md4c.h (15511B)
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_TABLE_DETAIL (for MD_BLOCK_TABLE),
95 * structure MD_BLOCK_TD_DETAIL (for MD_BLOCK_TH and MD_BLOCK_TD)
96 * Note all of these are used only if extension MD_FLAG_TABLES is enabled. */
97 MD_BLOCK_TABLE,
98 MD_BLOCK_THEAD,
99 MD_BLOCK_TBODY,
100 MD_BLOCK_TR,
101 MD_BLOCK_TH,
102 MD_BLOCK_TD
103 } MD_BLOCKTYPE;
104
105 /* Span represents an in-line piece of a document which should be rendered with
106 * the same font, color and other attributes. A sequence of spans forms a block
107 * like paragraph or list item. */
108 typedef enum MD_SPANTYPE {
109 /* <em>...</em> */
110 MD_SPAN_EM,
111
112 /* <strong>...</strong> */
113 MD_SPAN_STRONG,
114
115 /* <a href="xxx">...</a>
116 * Detail: Structure MD_SPAN_A_DETAIL. */
117 MD_SPAN_A,
118
119 /* <img src="xxx">...</a>
120 * Detail: Structure MD_SPAN_IMG_DETAIL.
121 * Note: Image text can contain nested spans and even nested images.
122 * If rendered into ALT attribute of HTML <IMG> tag, it's responsibility
123 * of the parser to deal with it.
124 */
125 MD_SPAN_IMG,
126
127 /* <code>...</code> */
128 MD_SPAN_CODE,
129
130 /* <del>...</del>
131 * Note: Recognized only when MD_FLAG_STRIKETHROUGH is enabled.
132 */
133 MD_SPAN_DEL,
134
135 /* For recognizing inline ($) and display ($) equations
136 * Note: Recognized only when MD_FLAG_LATEXMATHSPANS is enabled.
137 */
138 MD_SPAN_LATEXMATH,
139 MD_SPAN_LATEXMATH_DISPLAY,
140
141 /* Wiki links
142 * Note: Recognized only when MD_FLAG_WIKILINKS is enabled.
143 */
144 MD_SPAN_WIKILINK,
145
146 /* <u>...</u>
147 * Note: Recognized only when MD_FLAG_UNDERLINE is enabled. */
148 MD_SPAN_U,
149 MD_SPAN_FNT,
150 MD_SPAN_INV,
151 MD_SPAN_COC,
152 MD_SPAN_BLI,
153 MD_SPAN_ANCHOR,
154 /* This span type is issued by md4c
155 * MD_SPAN_COLOR allows supporting RGB colors:
156 * [text with colors](#1#13)
157 * md4c treats colors as MD_SPAN_A and the parsing of the color
158 * is done by the user.
159 */
160 MD_SPAN_COLOR,
161 } MD_SPANTYPE;
162
163 /* Text is the actual textual contents of span. */
164 typedef enum MD_TEXTTYPE {
165 /* Normal text. */
166 MD_TEXT_NORMAL = 0,
167
168 /* NULL character. CommonMark requires replacing NULL character with
169 * the replacement char U+FFFD, so this allows caller to do that easily. */
170 MD_TEXT_NULLCHAR,
171
172 /* Line breaks.
173 * Note these are not sent from blocks with verbatim output (MD_BLOCK_CODE
174 * or MD_BLOCK_HTML). In such cases, '\n' is part of the text itself. */
175 MD_TEXT_BR, /* <br> (hard break) */
176 MD_TEXT_SOFTBR, /* '\n' in source text where it is not semantically meaningful (soft break) */
177
178 /* Entity.
179 * (a) Named entity, e.g.
180 * (Note MD4C does not have a list of known entities.
181 * Anything matching the regexp /&[A-Za-z][A-Za-z0-9]{1,47};/ is
182 * treated as a named entity.)
183 * (b) Numerical entity, e.g. Ӓ
184 * (c) Hexadecimal entity, e.g. ካ
185 *
186 * As MD4C is mostly encoding agnostic, application gets the verbatim
187 * entity text into the MD_PARSER::text_callback(). */
188 MD_TEXT_ENTITY,
189
190 /* Text in a code block (inside MD_BLOCK_CODE) or inlined code (`code`).
191 * If it is inside MD_BLOCK_CODE, it includes spaces for indentation and
192 * '\n' for new lines. MD_TEXT_BR and MD_TEXT_SOFTBR are not sent for this
193 * kind of text. */
194 MD_TEXT_CODE,
195
196 /* Text is a raw HTML. If it is contents of a raw HTML block (i.e. not
197 * an inline raw HTML), then MD_TEXT_BR and MD_TEXT_SOFTBR are not used.
198 * The text contains verbatim '\n' for the new lines. */
199 MD_TEXT_HTML,
200
201 /* Text is inside an equation. This is processed the same way as inlined code
202 * spans (`code`). */
203 MD_TEXT_LATEXMATH
204 } MD_TEXTTYPE;
205
206
207 /* Alignment enumeration. */
208 typedef enum MD_ALIGN {
209 MD_ALIGN_DEFAULT = 0, /* When unspecified. */
210 MD_ALIGN_LEFT,
211 MD_ALIGN_CENTER,
212 MD_ALIGN_RIGHT
213 } MD_ALIGN;
214
215
216 /* String attribute.
217 *
218 * This wraps strings which are outside of a normal text flow and which are
219 * propagated within various detailed structures, but which still may contain
220 * string portions of different types like e.g. entities.
221 *
222 * So, for example, lets consider this image:
223 *
224 * ![image alt text](http://example.org/image.png 'foo " bar')
225 *
226 * The image alt text is propagated as a normal text via the MD_PARSER::text()
227 * callback. However, the image title ('foo " bar') is propagated as
228 * MD_ATTRIBUTE in MD_SPAN_IMG_DETAIL::title.
229 *
230 * Then the attribute MD_SPAN_IMG_DETAIL::title shall provide the following:
231 * -- [0]: "foo " (substr_types[0] == MD_TEXT_NORMAL; substr_offsets[0] == 0)
232 * -- [1]: """ (substr_types[1] == MD_TEXT_ENTITY; substr_offsets[1] == 4)
233 * -- [2]: " bar" (substr_types[2] == MD_TEXT_NORMAL; substr_offsets[2] == 10)
234 * -- [3]: (n/a) (n/a ; substr_offsets[3] == 14)
235 *
236 * Note that these invariants are always guaranteed:
237 * -- substr_offsets[0] == 0
238 * -- substr_offsets[LAST+1] == size
239 * -- Currently, only MD_TEXT_NORMAL, MD_TEXT_ENTITY, MD_TEXT_NULLCHAR
240 * substrings can appear. This could change only of the specification
241 * changes.
242 */
243 typedef struct MD_ATTRIBUTE {
244 const MD_CHAR* text;
245 MD_SIZE size;
246 const MD_TEXTTYPE* substr_types;
247 const MD_OFFSET* substr_offsets;
248 } MD_ATTRIBUTE;
249
250
251 /* Detailed info for MD_BLOCK_UL. */
252 typedef struct MD_BLOCK_UL_DETAIL {
253 int is_tight; /* Non-zero if tight list, zero if loose. */
254 MD_CHAR mark; /* Item bullet character in MarkDown source of the list, e.g. '-', '+', '*'. */
255 } MD_BLOCK_UL_DETAIL;
256
257 /* Detailed info for MD_BLOCK_OL. */
258 typedef struct MD_BLOCK_OL_DETAIL {
259 unsigned start; /* Start index of the ordered list. */
260 int is_tight; /* Non-zero if tight list, zero if loose. */
261 MD_CHAR mark_delimiter; /* Character delimiting the item marks in MarkDown source, e.g. '.' or ')' */
262 } MD_BLOCK_OL_DETAIL;
263
264 /* Detailed info for MD_BLOCK_LI. */
265 typedef struct MD_BLOCK_LI_DETAIL {
266 int is_task; /* Can be non-zero only with MD_FLAG_TASKLISTS */
267 MD_CHAR task_mark; /* If is_task, then one of 'x', 'X' or ' '. Undefined otherwise. */
268 MD_OFFSET task_mark_offset; /* If is_task, then offset in the input of the char between '[' and ']'. */
269 } MD_BLOCK_LI_DETAIL;
270
271 /* Detailed info for MD_BLOCK_H. */
272 typedef struct MD_BLOCK_H_DETAIL {
273 unsigned level; /* Header level (1 - 6) */
274 } MD_BLOCK_H_DETAIL;
275
276 /* Detailed info for MD_BLOCK_CODE. */
277 typedef struct MD_BLOCK_CODE_DETAIL {
278 MD_ATTRIBUTE info;
279 MD_ATTRIBUTE lang;
280 MD_CHAR fence_char; /* The character used for fenced code block; or zero for indented code block. */
281 } MD_BLOCK_CODE_DETAIL;
282
283 /* Detailed info for MD_BLOCK_TABLE. */
284 typedef struct MD_BLOCK_TABLE_DETAIL {
285 unsigned col_count; /* Count of columns in the table. */
286 unsigned head_row_count; /* Count of rows in the table header (currently always 1) */
287 unsigned body_row_count; /* Count of rows in the table body */
288 } MD_BLOCK_TABLE_DETAIL;
289
290 /* Detailed info for MD_BLOCK_TH and MD_BLOCK_TD. */
291 typedef struct MD_BLOCK_TD_DETAIL {
292 MD_ALIGN align;
293 } MD_BLOCK_TD_DETAIL;
294
295 /* Detailed info for MD_SPAN_A. */
296 typedef struct MD_SPAN_A_DETAIL {
297 MD_ATTRIBUTE href;
298 MD_ATTRIBUTE title;
299 } MD_SPAN_A_DETAIL;
300
301 /* Detailed info for MD_SPAN_IMG. */
302 typedef struct MD_SPAN_IMG_DETAIL {
303 MD_ATTRIBUTE src;
304 MD_ATTRIBUTE title;
305 } MD_SPAN_IMG_DETAIL;
306
307 /* Detailed info for MD_SPAN_WIKILINK. */
308 typedef struct MD_SPAN_WIKILINK {
309 MD_ATTRIBUTE target;
310 } MD_SPAN_WIKILINK_DETAIL;
311
312 /* Flags specifying extensions/deviations from CommonMark specification.
313 *
314 * By default (when MD_PARSER::flags == 0), we follow CommonMark specification.
315 * The following flags may allow some extensions or deviations from it.
316 */
317 #define MD_FLAG_COLLAPSEWHITESPACE 0x0001 /* In MD_TEXT_NORMAL, collapse non-trivial whitespace into single ' ' */
318 #define MD_FLAG_PERMISSIVEATXHEADERS 0x0002 /* Do not require space in ATX headers ( ###header ) */
319 #define MD_FLAG_PERMISSIVEURLAUTOLINKS 0x0004 /* Recognize URLs as autolinks even without '<', '>' */
320 #define MD_FLAG_PERMISSIVEEMAILAUTOLINKS 0x0008 /* Recognize e-mails as autolinks even without '<', '>' and 'mailto:' */
321 #define MD_FLAG_NOINDENTEDCODEBLOCKS 0x0010 /* Disable indented code blocks. (Only fenced code works.) */
322 #define MD_FLAG_NOHTMLBLOCKS 0x0020 /* Disable raw HTML blocks. */
323 #define MD_FLAG_NOHTMLSPANS 0x0040 /* Disable raw HTML (inline). */
324 #define MD_FLAG_TABLES 0x0100 /* Enable tables extension. */
325 #define MD_FLAG_STRIKETHROUGH 0x0200 /* Enable strikethrough extension. */
326 #define MD_FLAG_PERMISSIVEWWWAUTOLINKS 0x0400 /* Enable WWW autolinks (even without any scheme prefix, if they begin with 'www.') */
327 #define MD_FLAG_TASKLISTS 0x0800 /* Enable task list extension. */
328 #define MD_FLAG_LATEXMATHSPANS 0x1000 /* Enable $ and $ containing LaTeX equations. */
329 #define MD_FLAG_WIKILINKS 0x2000 /* Enable wiki links extension. */
330 #define MD_FLAG_UNDERLINE 0x4000 /* Enable underline extension (and disables '_' for normal emphasis). */
331
332 #define MD_FLAG_PERMISSIVEAUTOLINKS (MD_FLAG_PERMISSIVEEMAILAUTOLINKS | MD_FLAG_PERMISSIVEURLAUTOLINKS | MD_FLAG_PERMISSIVEWWWAUTOLINKS)
333 #define MD_FLAG_NOHTML (MD_FLAG_NOHTMLBLOCKS | MD_FLAG_NOHTMLSPANS)
334
335 /* Convenient sets of flags corresponding to well-known Markdown dialects.
336 *
337 * Note we may only support subset of features of the referred dialect.
338 * The constant just enables those extensions which bring us as close as
339 * possible given what features we implement.
340 *
341 * ABI compatibility note: Meaning of these can change in time as new
342 * extensions, bringing the dialect closer to the original, are implemented.
343 */
344 #define MD_DIALECT_COMMONMARK 0
345 #define MD_DIALECT_GITHUB (MD_FLAG_PERMISSIVEAUTOLINKS | MD_FLAG_TABLES | MD_FLAG_STRIKETHROUGH | MD_FLAG_TASKLISTS)
346
347 /* Parser structure.
348 */
349 typedef struct MD_PARSER {
350 /* Reserved. Set to zero.
351 */
352 unsigned abi_version;
353
354 /* Dialect options. Bitmask of MD_FLAG_xxxx values.
355 */
356 unsigned flags;
357
358 /* Caller-provided rendering callbacks.
359 *
360 * For some block/span types, more detailed information is provided in a
361 * type-specific structure pointed by the argument 'detail'.
362 *
363 * The last argument of all callbacks, 'userdata', is just propagated from
364 * md_parse() and is available for any use by the application.
365 *
366 * Note any strings provided to the callbacks as their arguments or as
367 * members of any detail structure are generally not zero-terminated.
368 * Application has to take the respective size information into account.
369 *
370 * Any rendering callback may abort further parsing of the document by
371 * returning non-zero.
372 */
373 int (*enter_block)(MD_BLOCKTYPE /*type*/, void* /*detail*/, void* /*userdata*/);
374 int (*leave_block)(MD_BLOCKTYPE /*type*/, void* /*detail*/, void* /*userdata*/);
375
376 int (*enter_span)(MD_SPANTYPE /*type*/, void* /*detail*/, void* /*userdata*/);
377 int (*leave_span)(MD_SPANTYPE /*type*/, void* /*detail*/, void* /*userdata*/);
378
379 int (*text)(MD_TEXTTYPE /*type*/, const MD_CHAR* /*text*/, MD_SIZE /*size*/, void* /*userdata*/);
380
381 /* Debug callback. Optional (may be NULL).
382 *
383 * If provided and something goes wrong, this function gets called.
384 * This is intended for debugging and problem diagnosis for developers;
385 * it is not intended to provide any errors suitable for displaying to an
386 * end user.
387 */
388 void (*debug_log)(const char* /*msg*/, void* /*userdata*/);
389
390 /* Reserved. Set to NULL.
391 */
392 void (*syntax)(void);
393 } MD_PARSER;
394
395
396 /* For backward compatibility. Do not use in new code.
397 */
398 typedef MD_PARSER MD_RENDERER;
399
400
401 /* Parse the Markdown document stored in the string 'text' of size 'size'.
402 * The parser provides callbacks to be called during the parsing so the
403 * caller can render the document on the screen or convert the Markdown
404 * to another format.
405 *
406 * Zero is returned on success. If a runtime error occurs (e.g. a memory
407 * fails), -1 is returned. If the processing is aborted due any callback
408 * returning non-zero, the return value of the callback is returned.
409 */
410 int md_parse(const MD_CHAR* text, MD_SIZE size, const MD_PARSER* parser, void* userdata);
411
412
413 #ifdef __cplusplus
414 } /* extern "C" { */
415 #endif
416
417 #endif /* MD4C_H */