gets called. 364 * This is intended for debugging and problem diagnosis for developers; 365 * it is not intended to provide any errors suitable for displaying to an 366 * end user. 367 */ 368 void (*debug_log)(const char* /*msg*/, void* /*userdata*/); 369 370 /* Reserved. Set to NULL. 371 */ 372 void (*syntax)(void); 373 } MD_PARSER; 374 375 376 /* For backward compatibility. Do not use in new code. 377 */ 378 typedef MD_PARSER MD_RENDERER; 379 380 381 /* Parse the Markdown document stored in the string 'text' of size 'size'. 382 * The parser provides callbacks to be called during the parsing so the 383 * caller can render the document on the screen or convert the Markdown 384 * to another format. 385 * 386 * Zero is returned on success. If a runtime error occurs (e.g. a memory 387 * fails), -1 is returned. If the processing is aborted due any callback 388 * returning non-zero, the return value of the callback is returned. 389 */ 390 int md_parse(const MD_CHAR* text, MD_SIZE size, const MD_PARSER* parser, void* userdata); 391 392 393 #ifdef __cplusplus 394 } /* extern "C" { */ 395 #endif 396 397 #endif /* MD4C_H */ 398