pdoc.markdown2
A fast and complete Python implementation of Markdown.
[from http://daringfireball.net/projects/markdown/]
Markdown is a text-to-HTML filter; it translates an easy-to-read / easy-to-write structured text format into HTML. Markdown's text format is most similar to that of plain text email, and supports features such as headers, emphasis, code blocks, blockquotes, and links.
Markdown's syntax is designed not as a generic markup language, but specifically to serve as a front-end to (X)HTML. You can use span-level HTML tags anywhere in a Markdown document, and you can use block level HTML tags (like
andas well).
Module usage:
>>> import markdown2 >>> markdown2.markdown("*boo!*") # or use `html = markdown_path(PATH)` u'<p><em>boo!</em></p>\n' >>> markdowner = Markdown() >>> markdowner.convert("*boo!*") u'<p><em>boo!</em></p>\n' >>> markdowner.convert("**boom!**") u'<p><strong>boom!</strong></p>\n'This implementation of Markdown implements the full "core" syntax plus a number of extras (e.g., code syntax coloring, footnotes) as described on https://github.com/trentm/python-markdown2/wiki/Extras.
1# fmt: off 2# flake8: noqa 3# type: ignore 4# Taken from here: https://github.com/trentm/python-markdown2/blob/8d3a65bc7d4f8b64af89f668eb6c60841dc0578c/lib/markdown2.py 5 6#!/usr/bin/env python 7# Copyright (c) 2012 Trent Mick. 8# Copyright (c) 2007-2008 ActiveState Corp. 9# License: MIT (http://www.opensource.org/licenses/mit-license.php) 10 11r"""A fast and complete Python implementation of Markdown. 12 13[from http://daringfireball.net/projects/markdown/] 14> Markdown is a text-to-HTML filter; it translates an easy-to-read / 15> easy-to-write structured text format into HTML. Markdown's text 16> format is most similar to that of plain text email, and supports 17> features such as headers, *emphasis*, code blocks, blockquotes, and 18> links. 19> 20> Markdown's syntax is designed not as a generic markup language, but 21> specifically to serve as a front-end to (X)HTML. You can use span-level 22> HTML tags anywhere in a Markdown document, and you can use block level 23> HTML tags (like <div> and <table> as well). 24 25Module usage: 26 27 >>> import markdown2 28 >>> markdown2.markdown("*boo!*") # or use `html = markdown_path(PATH)` 29 u'<p><em>boo!</em></p>\n' 30 31 >>> markdowner = Markdown() 32 >>> markdowner.convert("*boo!*") 33 u'<p><em>boo!</em></p>\n' 34 >>> markdowner.convert("**boom!**") 35 u'<p><strong>boom!</strong></p>\n' 36 37This implementation of Markdown implements the full "core" syntax plus a 38number of extras (e.g., code syntax coloring, footnotes) as described on 39<https://github.com/trentm/python-markdown2/wiki/Extras>. 40""" 41 42cmdln_desc = """A fast and complete Python implementation of Markdown, a 43text-to-HTML conversion tool for web writers. 44 45Supported extra syntax options (see -x|--extras option below and 46see <https://github.com/trentm/python-markdown2/wiki/Extras> for details): 47 48* admonitions: Enable parsing of RST admonitions. 49* breaks: Control where hard breaks are inserted in the markdown. 50 Options include: 51 - on_newline: Replace single new line characters with <br> when True 52 - on_backslash: Replace backslashes at the end of a line with <br> 53* break-on-newline: Alias for the on_newline option in the breaks extra. 54* code-friendly: Disable _ and __ for em and strong. 55* cuddled-lists: Allow lists to be cuddled to the preceding paragraph. 56* fenced-code-blocks: Allows a code block to not have to be indented 57 by fencing it with '```' on a line before and after. Based on 58 <http://github.github.com/github-flavored-markdown/> with support for 59 syntax highlighting. 60* footnotes: Support footnotes as in use on daringfireball.net and 61 implemented in other Markdown processors (tho not in Markdown.pl v1.0.1). 62* header-ids: Adds "id" attributes to headers. The id value is a slug of 63 the header text. 64* highlightjs-lang: Allows specifying the language which used for syntax 65 highlighting when using fenced-code-blocks and highlightjs. 66* html-classes: Takes a dict mapping html tag names (lowercase) to a 67 string to use for a "class" tag attribute. Currently only supports "img", 68 "table", "thead", "pre", "code", "ul" and "ol" tags. Add an issue if you require 69 this for other tags. 70* link-patterns: Auto-link given regex patterns in text (e.g. bug number 71 references, revision number references). 72* markdown-in-html: Allow the use of `markdown="1"` in a block HTML tag to 73 have markdown processing be done on its contents. Similar to 74 <http://michelf.com/projects/php-markdown/extra/#markdown-attr> but with 75 some limitations. 76* metadata: Extract metadata from a leading '---'-fenced block. 77 See <https://github.com/trentm/python-markdown2/issues/77> for details. 78* middle-word-em: Allows or disallows emphasis syntax in the middle of words, 79 defaulting to allow. Disabling this means that `this_text_here` will not be 80 converted to `this<em>text</em>here`. 81* nofollow: Add `rel="nofollow"` to add `<a>` tags with an href. See 82 <http://en.wikipedia.org/wiki/Nofollow>. 83* numbering: Support of generic counters. Non standard extension to 84 allow sequential numbering of figures, tables, equations, exhibits etc. 85* pyshell: Treats unindented Python interactive shell sessions as <code> 86 blocks. 87* smarty-pants: Replaces ' and " with curly quotation marks or curly 88 apostrophes. Replaces --, ---, ..., and . . . with en dashes, em dashes, 89 and ellipses. 90* spoiler: A special kind of blockquote commonly hidden behind a 91 click on SO. Syntax per <http://meta.stackexchange.com/a/72878>. 92* strike: text inside of double tilde is ~~strikethrough~~ 93* tag-friendly: Requires atx style headers to have a space between the # and 94 the header text. Useful for applications that require twitter style tags to 95 pass through the parser. 96* tables: Tables using the same format as GFM 97 <https://help.github.com/articles/github-flavored-markdown#tables> and 98 PHP-Markdown Extra <https://michelf.ca/projects/php-markdown/extra/#table>. 99* toc: The returned HTML string gets a new "toc_html" attribute which is 100 a Table of Contents for the document. (experimental) 101* use-file-vars: Look for an Emacs-style markdown-extras file variable to turn 102 on Extras. 103* wiki-tables: Google Code Wiki-style tables. See 104 <http://code.google.com/p/support/wiki/WikiSyntax#Tables>. 105* wavedrom: Support for generating Wavedrom digital timing diagrams 106* xml: Passes one-liner processing instructions and namespaced XML tags. 107""" 108 109# Dev Notes: 110# - Python's regex syntax doesn't have '\z', so I'm using '\Z'. I'm 111# not yet sure if there implications with this. Compare 'pydoc sre' 112# and 'perldoc perlre'. 113 114__version_info__ = (2, 5, 1) 115__version__ = '.'.join(map(str, __version_info__)) 116__author__ = "Trent Mick" 117 118import argparse 119import codecs 120import logging 121import re 122import sys 123from collections import defaultdict, OrderedDict 124from abc import ABC, abstractmethod 125import functools 126from hashlib import sha256 127from random import randint, random 128from typing import Any, Callable, Collection, Dict, List, Literal, Optional, Tuple, Type, TypedDict, Union 129from enum import IntEnum, auto 130 131if sys.version_info[1] < 9: 132 from typing import Iterable 133else: 134 from collections.abc import Iterable 135 136# ---- type defs 137_safe_mode = Literal['replace', 'escape'] 138_extras_dict = Dict[str, Any] 139_extras_param = Union[List[str], _extras_dict] 140_link_patterns = Iterable[Tuple[re.Pattern, Union[str, Callable[[re.Match], str]]]] 141 142# ---- globals 143 144DEBUG = False 145log = logging.getLogger("markdown") 146 147DEFAULT_TAB_WIDTH = 4 148 149 150SECRET_SALT = bytes(randint(0, 1000000)) 151# MD5 function was previously used for this; the "md5" prefix was kept for 152# backwards compatibility. 153def _hash_text(s: str) -> str: 154 return 'md5-' + sha256(SECRET_SALT + s.encode("utf-8")).hexdigest()[32:] 155 156# Table of hash values for escaped characters: 157g_escape_table = dict([(ch, _hash_text(ch)) 158 for ch in '\\`*_{}[]()>#+-.!']) 159 160# Ampersand-encoding based entirely on Nat Irons's Amputator MT plugin: 161# http://bumppo.net/projects/amputator/ 162_AMPERSAND_RE = re.compile(r'&(?!#?[xX]?(?:[0-9a-fA-F]+|\w+);)') 163 164 165# ---- exceptions 166class MarkdownError(Exception): 167 pass 168 169 170# ---- public api 171 172def markdown_path( 173 path: str, 174 encoding: str = "utf-8", 175 html4tags: bool = False, 176 tab_width: int = DEFAULT_TAB_WIDTH, 177 safe_mode: Optional[_safe_mode] = None, 178 extras: Optional[_extras_param] = None, 179 link_patterns: Optional[_link_patterns] = None, 180 footnote_title: Optional[str] = None, 181 footnote_return_symbol: Optional[str] = None, 182 use_file_vars: bool = False 183) -> 'UnicodeWithAttrs': 184 fp = codecs.open(path, 'r', encoding) 185 text = fp.read() 186 fp.close() 187 return Markdown(html4tags=html4tags, tab_width=tab_width, 188 safe_mode=safe_mode, extras=extras, 189 link_patterns=link_patterns, 190 footnote_title=footnote_title, 191 footnote_return_symbol=footnote_return_symbol, 192 use_file_vars=use_file_vars).convert(text) 193 194 195def markdown( 196 text: str, 197 html4tags: bool = False, 198 tab_width: int = DEFAULT_TAB_WIDTH, 199 safe_mode: Optional[_safe_mode] = None, 200 extras: Optional[_extras_param] = None, 201 link_patterns: Optional[_link_patterns] = None, 202 footnote_title: Optional[str] = None, 203 footnote_return_symbol: Optional[str] =None, 204 use_file_vars: bool = False, 205 cli: bool = False 206) -> 'UnicodeWithAttrs': 207 return Markdown(html4tags=html4tags, tab_width=tab_width, 208 safe_mode=safe_mode, extras=extras, 209 link_patterns=link_patterns, 210 footnote_title=footnote_title, 211 footnote_return_symbol=footnote_return_symbol, 212 use_file_vars=use_file_vars, cli=cli).convert(text) 213 214 215class Stage(IntEnum): 216 PREPROCESS = auto() 217 HASH_HTML = auto() 218 LINK_DEFS = auto() 219 220 BLOCK_GAMUT = auto() 221 HEADERS = auto() 222 LISTS = auto() 223 CODE_BLOCKS = auto() 224 BLOCK_QUOTES = auto() 225 PARAGRAPHS = auto() 226 227 SPAN_GAMUT = auto() 228 CODE_SPANS = auto() 229 ESCAPE_SPECIAL = auto() 230 LINKS = auto() # and auto links 231 ITALIC_AND_BOLD = auto() 232 233 POSTPROCESS = auto() 234 UNHASH_HTML = auto() 235 236 237def mark_stage(stage: Stage): 238 ''' 239 Decorator that handles executing relevant `Extra`s before and after this `Stage` executes. 240 ''' 241 def wrapper(func): 242 @functools.wraps(func) 243 def inner(md: 'Markdown', text, *args, **kwargs): 244 md.stage = stage 245 # set "order" prop so extras can tell if they're being invoked before/after the stage 246 md.order = stage - 0.5 247 248 if stage in Extra._exec_order: 249 for klass in Extra._exec_order[stage][0]: 250 if klass.name not in md.extra_classes: 251 continue 252 extra = md.extra_classes[klass.name] 253 if extra.test(text): 254 text = extra.run(text) 255 256 md.order = stage 257 text = func(md, text, *args, **kwargs) 258 md.order = stage + 0.5 259 260 if stage in Extra._exec_order: 261 for klass in Extra._exec_order[stage][1]: 262 if klass.name not in md.extra_classes: 263 continue 264 extra = md.extra_classes[klass.name] 265 if extra.test(text): 266 text = extra.run(text) 267 268 return text 269 270 return inner 271 272 return wrapper 273 274 275class Markdown(object): 276 # The dict of "extras" to enable in processing -- a mapping of 277 # extra name to argument for the extra. Most extras do not have an 278 # argument, in which case the value is None. 279 # 280 # This can be set via (a) subclassing and (b) the constructor 281 # "extras" argument. 282 extras: _extras_dict 283 # dict of `Extra` names and associated class instances, populated during _setup_extras 284 extra_classes: Dict[str, 'Extra'] 285 286 urls: Dict[str, str] 287 titles: Dict[str, str] 288 html_blocks: Dict[str, str] 289 html_spans: Dict[str, str] 290 html_removed_text: str = "{(#HTML#)}" # placeholder removed text that does not trigger bold 291 html_removed_text_compat: str = "[HTML_REMOVED]" # for compat with markdown.py 292 safe_mode: Optional[_safe_mode] 293 294 _toc: List[Tuple[int, str, str]] 295 296 # Used to track when we're inside an ordered or unordered list 297 # (see _ProcessListItems() for details): 298 list_level = 0 299 300 stage: Stage 301 '''Current "stage" of markdown conversion taking place''' 302 order: float 303 ''' 304 Same as `Stage` but will be +/- 0.5 of the value of `Stage`. 305 This allows extras to check if they are running before or after a particular stage 306 with `if md.order < md.stage`. 307 ''' 308 309 _ws_only_line_re = re.compile(r"^[ \t]+$", re.M) 310 311 def __init__( 312 self, 313 html4tags: bool = False, 314 tab_width: int = DEFAULT_TAB_WIDTH, 315 safe_mode: Optional[_safe_mode] = None, 316 extras: Optional[_extras_param] = None, 317 link_patterns: Optional[_link_patterns] = None, 318 footnote_title: Optional[str] = None, 319 footnote_return_symbol: Optional[str] = None, 320 use_file_vars: bool = False, 321 cli: bool = False 322 ): 323 if html4tags: 324 self.empty_element_suffix = ">" 325 else: 326 self.empty_element_suffix = " />" 327 self.tab_width = tab_width 328 self.tab = tab_width * " " 329 330 # For compatibility with earlier markdown2.py and with 331 # markdown.py's safe_mode being a boolean, 332 # safe_mode == True -> "replace" 333 if safe_mode is True: 334 self.safe_mode = "replace" 335 else: 336 self.safe_mode = safe_mode 337 338 # Massaging and building the "extras" info. 339 if getattr(self, 'extras', None) is None: 340 self.extras = {} 341 elif not isinstance(self.extras, dict): 342 # inheriting classes may set `self.extras` as List[str]. 343 # we can't allow it through type hints but we can convert it 344 self.extras = dict([(e, None) for e in self.extras]) # type:ignore 345 346 if extras: 347 if not isinstance(extras, dict): 348 extras = dict([(e, None) for e in extras]) 349 self.extras.update(extras) 350 assert isinstance(self.extras, dict) 351 352 if "toc" in self.extras: 353 if "header-ids" not in self.extras: 354 self.extras["header-ids"] = None # "toc" implies "header-ids" 355 356 if self.extras["toc"] is None: 357 self._toc_depth = 6 358 else: 359 self._toc_depth = self.extras["toc"].get("depth", 6) 360 361 if 'header-ids' in self.extras: 362 if not isinstance(self.extras['header-ids'], dict): 363 self.extras['header-ids'] = { 364 'mixed': False, 365 'prefix': self.extras['header-ids'], 366 'reset-count': True 367 } 368 369 if 'break-on-newline' in self.extras: 370 self.extras.setdefault('breaks', {}) 371 self.extras['breaks']['on_newline'] = True 372 373 if 'link-patterns' in self.extras: 374 # allow link patterns via extras dict without kwarg explicitly set 375 link_patterns = link_patterns or self.extras['link-patterns'] 376 if link_patterns is None: 377 # if you have specified that the link-patterns extra SHOULD 378 # be used (via self.extras) but you haven't provided anything 379 # via the link_patterns argument then an error is raised 380 raise MarkdownError("If the 'link-patterns' extra is used, an argument for 'link_patterns' is required") 381 self.extras['link-patterns'] = link_patterns 382 383 self._instance_extras = self.extras.copy() 384 self.link_patterns = link_patterns 385 self.footnote_title = footnote_title 386 self.footnote_return_symbol = footnote_return_symbol 387 self.use_file_vars = use_file_vars 388 self._outdent_re = re.compile(r'^(\t|[ ]{1,%d})' % tab_width, re.M) 389 self.cli = cli 390 391 self._escape_table = g_escape_table.copy() 392 self._code_table = {} 393 if "smarty-pants" in self.extras: 394 self._escape_table['"'] = _hash_text('"') 395 self._escape_table["'"] = _hash_text("'") 396 397 def reset(self): 398 self.urls = {} 399 self.titles = {} 400 self.html_blocks = {} 401 self.html_spans = {} 402 self.list_level = 0 403 self.extras = self._instance_extras.copy() 404 self._setup_extras() 405 self._toc = [] 406 407 def _setup_extras(self): 408 if "footnotes" in self.extras: 409 # order of insertion matters for footnotes. Use ordered dict for Python < 3.7 410 # https://docs.python.org/3/whatsnew/3.7.html#summary-release-highlights 411 self.footnotes = OrderedDict() 412 self.footnote_ids = [] 413 if "header-ids" in self.extras: 414 if not hasattr(self, '_count_from_header_id') or self.extras['header-ids'].get('reset-count', False): 415 self._count_from_header_id = defaultdict(int) 416 if "metadata" in self.extras: 417 self.metadata: Dict[str, Any] = {} 418 419 self.extra_classes = {} 420 for name, klass in Extra._registry.items(): 421 if name not in self.extras: 422 continue 423 self.extra_classes[name] = klass(self, (self.extras.get(name, {}))) 424 425 # Per <https://developer.mozilla.org/en-US/docs/HTML/Element/a> "rel" 426 # should only be used in <a> tags with an "href" attribute. 427 428 # Opens the linked document in a new window or tab 429 # should only used in <a> tags with an "href" attribute. 430 # same with _a_nofollow 431 _a_nofollow_or_blank_links = re.compile(r""" 432 <(a) 433 ( 434 [^>]* 435 href= # href is required 436 ['"]? # HTML5 attribute values do not have to be quoted 437 [^#'"] # We don't want to match href values that start with # (like footnotes) 438 ) 439 """, 440 re.IGNORECASE | re.VERBOSE 441 ) 442 443 def convert(self, text: str) -> 'UnicodeWithAttrs': 444 """Convert the given text.""" 445 # Main function. The order in which other subs are called here is 446 # essential. Link and image substitutions need to happen before 447 # _EscapeSpecialChars(), so that any *'s or _'s in the <a> 448 # and <img> tags get encoded. 449 450 # Clear the global hashes. If we don't clear these, you get conflicts 451 # from other articles when generating a page which contains more than 452 # one article (e.g. an index page that shows the N most recent 453 # articles): 454 self.reset() 455 456 if not isinstance(text, str): 457 # TODO: perhaps shouldn't presume UTF-8 for string input? 458 text = str(text, 'utf-8') 459 460 if self.use_file_vars: 461 # Look for emacs-style file variable hints. 462 text = self._emacs_oneliner_vars_pat.sub(self._emacs_vars_oneliner_sub, text) 463 emacs_vars = self._get_emacs_vars(text) 464 if "markdown-extras" in emacs_vars: 465 splitter = re.compile("[ ,]+") 466 for e in splitter.split(emacs_vars["markdown-extras"]): 467 if '=' in e: 468 ename, earg = e.split('=', 1) 469 try: 470 earg = int(earg) 471 except ValueError: 472 pass 473 else: 474 ename, earg = e, None 475 self.extras[ename] = earg 476 477 self._setup_extras() 478 479 # Standardize line endings: 480 text = text.replace("\r\n", "\n") 481 text = text.replace("\r", "\n") 482 483 # Make sure $text ends with a couple of newlines: 484 text += "\n\n" 485 486 # Convert all tabs to spaces. 487 text = self._detab(text) 488 489 # Strip any lines consisting only of spaces and tabs. 490 # This makes subsequent regexen easier to write, because we can 491 # match consecutive blank lines with /\n+/ instead of something 492 # contorted like /[ \t]*\n+/ . 493 text = self._ws_only_line_re.sub("", text) 494 495 # strip metadata from head and extract 496 if "metadata" in self.extras: 497 text = self._extract_metadata(text) 498 499 text = self.preprocess(text) 500 501 if self.safe_mode: 502 text = self._hash_html_spans(text) 503 504 # Turn block-level HTML blocks into hash entries 505 text = self._hash_html_blocks(text, raw=True) 506 507 # Strip link definitions, store in hashes. 508 if "footnotes" in self.extras: 509 # Must do footnotes first because an unlucky footnote defn 510 # looks like a link defn: 511 # [^4]: this "looks like a link defn" 512 text = self._strip_footnote_definitions(text) 513 text = self._strip_link_definitions(text) 514 515 text = self._run_block_gamut(text) 516 517 if "footnotes" in self.extras: 518 text = self._add_footnotes(text) 519 520 text = self.postprocess(text) 521 522 text = self._unescape_special_chars(text) 523 524 if self.safe_mode: 525 text = self._unhash_html_spans(text) 526 # return the removed text warning to its markdown.py compatible form 527 text = text.replace(self.html_removed_text, self.html_removed_text_compat) 528 529 do_target_blank_links = "target-blank-links" in self.extras 530 do_nofollow_links = "nofollow" in self.extras 531 532 if do_target_blank_links and do_nofollow_links: 533 text = self._a_nofollow_or_blank_links.sub(r'<\1 rel="nofollow noopener" target="_blank"\2', text) 534 elif do_target_blank_links: 535 text = self._a_nofollow_or_blank_links.sub(r'<\1 rel="noopener" target="_blank"\2', text) 536 elif do_nofollow_links: 537 text = self._a_nofollow_or_blank_links.sub(r'<\1 rel="nofollow"\2', text) 538 539 if "toc" in self.extras and self._toc: 540 if self.extras['header-ids'].get('mixed'): 541 # TOC will only be out of order if mixed headers is enabled 542 def toc_sort(entry): 543 '''Sort the TOC by order of appearance in text''' 544 match = re.search( 545 # header tag, any attrs, the ID, any attrs, the text, close tag 546 r'^<(h%d).*?id=(["\'])%s\2.*>%s</\1>$' % (entry[0], entry[1], re.escape(entry[2])), 547 text, re.M 548 ) 549 return match.start() if match else 0 550 551 self._toc.sort(key=toc_sort) 552 self._toc_html = calculate_toc_html(self._toc) 553 554 # Prepend toc html to output 555 if self.cli or (self.extras['toc'] is not None and self.extras['toc'].get('prepend', False)): 556 text = '{}\n{}'.format(self._toc_html, text) 557 558 text += "\n" 559 560 # Attach attrs to output 561 rv = UnicodeWithAttrs(text) 562 563 if "toc" in self.extras and self._toc: 564 rv.toc_html = self._toc_html 565 566 if "metadata" in self.extras: 567 rv.metadata = self.metadata 568 return rv 569 570 @mark_stage(Stage.POSTPROCESS) 571 def postprocess(self, text: str) -> str: 572 """A hook for subclasses to do some postprocessing of the html, if 573 desired. This is called before unescaping of special chars and 574 unhashing of raw HTML spans. 575 """ 576 return text 577 578 @mark_stage(Stage.PREPROCESS) 579 def preprocess(self, text: str) -> str: 580 """A hook for subclasses to do some preprocessing of the Markdown, if 581 desired. This is called after basic formatting of the text, but prior 582 to any extras, safe mode, etc. processing. 583 """ 584 return text 585 586 # Is metadata if the content starts with optional '---'-fenced `key: value` 587 # pairs. E.g. (indented for presentation): 588 # --- 589 # foo: bar 590 # another-var: blah blah 591 # --- 592 # # header 593 # or: 594 # foo: bar 595 # another-var: blah blah 596 # 597 # # header 598 _meta_data_pattern = re.compile(r''' 599 ^{0}( # optional opening fence 600 (?: 601 {1}:(?:\n+[ \t]+.*)+ # indented lists 602 )|(?: 603 (?:{1}:\s+>(?:\n\s+.*)+?) # multiline long descriptions 604 (?=\n{1}:\s*.*\n|\s*\Z) # match up until the start of the next key:value definition or the end of the input text 605 )|(?: 606 {1}:(?! >).*\n? # simple key:value pair, leading spaces allowed 607 ) 608 ){0} # optional closing fence 609 '''.format(r'(?:---[\ \t]*\n)?', r'[\S \t]*\w[\S \t]*\s*'), re.MULTILINE | re.VERBOSE 610 ) 611 612 _key_val_list_pat = re.compile( 613 r"^-(?:[ \t]*([^\n]*)(?:[ \t]*[:-][ \t]*(\S+))?)(?:\n((?:[ \t]+[^\n]+\n?)+))?", 614 re.MULTILINE, 615 ) 616 _key_val_dict_pat = re.compile( 617 r"^([^:\n]+)[ \t]*:[ \t]*([^\n]*)(?:((?:\n[ \t]+[^\n]+)+))?", re.MULTILINE 618 ) # grp0: key, grp1: value, grp2: multiline value 619 _meta_data_fence_pattern = re.compile(r'^---[\ \t]*\n', re.MULTILINE) 620 _meta_data_newline = re.compile("^\n", re.MULTILINE) 621 622 def _extract_metadata(self, text: str) -> str: 623 if text.startswith("---"): 624 fence_splits = re.split(self._meta_data_fence_pattern, text, maxsplit=2) 625 metadata_content = fence_splits[1] 626 tail = fence_splits[2] 627 else: 628 metadata_split = re.split(self._meta_data_newline, text, maxsplit=1) 629 metadata_content = metadata_split[0] 630 tail = metadata_split[1] 631 632 # _meta_data_pattern only has one capturing group, so we can assume 633 # the returned type to be list[str] 634 match: List[str] = re.findall(self._meta_data_pattern, metadata_content) 635 if not match: 636 return text 637 638 def parse_structured_value(value: str) -> Union[List[Any], Dict[str, Any]]: 639 vs = value.lstrip() 640 vs = value.replace(v[: len(value) - len(vs)], "\n")[1:] 641 642 # List 643 if vs.startswith("-"): 644 r: List[Any] = [] 645 # the regex used has multiple capturing groups, so 646 # returned type from findall will be List[List[str]] 647 match: List[str] 648 for match in re.findall(self._key_val_list_pat, vs): 649 if match[0] and not match[1] and not match[2]: 650 r.append(match[0].strip()) 651 elif match[0] == ">" and not match[1] and match[2]: 652 r.append(match[2].strip()) 653 elif match[0] and match[1]: 654 r.append({match[0].strip(): match[1].strip()}) 655 elif not match[0] and not match[1] and match[2]: 656 r.append(parse_structured_value(match[2])) 657 else: 658 # Broken case 659 pass 660 661 return r 662 663 # Dict 664 else: 665 return { 666 match[0].strip(): ( 667 match[1].strip() 668 if match[1] 669 else parse_structured_value(match[2]) 670 ) 671 for match in re.findall(self._key_val_dict_pat, vs) 672 } 673 674 for item in match: 675 676 k, v = item.split(":", 1) 677 678 # Multiline value 679 if v[:3] == " >\n": 680 self.metadata[k.strip()] = _dedent(v[3:]).strip() 681 682 # Empty value 683 elif v == "\n": 684 self.metadata[k.strip()] = "" 685 686 # Structured value 687 elif v[0] == "\n": 688 self.metadata[k.strip()] = parse_structured_value(v) 689 690 # Simple value 691 else: 692 self.metadata[k.strip()] = v.strip() 693 694 return tail 695 696 _emacs_oneliner_vars_pat = re.compile(r"((?:<!--)?\s*-\*-)\s*(?:(\S[^\r\n]*?)([\r\n]\s*)?)?(-\*-\s*(?:-->)?)", re.UNICODE) 697 # This regular expression is intended to match blocks like this: 698 # PREFIX Local Variables: SUFFIX 699 # PREFIX mode: Tcl SUFFIX 700 # PREFIX End: SUFFIX 701 # Some notes: 702 # - "[ \t]" is used instead of "\s" to specifically exclude newlines 703 # - "(\r\n|\n|\r)" is used instead of "$" because the sre engine does 704 # not like anything other than Unix-style line terminators. 705 _emacs_local_vars_pat = re.compile(r"""^ 706 (?P<prefix>(?:[^\r\n|\n|\r])*?) 707 [\ \t]*Local\ Variables:[\ \t]* 708 (?P<suffix>.*?)(?:\r\n|\n|\r) 709 (?P<content>.*?\1End:) 710 """, re.IGNORECASE | re.MULTILINE | re.DOTALL | re.VERBOSE) 711 712 def _emacs_vars_oneliner_sub(self, match: re.Match) -> str: 713 if match.group(1).strip() == '-*-' and match.group(4).strip() == '-*-': 714 lead_ws = re.findall(r'^\s*', match.group(1))[0] 715 tail_ws = re.findall(r'\s*$', match.group(4))[0] 716 return '%s<!-- %s %s %s -->%s' % (lead_ws, '-*-', match.group(2).strip(), '-*-', tail_ws) 717 718 start, end = match.span() 719 return match.string[start: end] 720 721 def _get_emacs_vars(self, text: str) -> Dict[str, str]: 722 """Return a dictionary of emacs-style local variables. 723 724 Parsing is done loosely according to this spec (and according to 725 some in-practice deviations from this): 726 http://www.gnu.org/software/emacs/manual/html_node/emacs/Specifying-File-Variables.html#Specifying-File-Variables 727 """ 728 emacs_vars = {} 729 SIZE = pow(2, 13) # 8kB 730 731 # Search near the start for a '-*-'-style one-liner of variables. 732 head = text[:SIZE] 733 if "-*-" in head: 734 match = self._emacs_oneliner_vars_pat.search(head) 735 if match: 736 emacs_vars_str = match.group(2) 737 assert '\n' not in emacs_vars_str 738 emacs_var_strs = [s.strip() for s in emacs_vars_str.split(';') 739 if s.strip()] 740 if len(emacs_var_strs) == 1 and ':' not in emacs_var_strs[0]: 741 # While not in the spec, this form is allowed by emacs: 742 # -*- Tcl -*- 743 # where the implied "variable" is "mode". This form 744 # is only allowed if there are no other variables. 745 emacs_vars["mode"] = emacs_var_strs[0].strip() 746 else: 747 for emacs_var_str in emacs_var_strs: 748 try: 749 variable, value = emacs_var_str.strip().split(':', 1) 750 except ValueError: 751 log.debug("emacs variables error: malformed -*- " 752 "line: %r", emacs_var_str) 753 continue 754 # Lowercase the variable name because Emacs allows "Mode" 755 # or "mode" or "MoDe", etc. 756 emacs_vars[variable.lower()] = value.strip() 757 758 tail = text[-SIZE:] 759 if "Local Variables" in tail: 760 match = self._emacs_local_vars_pat.search(tail) 761 if match: 762 prefix = match.group("prefix") 763 suffix = match.group("suffix") 764 lines = match.group("content").splitlines(False) 765 # print "prefix=%r, suffix=%r, content=%r, lines: %s"\ 766 # % (prefix, suffix, match.group("content"), lines) 767 768 # Validate the Local Variables block: proper prefix and suffix 769 # usage. 770 for i, line in enumerate(lines): 771 if not line.startswith(prefix): 772 log.debug("emacs variables error: line '%s' " 773 "does not use proper prefix '%s'" 774 % (line, prefix)) 775 return {} 776 # Don't validate suffix on last line. Emacs doesn't care, 777 # neither should we. 778 if i != len(lines)-1 and not line.endswith(suffix): 779 log.debug("emacs variables error: line '%s' " 780 "does not use proper suffix '%s'" 781 % (line, suffix)) 782 return {} 783 784 # Parse out one emacs var per line. 785 continued_for = None 786 for line in lines[:-1]: # no var on the last line ("PREFIX End:") 787 if prefix: 788 line = line[len(prefix):] # strip prefix 789 if suffix: 790 line = line[:-len(suffix)] # strip suffix 791 line = line.strip() 792 if continued_for: 793 variable = continued_for 794 if line.endswith('\\'): 795 line = line[:-1].rstrip() 796 else: 797 continued_for = None 798 emacs_vars[variable] += ' ' + line 799 else: 800 try: 801 variable, value = line.split(':', 1) 802 except ValueError: 803 log.debug("local variables error: missing colon " 804 "in local variables entry: '%s'" % line) 805 continue 806 # Do NOT lowercase the variable name, because Emacs only 807 # allows "mode" (and not "Mode", "MoDe", etc.) in this block. 808 value = value.strip() 809 if value.endswith('\\'): 810 value = value[:-1].rstrip() 811 continued_for = variable 812 else: 813 continued_for = None 814 emacs_vars[variable] = value 815 816 # Unquote values. 817 for var, val in list(emacs_vars.items()): 818 if len(val) > 1 and (val.startswith('"') and val.endswith('"') 819 or val.startswith('"') and val.endswith('"')): 820 emacs_vars[var] = val[1:-1] 821 822 return emacs_vars 823 824 def _detab_line(self, line: str) -> str: 825 r"""Recusively convert tabs to spaces in a single line. 826 827 Called from _detab().""" 828 if '\t' not in line: 829 return line 830 chunk1, chunk2 = line.split('\t', 1) 831 chunk1 += (' ' * (self.tab_width - len(chunk1) % self.tab_width)) 832 output = chunk1 + chunk2 833 return self._detab_line(output) 834 835 def _detab(self, text: str) -> str: 836 r"""Iterate text line by line and convert tabs to spaces. 837 838 >>> m = Markdown() 839 >>> m._detab("\tfoo") 840 ' foo' 841 >>> m._detab(" \tfoo") 842 ' foo' 843 >>> m._detab("\t foo") 844 ' foo' 845 >>> m._detab(" foo") 846 ' foo' 847 >>> m._detab(" foo\n\tbar\tblam") 848 ' foo\n bar blam' 849 """ 850 if '\t' not in text: 851 return text 852 output = [] 853 for line in text.splitlines(): 854 output.append(self._detab_line(line)) 855 return '\n'.join(output) 856 857 # I broke out the html5 tags here and add them to _block_tags_a and 858 # _block_tags_b. This way html5 tags are easy to keep track of. 859 _html5tags = '|article|aside|header|hgroup|footer|nav|section|figure|figcaption' 860 861 _block_tags_a = 'p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|script|noscript|form|fieldset|iframe|math|ins|del|style|html|head|body' 862 _block_tags_a += _html5tags 863 864 _strict_tag_block_re = re.compile(r""" 865 ( # save in \1 866 ^ # start of line (with re.M) 867 <(%s) # start tag = \2 868 \b # word break 869 (.*\n)*? # any number of lines, minimally matching 870 </\2> # the matching end tag 871 [ \t]* # trailing spaces/tabs 872 (?=\n+|\Z) # followed by a newline or end of document 873 ) 874 """ % _block_tags_a, 875 re.X | re.M) 876 877 _block_tags_b = 'p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|script|noscript|form|fieldset|iframe|math' 878 _block_tags_b += _html5tags 879 880 _span_tags = ( 881 'a|abbr|acronym|b|bdo|big|br|button|cite|code|dfn|em|i|img|input|kbd|label|map|object|output|q' 882 '|samp|script|select|small|span|strong|sub|sup|textarea|time|tt|var' 883 ) 884 885 _liberal_tag_block_re = re.compile(r""" 886 ( # save in \1 887 ^ # start of line (with re.M) 888 <(%s) # start tag = \2 889 \b # word break 890 (.*\n)*? # any number of lines, minimally matching 891 .*</\2> # the matching end tag 892 [ \t]* # trailing spaces/tabs 893 (?=\n+|\Z) # followed by a newline or end of document 894 ) 895 """ % _block_tags_b, 896 re.X | re.M) 897 898 _html_markdown_attr_re = re.compile( 899 r'''\s+markdown=("1"|'1')''') 900 def _hash_html_block_sub( 901 self, 902 match: Union[re.Match, str], 903 raw: bool = False 904 ) -> str: 905 if isinstance(match, str): 906 html = match 907 tag = None 908 else: 909 html = match.group(1) 910 try: 911 tag = match.group(2) 912 except IndexError: 913 tag = None 914 915 if not tag: 916 m = re.match(r'.*?<(\S).*?\s*>', html) 917 # tag shouldn't be none but make the assertion for type checker 918 assert m is not None 919 tag = m.group(1) 920 921 if raw and self.safe_mode: 922 html = self._sanitize_html(html) 923 elif 'markdown-in-html' in self.extras and 'markdown=' in html: 924 first_line = html.split('\n', 1)[0] 925 m = self._html_markdown_attr_re.search(first_line) 926 if m: 927 lines = html.split('\n') 928 # if MD is on same line as opening tag then split across two lines 929 lines = list(filter(None, (re.split(r'(.*?<%s.*markdown=.*?>)' % tag, lines[0])))) + lines[1:] 930 # if MD on same line as closing tag, split across two lines 931 lines = lines[:-1] + list(filter(None, re.split(r'(\s*?</%s>.*?$)' % tag, lines[-1]))) 932 # extract key sections of the match 933 first_line = lines[0] 934 middle = '\n'.join(lines[1:-1]) 935 last_line = lines[-1] 936 # remove `markdown="1"` attr from tag 937 first_line = first_line[:m.start()] + first_line[m.end():] 938 # hash the HTML segments to protect them 939 f_key = _hash_text(first_line) 940 self.html_blocks[f_key] = first_line 941 l_key = _hash_text(last_line) 942 self.html_blocks[l_key] = last_line 943 return ''.join(["\n\n", f_key, 944 "\n\n", middle, "\n\n", 945 l_key, "\n\n"]) 946 elif self.extras.get('header-ids', {}).get('mixed') and self._h_tag_re.match(html): 947 html = self._h_tag_re.sub(self._h_tag_sub, html) 948 key = _hash_text(html) 949 self.html_blocks[key] = html 950 return "\n\n" + key + "\n\n" 951 952 @mark_stage(Stage.HASH_HTML) 953 def _hash_html_blocks(self, text: str, raw: bool = False) -> str: 954 """Hashify HTML blocks 955 956 We only want to do this for block-level HTML tags, such as headers, 957 lists, and tables. That's because we still want to wrap <p>s around 958 "paragraphs" that are wrapped in non-block-level tags, such as anchors, 959 phrase emphasis, and spans. The list of tags we're looking for is 960 hard-coded. 961 962 @param raw {boolean} indicates if these are raw HTML blocks in 963 the original source. It makes a difference in "safe" mode. 964 """ 965 if '<' not in text: 966 return text 967 968 # Pass `raw` value into our calls to self._hash_html_block_sub. 969 hash_html_block_sub = _curry(self._hash_html_block_sub, raw=raw) 970 971 # First, look for nested blocks, e.g.: 972 # <div> 973 # <div> 974 # tags for inner block must be indented. 975 # </div> 976 # </div> 977 # 978 # The outermost tags must start at the left margin for this to match, and 979 # the inner nested divs must be indented. 980 # We need to do this before the next, more liberal match, because the next 981 # match will start at the first `<div>` and stop at the first `</div>`. 982 text = self._strict_tag_block_sub(text, self._block_tags_a, hash_html_block_sub) 983 984 # Now match more liberally, simply from `\n<tag>` to `</tag>\n` 985 text = self._liberal_tag_block_re.sub(hash_html_block_sub, text) 986 987 # now do the same for spans that are acting like blocks 988 # eg: an anchor split over multiple lines for readability 989 text = self._strict_tag_block_sub( 990 text, self._span_tags, 991 # inline elements can't contain block level elements, so only span gamut is required 992 lambda t: hash_html_block_sub(self._run_span_gamut(t)) 993 ) 994 995 # Special case just for <hr />. It was easier to make a special 996 # case than to make the other regex more complicated. 997 if "<hr" in text: 998 _hr_tag_re = _hr_tag_re_from_tab_width(self.tab_width) 999 text = _hr_tag_re.sub(hash_html_block_sub, text) 1000 1001 # Special case for standalone HTML comments: 1002 if "<!--" in text: 1003 start = 0 1004 while True: 1005 # Delimiters for next comment block. 1006 try: 1007 start_idx = text.index("<!--", start) 1008 except ValueError: 1009 break 1010 try: 1011 end_idx = text.index("-->", start_idx) + 3 1012 except ValueError: 1013 break 1014 1015 # Start position for next comment block search. 1016 start = end_idx 1017 1018 # Validate whitespace before comment. 1019 if start_idx: 1020 # - Up to `tab_width - 1` spaces before start_idx. 1021 for i in range(self.tab_width - 1): 1022 if text[start_idx - 1] != ' ': 1023 break 1024 start_idx -= 1 1025 if start_idx == 0: 1026 break 1027 # - Must be preceded by 2 newlines or hit the start of 1028 # the document. 1029 if start_idx == 0: 1030 pass 1031 elif start_idx == 1 and text[0] == '\n': 1032 start_idx = 0 # to match minute detail of Markdown.pl regex 1033 elif text[start_idx-2:start_idx] == '\n\n': 1034 pass 1035 else: 1036 break 1037 1038 # Validate whitespace after comment. 1039 # - Any number of spaces and tabs. 1040 while end_idx < len(text): 1041 if text[end_idx] not in ' \t': 1042 break 1043 end_idx += 1 1044 # - Must be following by 2 newlines or hit end of text. 1045 if text[end_idx:end_idx+2] not in ('', '\n', '\n\n'): 1046 continue 1047 1048 # Escape and hash (must match `_hash_html_block_sub`). 1049 html = text[start_idx:end_idx] 1050 if raw and self.safe_mode: 1051 html = self._sanitize_html(html) 1052 key = _hash_text(html) 1053 self.html_blocks[key] = html 1054 text = text[:start_idx] + "\n\n" + key + "\n\n" + text[end_idx:] 1055 1056 if "xml" in self.extras: 1057 # Treat XML processing instructions and namespaced one-liner 1058 # tags as if they were block HTML tags. E.g., if standalone 1059 # (i.e. are their own paragraph), the following do not get 1060 # wrapped in a <p> tag: 1061 # <?foo bar?> 1062 # 1063 # <xi:include xmlns:xi="http://www.w3.org/2001/XInclude" href="chapter_1.md"/> 1064 _xml_oneliner_re = _xml_oneliner_re_from_tab_width(self.tab_width) 1065 text = _xml_oneliner_re.sub(hash_html_block_sub, text) 1066 1067 return text 1068 1069 def _strict_tag_block_sub( 1070 self, 1071 text: str, 1072 html_tags_re: str, 1073 callback: Callable[[str], str], 1074 allow_indent: bool = False 1075 ) -> str: 1076 ''' 1077 Finds and substitutes HTML blocks within blocks of text 1078 1079 Args: 1080 text: the text to search 1081 html_tags_re: a regex pattern of HTML block tags to match against. 1082 For example, `Markdown._block_tags_a` 1083 callback: callback function that receives the found HTML text block and returns a new str 1084 allow_indent: allow matching HTML blocks that are not completely outdented 1085 ''' 1086 tag_count = 0 1087 current_tag = html_tags_re 1088 block = '' 1089 result = '' 1090 1091 for chunk in text.splitlines(True): 1092 is_markup = re.match( 1093 r'^(\s{0,%s})(?:</code>(?=</pre>))?(</?(%s)\b>?)' % ('' if allow_indent else '0', current_tag), chunk 1094 ) 1095 block += chunk 1096 1097 if is_markup: 1098 if chunk.startswith('%s</' % is_markup.group(1)): 1099 tag_count -= 1 1100 else: 1101 # if close tag is in same line 1102 if self._tag_is_closed(is_markup.group(3), chunk): 1103 # we must ignore these 1104 is_markup = None 1105 else: 1106 tag_count += 1 1107 current_tag = is_markup.group(3) 1108 1109 if tag_count == 0: 1110 if is_markup: 1111 block = callback(block.rstrip('\n')) # remove trailing newline 1112 current_tag = html_tags_re 1113 result += block 1114 block = '' 1115 1116 result += block 1117 1118 return result 1119 1120 def _tag_is_closed(self, tag_name: str, text: str) -> bool: 1121 # super basic check if number of open tags == number of closing tags 1122 return len(re.findall('<%s(?:.*?)>' % tag_name, text)) == len(re.findall('</%s>' % tag_name, text)) 1123 1124 @mark_stage(Stage.LINK_DEFS) 1125 def _strip_link_definitions(self, text: str) -> str: 1126 # Strips link definitions from text, stores the URLs and titles in 1127 # hash references. 1128 less_than_tab = self.tab_width - 1 1129 1130 # Link defs are in the form: 1131 # [id]: url "optional title" 1132 _link_def_re = re.compile(r""" 1133 ^[ ]{0,%d}\[(.+)\]: # id = \1 1134 [ \t]* 1135 \n? # maybe *one* newline 1136 [ \t]* 1137 <?(.+?)>? # url = \2 1138 [ \t]* 1139 (?: 1140 \n? # maybe one newline 1141 [ \t]* 1142 (?<=\s) # lookbehind for whitespace 1143 ['"(] 1144 ([^\n]*) # title = \3 1145 ['")] 1146 [ \t]* 1147 )? # title is optional 1148 (?:\n+|\Z) 1149 """ % less_than_tab, re.X | re.M | re.U) 1150 return _link_def_re.sub(self._extract_link_def_sub, text) 1151 1152 def _extract_link_def_sub(self, match: re.Match) -> str: 1153 id, url, title = match.groups() 1154 key = id.lower() # Link IDs are case-insensitive 1155 self.urls[key] = self._encode_amps_and_angles(url) 1156 if title: 1157 self.titles[key] = title 1158 return "" 1159 1160 def _extract_footnote_def_sub(self, match: re.Match) -> str: 1161 id, text = match.groups() 1162 text = _dedent(text, skip_first_line=not text.startswith('\n')).strip() 1163 normed_id = re.sub(r'\W', '-', id) 1164 # Ensure footnote text ends with a couple newlines (for some 1165 # block gamut matches). 1166 self.footnotes[normed_id] = text + "\n\n" 1167 return "" 1168 1169 def _strip_footnote_definitions(self, text: str) -> str: 1170 """A footnote definition looks like this: 1171 1172 [^note-id]: Text of the note. 1173 1174 May include one or more indented paragraphs. 1175 1176 Where, 1177 - The 'note-id' can be pretty much anything, though typically it 1178 is the number of the footnote. 1179 - The first paragraph may start on the next line, like so: 1180 1181 [^note-id]: 1182 Text of the note. 1183 """ 1184 less_than_tab = self.tab_width - 1 1185 footnote_def_re = re.compile(r''' 1186 ^[ ]{0,%d}\[\^(.+)\]: # id = \1 1187 [ \t]* 1188 ( # footnote text = \2 1189 # First line need not start with the spaces. 1190 (?:\s*.*\n+) 1191 (?: 1192 (?:[ ]{%d} | \t) # Subsequent lines must be indented. 1193 .*\n+ 1194 )* 1195 ) 1196 # Lookahead for non-space at line-start, or end of doc. 1197 (?:(?=^[ ]{0,%d}\S)|\Z) 1198 ''' % (less_than_tab, self.tab_width, self.tab_width), 1199 re.X | re.M) 1200 return footnote_def_re.sub(self._extract_footnote_def_sub, text) 1201 1202 _hr_re = re.compile(r'^[ ]{0,3}([-_*])[ ]{0,2}(\1[ ]{0,2}){2,}$', re.M) 1203 1204 @mark_stage(Stage.BLOCK_GAMUT) 1205 def _run_block_gamut(self, text: str) -> str: 1206 # These are all the transformations that form block-level 1207 # tags like paragraphs, headers, and list items. 1208 1209 text = self._do_headers(text) 1210 1211 # Do Horizontal Rules: 1212 # On the number of spaces in horizontal rules: The spec is fuzzy: "If 1213 # you wish, you may use spaces between the hyphens or asterisks." 1214 # Markdown.pl 1.0.1's hr regexes limit the number of spaces between the 1215 # hr chars to one or two. We'll reproduce that limit here. 1216 hr = "\n<hr"+self.empty_element_suffix+"\n" 1217 text = re.sub(self._hr_re, hr, text) 1218 1219 text = self._do_lists(text) 1220 1221 text = self._do_code_blocks(text) 1222 1223 text = self._do_block_quotes(text) 1224 1225 # We already ran _HashHTMLBlocks() before, in Markdown(), but that 1226 # was to escape raw HTML in the original Markdown source. This time, 1227 # we're escaping the markup we've just created, so that we don't wrap 1228 # <p> tags around block-level tags. 1229 text = self._hash_html_blocks(text) 1230 1231 text = self._form_paragraphs(text) 1232 1233 return text 1234 1235 @mark_stage(Stage.SPAN_GAMUT) 1236 def _run_span_gamut(self, text: str) -> str: 1237 # These are all the transformations that occur *within* block-level 1238 # tags like paragraphs, headers, and list items. 1239 1240 text = self._do_code_spans(text) 1241 1242 text = self._escape_special_chars(text) 1243 1244 # Process anchor and image tags. 1245 text = self._do_links(text) 1246 1247 # Make links out of things like `<http://example.com/>` 1248 # Must come after _do_links(), because you can use < and > 1249 # delimiters in inline links like [this](<url>). 1250 text = self._do_auto_links(text) 1251 1252 text = self._encode_amps_and_angles(text) 1253 1254 text = self._do_italics_and_bold(text) 1255 1256 # Do hard breaks 1257 text = re.sub(r" {2,}\n(?!\<(?:\/?(ul|ol|li))\>)", "<br%s\n" % self.empty_element_suffix, text) 1258 1259 return text 1260 1261 # "Sorta" because auto-links are identified as "tag" tokens. 1262 _sorta_html_tokenize_re = re.compile(r""" 1263 ( 1264 \\* # escapes 1265 (?: 1266 # tag 1267 </? 1268 (?:\w+) # tag name 1269 (?:\s+(?:[\w-]+:)?[\w-]+=(?:".*?"|'.*?'))* # attributes 1270 \s*/?> 1271 | 1272 # auto-link (e.g., <http://www.activestate.com/>) 1273 <[\w~:/?#\[\]@!$&'\(\)*+,;%=\.\\-]+> 1274 | 1275 <!--.*?--> # comment 1276 | 1277 <\?.*?\?> # processing instruction 1278 ) 1279 ) 1280 """, re.X) 1281 1282 @mark_stage(Stage.ESCAPE_SPECIAL) 1283 def _escape_special_chars(self, text: str) -> str: 1284 # Python markdown note: the HTML tokenization here differs from 1285 # that in Markdown.pl, hence the behaviour for subtle cases can 1286 # differ (I believe the tokenizer here does a better job because 1287 # it isn't susceptible to unmatched '<' and '>' in HTML tags). 1288 # Note, however, that '>' is not allowed in an auto-link URL 1289 # here. 1290 lead_escape_re = re.compile(r'^((?:\\\\)*(?!\\))') 1291 escaped = [] 1292 is_html_markup = False 1293 for token in self._sorta_html_tokenize_re.split(text): 1294 # check token is preceded by 0 or more PAIRS of escapes, because escape pairs 1295 # escape themselves and don't affect the token 1296 if is_html_markup and lead_escape_re.match(token): 1297 # Within tags/HTML-comments/auto-links, encode * and _ 1298 # so they don't conflict with their use in Markdown for 1299 # italics and strong. We're replacing each such 1300 # character with its corresponding MD5 checksum value; 1301 # this is likely overkill, but it should prevent us from 1302 # colliding with the escape values by accident. 1303 escape_seq, token = lead_escape_re.split(token)[1:] or ('', token) 1304 escaped.append( 1305 escape_seq.replace('\\\\', self._escape_table['\\']) 1306 + token.replace('*', self._escape_table['*']) 1307 .replace('_', self._escape_table['_']) 1308 ) 1309 else: 1310 escaped.append(self._encode_backslash_escapes(token.replace('\\<', '<'))) 1311 is_html_markup = not is_html_markup 1312 return ''.join(escaped) 1313 1314 @mark_stage(Stage.HASH_HTML) 1315 def _hash_html_spans(self, text: str) -> str: 1316 # Used for safe_mode. 1317 1318 def _is_auto_link(s): 1319 if ':' in s and self._auto_link_re.match(s): 1320 return True 1321 elif '@' in s and self._auto_email_link_re.match(s): 1322 return True 1323 return False 1324 1325 def _is_code_span(index, token): 1326 try: 1327 if token == '<code>': 1328 peek_tokens = split_tokens[index: index + 3] 1329 elif token == '</code>': 1330 peek_tokens = split_tokens[index - 2: index + 1] 1331 else: 1332 return False 1333 except IndexError: 1334 return False 1335 1336 return re.match(r'<code>md5-[A-Fa-f0-9]{32}</code>', ''.join(peek_tokens)) 1337 1338 def _is_comment(token): 1339 if self.safe_mode == 'replace': 1340 # don't bother processing each section of comment in replace mode. Just do the whole thing 1341 return 1342 return re.match(r'(<!--)(.*)(-->)', token) 1343 1344 def _hash(token): 1345 key = _hash_text(token) 1346 self.html_spans[key] = token 1347 return key 1348 1349 tokens = [] 1350 split_tokens = self._sorta_html_tokenize_re.split(text) 1351 is_html_markup = False 1352 for index, token in enumerate(split_tokens): 1353 if is_html_markup and not _is_auto_link(token) and not _is_code_span(index, token): 1354 is_comment = _is_comment(token) 1355 if is_comment: 1356 tokens.append(_hash(self._sanitize_html(is_comment.group(1)))) 1357 # sanitise but leave comment body intact for further markdown processing 1358 tokens.append(self._sanitize_html(is_comment.group(2))) 1359 tokens.append(_hash(self._sanitize_html(is_comment.group(3)))) 1360 else: 1361 tokens.append(_hash(self._sanitize_html(token))) 1362 else: 1363 tokens.append(self._encode_incomplete_tags(token)) 1364 is_html_markup = not is_html_markup 1365 return ''.join(tokens) 1366 1367 def _unhash_html_spans(self, text: str) -> str: 1368 for key, sanitized in list(self.html_spans.items()): 1369 text = text.replace(key, sanitized) 1370 return text 1371 1372 def _sanitize_html(self, s: str) -> str: 1373 if self.safe_mode == "replace": 1374 return self.html_removed_text 1375 elif self.safe_mode == "escape": 1376 replacements = [ 1377 ('&', '&'), 1378 ('<', '<'), 1379 ('>', '>'), 1380 ] 1381 for before, after in replacements: 1382 s = s.replace(before, after) 1383 return s 1384 else: 1385 raise MarkdownError("invalid value for 'safe_mode': %r (must be " 1386 "'escape' or 'replace')" % self.safe_mode) 1387 1388 _inline_link_title = re.compile(r''' 1389 ( # \1 1390 [ \t]+ 1391 (['"]) # quote char = \2 1392 (?P<title>.*?) 1393 \2 1394 )? # title is optional 1395 \)$ 1396 ''', re.X | re.S) 1397 _tail_of_reference_link_re = re.compile(r''' 1398 # Match tail of: [text][id] 1399 [ ]? # one optional space 1400 (?:\n[ ]*)? # one optional newline followed by spaces 1401 \[ 1402 (?P<id>.*?) 1403 \] 1404 ''', re.X | re.S) 1405 1406 _whitespace = re.compile(r'\s*') 1407 1408 _strip_anglebrackets = re.compile(r'<(.*)>.*') 1409 1410 def _find_non_whitespace(self, text: str, start: int) -> int: 1411 """Returns the index of the first non-whitespace character in text 1412 after (and including) start 1413 """ 1414 match = self._whitespace.match(text, start) 1415 return match.end() if match else len(text) 1416 1417 def _find_balanced(self, text: str, start: int, open_c: str, close_c: str) -> int: 1418 """Returns the index where the open_c and close_c characters balance 1419 out - the same number of open_c and close_c are encountered - or the 1420 end of string if it's reached before the balance point is found. 1421 """ 1422 i = start 1423 l = len(text) 1424 count = 1 1425 while count > 0 and i < l: 1426 if text[i] == open_c: 1427 count += 1 1428 elif text[i] == close_c: 1429 count -= 1 1430 i += 1 1431 return i 1432 1433 def _extract_url_and_title(self, text: str, start: int) -> Union[Tuple[str, str, int], Tuple[None, None, None]]: 1434 """Extracts the url and (optional) title from the tail of a link""" 1435 # text[start] equals the opening parenthesis 1436 idx = self._find_non_whitespace(text, start+1) 1437 if idx == len(text): 1438 return None, None, None 1439 end_idx = idx 1440 has_anglebrackets = text[idx] == "<" 1441 if has_anglebrackets: 1442 end_idx = self._find_balanced(text, end_idx+1, "<", ">") 1443 end_idx = self._find_balanced(text, end_idx, "(", ")") 1444 match = self._inline_link_title.search(text, idx, end_idx) 1445 if not match: 1446 return None, None, None 1447 url, title = text[idx:match.start()], match.group("title") 1448 if has_anglebrackets: 1449 url = self._strip_anglebrackets.sub(r'\1', url) 1450 return url, title, end_idx 1451 1452 # https://developer.mozilla.org/en-US/docs/web/http/basics_of_http/data_urls 1453 # https://developer.mozilla.org/en-US/docs/Web/HTTP/Basics_of_HTTP/MIME_types 1454 _data_url_re = re.compile(r''' 1455 data: 1456 # in format type/subtype;parameter=optional 1457 (?P<mime>\w+/[\w+\.-]+(?:;\w+=[\w+\.-]+)?)? 1458 # optional base64 token 1459 (?P<token>;base64)? 1460 ,(?P<data>.*) 1461 ''', re.X) 1462 1463 def _protect_url(self, url: str) -> str: 1464 ''' 1465 Function that passes a URL through `_html_escape_url` to remove any nasty characters, 1466 and then hashes the now "safe" URL to prevent other safety mechanisms from tampering 1467 with it (eg: escaping "&" in URL parameters) 1468 ''' 1469 data_url = self._data_url_re.match(url) 1470 charset = None 1471 if data_url is not None: 1472 mime = data_url.group('mime') or '' 1473 if mime.startswith('image/') and data_url.group('token') == ';base64': 1474 charset='base64' 1475 url = _html_escape_url(url, safe_mode=self.safe_mode, charset=charset) 1476 key = _hash_text(url) 1477 self._escape_table[url] = key 1478 return key 1479 1480 _safe_protocols = r'(?:https?|ftp):\/\/|(?:mailto|tel):' 1481 1482 @property 1483 def _safe_href(self): 1484 ''' 1485 _safe_href is adapted from pagedown's Markdown.Sanitizer.js 1486 From: https://github.com/StackExchange/pagedown/blob/master/LICENSE.txt 1487 Original Showdown code copyright (c) 2007 John Fraser 1488 Modifications and bugfixes (c) 2009 Dana Robinson 1489 Modifications and bugfixes (c) 2009-2014 Stack Exchange Inc. 1490 ''' 1491 safe = r'-\w' 1492 # omitted ['"<>] for XSS reasons 1493 less_safe = r'#/\.!#$%&\(\)\+,/:;=\?@\[\]^`\{\}\|~' 1494 # dot seperated hostname, optional port number, not followed by protocol seperator 1495 domain = r'(?:[%s]+(?:\.[%s]+)*)(?:(?<!tel):\d+/?)?(?![^:/]*:/*)' % (safe, safe) 1496 fragment = r'[%s]*' % (safe + less_safe) 1497 1498 return re.compile(r'^(?:(%s)?(%s)(%s)|(#|\.{,2}/)(%s))$' % (self._safe_protocols, domain, fragment, fragment), re.I) 1499 1500 @mark_stage(Stage.LINKS) 1501 def _do_links(self, text: str) -> str: 1502 """Turn Markdown link shortcuts into XHTML <a> and <img> tags. 1503 1504 This is a combination of Markdown.pl's _DoAnchors() and 1505 _DoImages(). They are done together because that simplified the 1506 approach. It was necessary to use a different approach than 1507 Markdown.pl because of the lack of atomic matching support in 1508 Python's regex engine used in $g_nested_brackets. 1509 """ 1510 MAX_LINK_TEXT_SENTINEL = 3000 # markdown2 issue 24 1511 1512 # `anchor_allowed_pos` is used to support img links inside 1513 # anchors, but not anchors inside anchors. An anchor's start 1514 # pos must be `>= anchor_allowed_pos`. 1515 anchor_allowed_pos = 0 1516 1517 curr_pos = 0 1518 while True: # Handle the next link. 1519 # The next '[' is the start of: 1520 # - an inline anchor: [text](url "title") 1521 # - a reference anchor: [text][id] 1522 # - an inline img:  1523 # - a reference img: ![text][id] 1524 # - a footnote ref: [^id] 1525 # (Only if 'footnotes' extra enabled) 1526 # - a footnote defn: [^id]: ... 1527 # (Only if 'footnotes' extra enabled) These have already 1528 # been stripped in _strip_footnote_definitions() so no 1529 # need to watch for them. 1530 # - a link definition: [id]: url "title" 1531 # These have already been stripped in 1532 # _strip_link_definitions() so no need to watch for them. 1533 # - not markup: [...anything else... 1534 try: 1535 start_idx = text.index('[', curr_pos) 1536 except ValueError: 1537 break 1538 text_length = len(text) 1539 1540 # Find the matching closing ']'. 1541 # Markdown.pl allows *matching* brackets in link text so we 1542 # will here too. Markdown.pl *doesn't* currently allow 1543 # matching brackets in img alt text -- we'll differ in that 1544 # regard. 1545 bracket_depth = 0 1546 for p in range(start_idx+1, min(start_idx+MAX_LINK_TEXT_SENTINEL, 1547 text_length)): 1548 ch = text[p] 1549 if ch == ']': 1550 bracket_depth -= 1 1551 if bracket_depth < 0: 1552 break 1553 elif ch == '[': 1554 bracket_depth += 1 1555 else: 1556 # Closing bracket not found within sentinel length. 1557 # This isn't markup. 1558 curr_pos = start_idx + 1 1559 continue 1560 link_text = text[start_idx+1:p] 1561 1562 # Fix for issue 341 - Injecting XSS into link text 1563 if self.safe_mode: 1564 link_text = self._hash_html_spans(link_text) 1565 link_text = self._unhash_html_spans(link_text) 1566 1567 # Possibly a footnote ref? 1568 if "footnotes" in self.extras and link_text.startswith("^"): 1569 normed_id = re.sub(r'\W', '-', link_text[1:]) 1570 if normed_id in self.footnotes: 1571 self.footnote_ids.append(normed_id) 1572 result = '<sup class="footnote-ref" id="fnref-%s">' \ 1573 '<a href="#fn-%s">%s</a></sup>' \ 1574 % (normed_id, normed_id, len(self.footnote_ids)) 1575 text = text[:start_idx] + result + text[p+1:] 1576 else: 1577 # This id isn't defined, leave the markup alone. 1578 curr_pos = p+1 1579 continue 1580 1581 # Now determine what this is by the remainder. 1582 p += 1 1583 1584 # Inline anchor or img? 1585 if text[p:p + 1] == '(': # attempt at perf improvement 1586 url, title, url_end_idx = self._extract_url_and_title(text, p) 1587 if url is not None: 1588 # Handle an inline anchor or img. 1589 is_img = start_idx > 0 and text[start_idx-1] == "!" 1590 if is_img: 1591 start_idx -= 1 1592 1593 # We've got to encode these to avoid conflicting 1594 # with italics/bold. 1595 url = url.replace('*', self._escape_table['*']) \ 1596 .replace('_', self._escape_table['_']) 1597 if title: 1598 title_str = ' title="%s"' % ( 1599 _xml_escape_attr(title) 1600 .replace('*', self._escape_table['*']) 1601 .replace('_', self._escape_table['_'])) 1602 else: 1603 title_str = '' 1604 if is_img: 1605 img_class_str = self._html_class_str_from_tag("img") 1606 result = '<img src="%s" alt="%s"%s%s%s' \ 1607 % (self._protect_url(url), 1608 _xml_escape_attr(link_text), 1609 title_str, 1610 img_class_str, 1611 self.empty_element_suffix) 1612 if "smarty-pants" in self.extras: 1613 result = result.replace('"', self._escape_table['"']) 1614 curr_pos = start_idx + len(result) 1615 anchor_allowed_pos = start_idx + len(result) 1616 text = text[:start_idx] + result + text[url_end_idx:] 1617 elif start_idx >= anchor_allowed_pos: 1618 safe_link = self._safe_href.match(url) 1619 if self.safe_mode and not safe_link: 1620 result_head = '<a href="#"%s>' % (title_str) 1621 else: 1622 result_head = '<a href="%s"%s>' % (self._protect_url(url), title_str) 1623 result = '%s%s</a>' % (result_head, link_text) 1624 if "smarty-pants" in self.extras: 1625 result = result.replace('"', self._escape_table['"']) 1626 # <img> allowed from curr_pos on, <a> from 1627 # anchor_allowed_pos on. 1628 curr_pos = start_idx + len(result_head) 1629 anchor_allowed_pos = start_idx + len(result) 1630 text = text[:start_idx] + result + text[url_end_idx:] 1631 else: 1632 # Anchor not allowed here. 1633 curr_pos = start_idx + 1 1634 continue 1635 1636 # Reference anchor or img? 1637 else: 1638 match = self._tail_of_reference_link_re.match(text, p) 1639 if match: 1640 # Handle a reference-style anchor or img. 1641 is_img = start_idx > 0 and text[start_idx-1] == "!" 1642 if is_img: 1643 start_idx -= 1 1644 link_id = match.group("id").lower() 1645 if not link_id: 1646 link_id = link_text.lower() # for links like [this][] 1647 if link_id in self.urls: 1648 url = self.urls[link_id] 1649 # We've got to encode these to avoid conflicting 1650 # with italics/bold. 1651 url = url.replace('*', self._escape_table['*']) \ 1652 .replace('_', self._escape_table['_']) 1653 title = self.titles.get(link_id) 1654 if title: 1655 title = _xml_escape_attr(title) \ 1656 .replace('*', self._escape_table['*']) \ 1657 .replace('_', self._escape_table['_']) 1658 title_str = ' title="%s"' % title 1659 else: 1660 title_str = '' 1661 if is_img: 1662 img_class_str = self._html_class_str_from_tag("img") 1663 result = '<img src="%s" alt="%s"%s%s%s' \ 1664 % (self._protect_url(url), 1665 _xml_escape_attr(link_text), 1666 title_str, 1667 img_class_str, 1668 self.empty_element_suffix) 1669 if "smarty-pants" in self.extras: 1670 result = result.replace('"', self._escape_table['"']) 1671 curr_pos = start_idx + len(result) 1672 text = text[:start_idx] + result + text[match.end():] 1673 elif start_idx >= anchor_allowed_pos: 1674 if self.safe_mode and not self._safe_href.match(url): 1675 result_head = '<a href="#"%s>' % (title_str) 1676 else: 1677 result_head = '<a href="%s"%s>' % (self._protect_url(url), title_str) 1678 result = '%s%s</a>' % (result_head, link_text) 1679 if "smarty-pants" in self.extras: 1680 result = result.replace('"', self._escape_table['"']) 1681 # <img> allowed from curr_pos on, <a> from 1682 # anchor_allowed_pos on. 1683 curr_pos = start_idx + len(result_head) 1684 anchor_allowed_pos = start_idx + len(result) 1685 text = text[:start_idx] + result + text[match.end():] 1686 else: 1687 # Anchor not allowed here. 1688 curr_pos = start_idx + 1 1689 else: 1690 # This id isn't defined, leave the markup alone. 1691 # set current pos to end of link title and continue from there 1692 curr_pos = p 1693 continue 1694 1695 # Otherwise, it isn't markup. 1696 curr_pos = start_idx + 1 1697 1698 return text 1699 1700 def header_id_from_text(self, 1701 text: str, 1702 prefix: str, 1703 n: Optional[int] = None 1704 ) -> str: 1705 """Generate a header id attribute value from the given header 1706 HTML content. 1707 1708 This is only called if the "header-ids" extra is enabled. 1709 Subclasses may override this for different header ids. 1710 1711 @param text {str} The text of the header tag 1712 @param prefix {str} The requested prefix for header ids. This is the 1713 value of the "header-ids" extra key, if any. Otherwise, None. 1714 @param n {int} (unused) The <hN> tag number, i.e. `1` for an <h1> tag. 1715 @returns {str} The value for the header tag's "id" attribute. Return 1716 None to not have an id attribute and to exclude this header from 1717 the TOC (if the "toc" extra is specified). 1718 """ 1719 header_id = _slugify(text) 1720 if prefix and isinstance(prefix, str): 1721 header_id = prefix + '-' + header_id 1722 1723 self._count_from_header_id[header_id] += 1 1724 if 0 == len(header_id) or self._count_from_header_id[header_id] > 1: 1725 header_id += '-%s' % self._count_from_header_id[header_id] 1726 1727 return header_id 1728 1729 def _header_id_exists(self, text: str) -> bool: 1730 header_id = _slugify(text) 1731 prefix = self.extras['header-ids'].get('prefix') 1732 if prefix and isinstance(prefix, str): 1733 header_id = prefix + '-' + header_id 1734 return header_id in self._count_from_header_id or header_id in map(lambda x: x[1], self._toc) 1735 1736 def _toc_add_entry(self, level: int, id: str, name: str) -> None: 1737 if level > self._toc_depth: 1738 return 1739 if self._toc is None: 1740 self._toc = [] 1741 self._toc.append((level, id, self._unescape_special_chars(name))) 1742 1743 _h_re_base = r''' 1744 (^(.+)[ \t]{0,99}\n(=+|-+)[ \t]*\n+) 1745 | 1746 (^(\#{1,6}) # \1 = string of #'s 1747 [ \t]%s 1748 (.+?) # \2 = Header text 1749 [ \t]{0,99} 1750 (?<!\\) # ensure not an escaped trailing '#' 1751 \#* # optional closing #'s (not counted) 1752 \n+ 1753 ) 1754 ''' 1755 1756 _h_re = re.compile(_h_re_base % '*', re.X | re.M) 1757 _h_re_tag_friendly = re.compile(_h_re_base % '+', re.X | re.M) 1758 1759 def _h_sub(self, match: re.Match) -> str: 1760 '''Handles processing markdown headers''' 1761 if match.group(1) is not None and match.group(3) == "-": 1762 return match.group(1) 1763 elif match.group(1) is not None: 1764 # Setext header 1765 n = {"=": 1, "-": 2}[match.group(3)[0]] 1766 header_group = match.group(2) 1767 else: 1768 # atx header 1769 n = len(match.group(5)) 1770 header_group = match.group(6) 1771 1772 demote_headers = self.extras.get("demote-headers") 1773 if demote_headers: 1774 n = min(n + demote_headers, 6) 1775 header_id_attr = "" 1776 if "header-ids" in self.extras: 1777 header_id = self.header_id_from_text(header_group, 1778 self.extras["header-ids"].get('prefix'), n) 1779 if header_id: 1780 header_id_attr = ' id="%s"' % header_id 1781 html = self._run_span_gamut(header_group) 1782 if "toc" in self.extras and header_id: 1783 self._toc_add_entry(n, header_id, html) 1784 return "<h%d%s>%s</h%d>\n\n" % (n, header_id_attr, html, n) 1785 1786 _h_tag_re = re.compile(r''' 1787 ^<h([1-6])(.*)> # \1 tag num, \2 attrs 1788 (.*) # \3 text 1789 </h\1> 1790 ''', re.X | re.M) 1791 1792 def _h_tag_sub(self, match: re.Match) -> str: 1793 '''Different to `_h_sub` in that this function handles existing HTML headers''' 1794 text = match.string[match.start(): match.end()] 1795 h_level = int(match.group(1)) 1796 # extract id= attr from tag, trying to account for regex "misses" 1797 id_attr = (re.match(r'.*?id=(\S+)?.*', match.group(2) or '') or '') 1798 if id_attr: 1799 # if id attr exists, extract that 1800 id_attr = id_attr.group(1) or '' 1801 id_attr = id_attr.strip('\'" ') 1802 h_text = match.group(3) 1803 1804 # check if header was already processed (ie: was a markdown header rather than HTML) 1805 if id_attr and self._header_id_exists(id_attr): 1806 return text 1807 1808 # generate new header id if none existed 1809 header_id = id_attr or self.header_id_from_text(h_text, self.extras['header-ids'].get('prefix'), h_level) 1810 if "toc" in self.extras: 1811 self._toc_add_entry(h_level, header_id, h_text) 1812 if header_id and not id_attr: 1813 # '<h[digit]' + new ID + '...' 1814 return text[:3] + ' id="%s"' % header_id + text[3:] 1815 return text 1816 1817 @mark_stage(Stage.HEADERS) 1818 def _do_headers(self, text: str) -> str: 1819 # Setext-style headers: 1820 # Header 1 1821 # ======== 1822 # 1823 # Header 2 1824 # -------- 1825 1826 # atx-style headers: 1827 # # Header 1 1828 # ## Header 2 1829 # ## Header 2 with closing hashes ## 1830 # ... 1831 # ###### Header 6 1832 1833 if 'tag-friendly' in self.extras: 1834 return self._h_re_tag_friendly.sub(self._h_sub, text) 1835 return self._h_re.sub(self._h_sub, text) 1836 1837 _marker_ul_chars = '*+-' 1838 _marker_any = r'(?:[%s]|\d+\.)' % _marker_ul_chars 1839 _marker_ul = '(?:[%s])' % _marker_ul_chars 1840 _marker_ol = r'(?:\d+\.)' 1841 1842 def _list_sub(self, match: re.Match) -> str: 1843 lst = match.group(1) 1844 lst_type = match.group(4) in self._marker_ul_chars and "ul" or "ol" 1845 1846 if lst_type == 'ol' and match.group(4) != '1.': 1847 # if list doesn't start at 1 then set the ol start attribute 1848 lst_opts = ' start="%s"' % match.group(4)[:-1] 1849 else: 1850 lst_opts = '' 1851 1852 lst_opts = lst_opts + self._html_class_str_from_tag(lst_type) 1853 1854 result = self._process_list_items(lst) 1855 if self.list_level: 1856 return "<%s%s>\n%s</%s>\n" % (lst_type, lst_opts, result, lst_type) 1857 else: 1858 return "<%s%s>\n%s</%s>\n\n" % (lst_type, lst_opts, result, lst_type) 1859 1860 @mark_stage(Stage.LISTS) 1861 def _do_lists(self, text: str) -> str: 1862 # Form HTML ordered (numbered) and unordered (bulleted) lists. 1863 1864 # Iterate over each *non-overlapping* list match. 1865 pos = 0 1866 while True: 1867 # Find the *first* hit for either list style (ul or ol). We 1868 # match ul and ol separately to avoid adjacent lists of different 1869 # types running into each other (see issue #16). 1870 hits = [] 1871 for marker_pat in (self._marker_ul, self._marker_ol): 1872 less_than_tab = self.tab_width - 1 1873 other_marker_pat = self._marker_ul if marker_pat == self._marker_ol else self._marker_ol 1874 whole_list = r''' 1875 ( # \1 = whole list 1876 ( # \2 1877 ([ ]{0,%d}) # \3 = the indentation level of the list item marker 1878 (%s) # \4 = first list item marker 1879 [ \t]+ 1880 (?!\ *\4\ ) # '- - - ...' isn't a list. See 'not_quite_a_list' test case. 1881 ) 1882 (?:.+?) 1883 ( # \5 1884 \Z 1885 | 1886 \n{2,} 1887 (?=\S) 1888 (?! # Negative lookahead for another list item marker 1889 [ \t]* 1890 %s[ \t]+ 1891 ) 1892 | 1893 \n+ 1894 (?= 1895 \3 # lookahead for a different style of list item marker 1896 %s[ \t]+ 1897 ) 1898 ) 1899 ) 1900 ''' % (less_than_tab, marker_pat, marker_pat, other_marker_pat) 1901 if self.list_level: # sub-list 1902 list_re = re.compile("^"+whole_list, re.X | re.M | re.S) 1903 else: 1904 list_re = re.compile(r"(?:(?<=\n\n)|\A\n?)"+whole_list, 1905 re.X | re.M | re.S) 1906 match = list_re.search(text, pos) 1907 if match: 1908 hits.append((match.start(), match)) 1909 if not hits: 1910 break 1911 hits.sort() 1912 match = hits[0][1] 1913 start, end = match.span() 1914 middle = self._list_sub(match) 1915 text = text[:start] + middle + text[end:] 1916 pos = start + len(middle) # start pos for next attempted match 1917 1918 return text 1919 1920 _list_item_re = re.compile(r''' 1921 (\n)? # leading line = \1 1922 (^[ \t]*) # leading whitespace = \2 1923 (?P<marker>%s) [ \t]+ # list marker = \3 1924 ((?:.+?) # list item text = \4 1925 (\n{1,2})) # eols = \5 1926 (?= \n* (\Z | \2 (?P<next_marker>%s) [ \t]+)) 1927 ''' % (_marker_any, _marker_any), 1928 re.M | re.X | re.S) 1929 1930 _task_list_item_re = re.compile(r''' 1931 (\[[\ xX]\])[ \t]+ # tasklist marker = \1 1932 (.*) # list item text = \2 1933 ''', re.M | re.X | re.S) 1934 1935 _task_list_warpper_str = r'<input type="checkbox" class="task-list-item-checkbox" %sdisabled> %s' 1936 1937 def _task_list_item_sub(self, match: re.Match) -> str: 1938 marker = match.group(1) 1939 item_text = match.group(2) 1940 if marker in ['[x]','[X]']: 1941 return self._task_list_warpper_str % ('checked ', item_text) 1942 elif marker == '[ ]': 1943 return self._task_list_warpper_str % ('', item_text) 1944 # returning None has same effect as returning empty str, but only 1945 # one makes the type checker happy 1946 return '' 1947 1948 _last_li_endswith_two_eols = False 1949 def _list_item_sub(self, match: re.Match) -> str: 1950 item = match.group(4) 1951 leading_line = match.group(1) 1952 if leading_line or "\n\n" in item or self._last_li_endswith_two_eols: 1953 item = self._uniform_outdent(item, min_outdent=' ', max_outdent=self.tab)[1] 1954 item = self._run_block_gamut(item) 1955 else: 1956 # Recursion for sub-lists: 1957 item = self._do_lists(self._uniform_outdent(item, min_outdent=' ')[1]) 1958 if item.endswith('\n'): 1959 item = item[:-1] 1960 item = self._run_span_gamut(item) 1961 self._last_li_endswith_two_eols = (len(match.group(5)) == 2) 1962 1963 if "task_list" in self.extras: 1964 item = self._task_list_item_re.sub(self._task_list_item_sub, item) 1965 1966 return "<li>%s</li>\n" % item 1967 1968 def _process_list_items(self, list_str: str) -> str: 1969 # Process the contents of a single ordered or unordered list, 1970 # splitting it into individual list items. 1971 1972 # The $g_list_level global keeps track of when we're inside a list. 1973 # Each time we enter a list, we increment it; when we leave a list, 1974 # we decrement. If it's zero, we're not in a list anymore. 1975 # 1976 # We do this because when we're not inside a list, we want to treat 1977 # something like this: 1978 # 1979 # I recommend upgrading to version 1980 # 8. Oops, now this line is treated 1981 # as a sub-list. 1982 # 1983 # As a single paragraph, despite the fact that the second line starts 1984 # with a digit-period-space sequence. 1985 # 1986 # Whereas when we're inside a list (or sub-list), that line will be 1987 # treated as the start of a sub-list. What a kludge, huh? This is 1988 # an aspect of Markdown's syntax that's hard to parse perfectly 1989 # without resorting to mind-reading. Perhaps the solution is to 1990 # change the syntax rules such that sub-lists must start with a 1991 # starting cardinal number; e.g. "1." or "a.". 1992 self.list_level += 1 1993 self._last_li_endswith_two_eols = False 1994 list_str = list_str.rstrip('\n') + '\n' 1995 list_str = self._list_item_re.sub(self._list_item_sub, list_str) 1996 self.list_level -= 1 1997 return list_str 1998 1999 def _get_pygments_lexer(self, lexer_name: str): 2000 ''' 2001 Returns: 2002 `pygments.Lexer` or None if a lexer matching `lexer_name` is 2003 not found 2004 ''' 2005 try: 2006 from pygments import lexers, util 2007 except ImportError: 2008 return None 2009 try: 2010 return lexers.get_lexer_by_name(lexer_name) 2011 except util.ClassNotFound: 2012 return None 2013 2014 def _color_with_pygments( 2015 self, 2016 codeblock: str, 2017 lexer, 2018 **formatter_opts 2019 ) -> str: 2020 ''' 2021 TODO: this function is only referenced by the `FencedCodeBlocks` 2022 extra. May be worth moving over there 2023 2024 Args: 2025 codeblock: the codeblock to highlight 2026 lexer (pygments.Lexer): lexer to use 2027 formatter_opts: pygments HtmlFormatter options 2028 ''' 2029 import pygments 2030 import pygments.formatters 2031 2032 class HtmlCodeFormatter(pygments.formatters.HtmlFormatter): 2033 def _wrap_code(self, inner): 2034 """A function for use in a Pygments Formatter which 2035 wraps in <code> tags. 2036 """ 2037 yield 0, "<code>" 2038 for tup in inner: 2039 yield tup 2040 yield 0, "</code>" 2041 2042 def _add_newline(self, inner): 2043 # Add newlines around the inner contents so that _strict_tag_block_re matches the outer div. 2044 yield 0, "\n" 2045 yield from inner 2046 yield 0, "\n" 2047 2048 def wrap(self, source, outfile=None): 2049 """Return the source with a code, pre, and div.""" 2050 if outfile is None: 2051 # pygments >= 2.12 2052 return self._add_newline(self._wrap_pre(self._wrap_code(source))) 2053 else: 2054 # pygments < 2.12 2055 return self._wrap_div(self._add_newline(self._wrap_pre(self._wrap_code(source)))) 2056 2057 formatter_opts.setdefault("cssclass", "codehilite") 2058 formatter = HtmlCodeFormatter(**formatter_opts) 2059 return pygments.highlight(codeblock, lexer, formatter) 2060 2061 def _code_block_sub(self, match: re.Match) -> str: 2062 codeblock = match.group(1) 2063 codeblock = self._outdent(codeblock) 2064 codeblock = self._detab(codeblock) 2065 codeblock = codeblock.lstrip('\n') # trim leading newlines 2066 codeblock = codeblock.rstrip() # trim trailing whitespace 2067 2068 pre_class_str = self._html_class_str_from_tag("pre") 2069 code_class_str = self._html_class_str_from_tag("code") 2070 2071 codeblock = self._encode_code(codeblock) 2072 2073 return "\n<pre%s><code%s>%s\n</code></pre>\n" % ( 2074 pre_class_str, code_class_str, codeblock) 2075 2076 def _html_class_str_from_tag(self, tag: str) -> str: 2077 """Get the appropriate ' class="..."' string (note the leading 2078 space), if any, for the given tag. 2079 """ 2080 if "html-classes" not in self.extras: 2081 return "" 2082 try: 2083 html_classes_from_tag = self.extras["html-classes"] 2084 except TypeError: 2085 return "" 2086 else: 2087 if isinstance(html_classes_from_tag, dict): 2088 if tag in html_classes_from_tag: 2089 return ' class="%s"' % html_classes_from_tag[tag] 2090 return "" 2091 2092 @mark_stage(Stage.CODE_BLOCKS) 2093 def _do_code_blocks(self, text: str) -> str: 2094 """Process Markdown `<pre><code>` blocks.""" 2095 code_block_re = re.compile(r''' 2096 (?:\n\n|\A\n?) 2097 ( # $1 = the code block -- one or more lines, starting with a space/tab 2098 (?: 2099 (?:[ ]{%d} | \t) # Lines must start with a tab or a tab-width of spaces 2100 .*\n+ 2101 )+ 2102 ) 2103 ((?=^[ ]{0,%d}\S)|\Z) # Lookahead for non-space at line-start, or end of doc 2104 # Lookahead to make sure this block isn't already in a code block. 2105 # Needed when syntax highlighting is being used. 2106 (?!([^<]|<(/?)span)*\</code\>) 2107 ''' % (self.tab_width, self.tab_width), 2108 re.M | re.X) 2109 return code_block_re.sub(self._code_block_sub, text) 2110 2111 # Rules for a code span: 2112 # - backslash escapes are not interpreted in a code span 2113 # - to include one or or a run of more backticks the delimiters must 2114 # be a longer run of backticks 2115 # - cannot start or end a code span with a backtick; pad with a 2116 # space and that space will be removed in the emitted HTML 2117 # See `test/tm-cases/escapes.text` for a number of edge-case 2118 # examples. 2119 _code_span_re = re.compile(r''' 2120 (?<!\\) 2121 (`+) # \1 = Opening run of ` 2122 (?!`) # See Note A test/tm-cases/escapes.text 2123 (.+?) # \2 = The code block 2124 (?<!`) 2125 \1 # Matching closer 2126 (?!`) 2127 ''', re.X | re.S) 2128 2129 def _code_span_sub(self, match: re.Match) -> str: 2130 c = match.group(2).strip(" \t") 2131 c = self._encode_code(c) 2132 return "<code%s>%s</code>" % (self._html_class_str_from_tag("code"), c) 2133 2134 @mark_stage(Stage.CODE_SPANS) 2135 def _do_code_spans(self, text: str) -> str: 2136 # * Backtick quotes are used for <code></code> spans. 2137 # 2138 # * You can use multiple backticks as the delimiters if you want to 2139 # include literal backticks in the code span. So, this input: 2140 # 2141 # Just type ``foo `bar` baz`` at the prompt. 2142 # 2143 # Will translate to: 2144 # 2145 # <p>Just type <code>foo `bar` baz</code> at the prompt.</p> 2146 # 2147 # There's no arbitrary limit to the number of backticks you 2148 # can use as delimters. If you need three consecutive backticks 2149 # in your code, use four for delimiters, etc. 2150 # 2151 # * You can use spaces to get literal backticks at the edges: 2152 # 2153 # ... type `` `bar` `` ... 2154 # 2155 # Turns to: 2156 # 2157 # ... type <code>`bar`</code> ... 2158 return self._code_span_re.sub(self._code_span_sub, text) 2159 2160 def _encode_code(self, text: str) -> str: 2161 """Encode/escape certain characters inside Markdown code runs. 2162 The point is that in code, these characters are literals, 2163 and lose their special Markdown meanings. 2164 """ 2165 replacements = [ 2166 # Encode all ampersands; HTML entities are not 2167 # entities within a Markdown code span. 2168 ('&', '&'), 2169 # Do the angle bracket song and dance: 2170 ('<', '<'), 2171 ('>', '>'), 2172 ] 2173 for before, after in replacements: 2174 text = text.replace(before, after) 2175 hashed = _hash_text(text) 2176 self._code_table[text] = hashed 2177 return hashed 2178 2179 _strong_re = re.compile(r"(\*\*|__)(?=\S)(.+?[*_]?)(?<=\S)\1", re.S) 2180 _em_re = re.compile(r"(\*|_)(?=\S)(.*?\S)\1", re.S) 2181 2182 @mark_stage(Stage.ITALIC_AND_BOLD) 2183 def _do_italics_and_bold(self, text: str) -> str: 2184 # <strong> must go first: 2185 text = self._strong_re.sub(r"<strong>\2</strong>", text) 2186 text = self._em_re.sub(r"<em>\2</em>", text) 2187 return text 2188 2189 _block_quote_base = r''' 2190 ( # Wrap whole match in \1 2191 ( 2192 ^[ \t]*>%s[ \t]? # '>' at the start of a line 2193 .+\n # rest of the first line 2194 (.+\n)* # subsequent consecutive lines 2195 )+ 2196 ) 2197 ''' 2198 _block_quote_re = re.compile(_block_quote_base % '', re.M | re.X) 2199 _block_quote_re_spoiler = re.compile(_block_quote_base % '[ \t]*?!?', re.M | re.X) 2200 _bq_one_level_re = re.compile('^[ \t]*>[ \t]?', re.M) 2201 _bq_one_level_re_spoiler = re.compile('^[ \t]*>[ \t]*?![ \t]?', re.M) 2202 _bq_all_lines_spoilers = re.compile(r'\A(?:^[ \t]*>[ \t]*?!.*[\n\r]*)+\Z', re.M) 2203 _html_pre_block_re = re.compile(r'(\s*<pre>.+?</pre>)', re.S) 2204 def _dedent_two_spaces_sub(self, match: re.Match) -> str: 2205 return re.sub(r'(?m)^ ', '', match.group(1)) 2206 2207 def _block_quote_sub(self, match: re.Match) -> str: 2208 bq = match.group(1) 2209 is_spoiler = 'spoiler' in self.extras and self._bq_all_lines_spoilers.match(bq) 2210 # trim one level of quoting 2211 if is_spoiler: 2212 bq = self._bq_one_level_re_spoiler.sub('', bq) 2213 else: 2214 bq = self._bq_one_level_re.sub('', bq) 2215 # trim whitespace-only lines 2216 bq = self._ws_only_line_re.sub('', bq) 2217 bq = self._run_block_gamut(bq) # recurse 2218 2219 bq = re.sub('(?m)^', ' ', bq) 2220 # These leading spaces screw with <pre> content, so we need to fix that: 2221 bq = self._html_pre_block_re.sub(self._dedent_two_spaces_sub, bq) 2222 2223 if is_spoiler: 2224 return '<blockquote class="spoiler">\n%s\n</blockquote>\n\n' % bq 2225 else: 2226 return '<blockquote>\n%s\n</blockquote>\n\n' % bq 2227 2228 @mark_stage(Stage.BLOCK_QUOTES) 2229 def _do_block_quotes(self, text: str) -> str: 2230 if '>' not in text: 2231 return text 2232 if 'spoiler' in self.extras: 2233 return self._block_quote_re_spoiler.sub(self._block_quote_sub, text) 2234 else: 2235 return self._block_quote_re.sub(self._block_quote_sub, text) 2236 2237 @mark_stage(Stage.PARAGRAPHS) 2238 def _form_paragraphs(self, text: str) -> str: 2239 # Strip leading and trailing lines: 2240 text = text.strip('\n') 2241 2242 # Wrap <p> tags. 2243 grafs = [] 2244 for i, graf in enumerate(re.split(r"\n{2,}", text)): 2245 if graf in self.html_blocks: 2246 # Unhashify HTML blocks 2247 grafs.append(self.html_blocks[graf]) 2248 else: 2249 cuddled_list = None 2250 if "cuddled-lists" in self.extras: 2251 # Need to put back trailing '\n' for `_list_item_re` 2252 # match at the end of the paragraph. 2253 li = self._list_item_re.search(graf + '\n') 2254 # Two of the same list marker in this paragraph: a likely 2255 # candidate for a list cuddled to preceding paragraph 2256 # text (issue 33). Note the `[-1]` is a quick way to 2257 # consider numeric bullets (e.g. "1." and "2.") to be 2258 # equal. 2259 if (li and len(li.group(2)) <= 3 2260 and ( 2261 (li.group("next_marker") and li.group("marker")[-1] == li.group("next_marker")[-1]) 2262 or 2263 li.group("next_marker") is None 2264 ) 2265 ): 2266 start = li.start() 2267 cuddled_list = self._do_lists(graf[start:]).rstrip("\n") 2268 if re.match(r'^<(?:ul|ol).*?>', cuddled_list): 2269 graf = graf[:start] 2270 else: 2271 # Not quite a cuddled list. (See not_quite_a_list_cuddled_lists test case) 2272 # Store as a simple paragraph. 2273 graf = cuddled_list 2274 cuddled_list = None 2275 2276 # Wrap <p> tags. 2277 graf = self._run_span_gamut(graf) 2278 grafs.append("<p%s>" % self._html_class_str_from_tag('p') + graf.lstrip(" \t") + "</p>") 2279 2280 if cuddled_list: 2281 grafs.append(cuddled_list) 2282 2283 return "\n\n".join(grafs) 2284 2285 def _add_footnotes(self, text: str) -> str: 2286 if self.footnotes: 2287 footer = [ 2288 '<div class="footnotes">', 2289 '<hr' + self.empty_element_suffix, 2290 '<ol>', 2291 ] 2292 2293 if not self.footnote_title: 2294 self.footnote_title = "Jump back to footnote %d in the text." 2295 if not self.footnote_return_symbol: 2296 self.footnote_return_symbol = "↩" 2297 2298 # self.footnotes is generated in _strip_footnote_definitions, which runs re.sub on the whole 2299 # text. This means that the dict keys are inserted in order of appearance. Use the dict to 2300 # sort footnote ids by that same order 2301 self.footnote_ids.sort(key=lambda a: list(self.footnotes.keys()).index(a)) 2302 for i, id in enumerate(self.footnote_ids): 2303 if i != 0: 2304 footer.append('') 2305 footer.append('<li id="fn-%s">' % id) 2306 footer.append(self._run_block_gamut(self.footnotes[id])) 2307 try: 2308 backlink = ('<a href="#fnref-%s" ' + 2309 'class="footnoteBackLink" ' + 2310 'title="' + self.footnote_title + '">' + 2311 self.footnote_return_symbol + 2312 '</a>') % (id, i+1) 2313 except TypeError: 2314 log.debug("Footnote error. `footnote_title` " 2315 "must include parameter. Using defaults.") 2316 backlink = ('<a href="#fnref-%s" ' 2317 'class="footnoteBackLink" ' 2318 'title="Jump back to footnote %d in the text.">' 2319 '↩</a>' % (id, i+1)) 2320 2321 if footer[-1].endswith("</p>"): 2322 footer[-1] = footer[-1][:-len("</p>")] \ 2323 + ' ' + backlink + "</p>" 2324 else: 2325 footer.append("\n<p>%s</p>" % backlink) 2326 footer.append('</li>') 2327 footer.append('</ol>') 2328 footer.append('</div>') 2329 return text + '\n\n' + '\n'.join(footer) 2330 else: 2331 return text 2332 2333 _naked_lt_re = re.compile(r'<(?![a-z/?\$!])', re.I) 2334 _naked_gt_re = re.compile(r'''(?<![a-z0-9?!/'"-])>''', re.I) 2335 2336 def _encode_amps_and_angles(self, text: str) -> str: 2337 # Smart processing for ampersands and angle brackets that need 2338 # to be encoded. 2339 text = _AMPERSAND_RE.sub('&', text) 2340 2341 # Encode naked <'s 2342 text = self._naked_lt_re.sub('<', text) 2343 2344 # Encode naked >'s 2345 # Note: Other markdown implementations (e.g. Markdown.pl, PHP 2346 # Markdown) don't do this. 2347 text = self._naked_gt_re.sub('>', text) 2348 return text 2349 2350 _incomplete_tags_re = re.compile(r"<(!--|/?\w+?(?!\w)\s*?.+?(?:[\s/]+?|$))") 2351 2352 def _encode_incomplete_tags(self, text: str) -> str: 2353 if self.safe_mode not in ("replace", "escape"): 2354 return text 2355 2356 if text.endswith(">"): 2357 return text # this is not an incomplete tag, this is a link in the form <http://x.y.z> 2358 2359 def incomplete_tags_sub(match): 2360 return match.group().replace('<', '<') 2361 2362 return self._incomplete_tags_re.sub(incomplete_tags_sub, text) 2363 2364 def _encode_backslash_escapes(self, text: str) -> str: 2365 for ch, escape in list(self._escape_table.items()): 2366 text = text.replace("\\"+ch, escape) 2367 return text 2368 2369 _auto_link_re = re.compile(r'<((https?|ftp):[^\'">\s]+)>', re.I) 2370 def _auto_link_sub(self, match: re.Match) -> str: 2371 g1 = match.group(1) 2372 return '<a href="%s">%s</a>' % (self._protect_url(g1), g1) 2373 2374 _auto_email_link_re = re.compile(r""" 2375 < 2376 (?:mailto:)? 2377 ( 2378 [-.\w]+ 2379 \@ 2380 [-\w]+(\.[-\w]+)*\.[a-z]+ 2381 ) 2382 > 2383 """, re.I | re.X | re.U) 2384 def _auto_email_link_sub(self, match: re.Match) -> str: 2385 return self._encode_email_address( 2386 self._unescape_special_chars(match.group(1))) 2387 2388 def _do_auto_links(self, text: str) -> str: 2389 text = self._auto_link_re.sub(self._auto_link_sub, text) 2390 text = self._auto_email_link_re.sub(self._auto_email_link_sub, text) 2391 return text 2392 2393 def _encode_email_address(self, addr: str) -> str: 2394 # Input: an email address, e.g. "foo@example.com" 2395 # 2396 # Output: the email address as a mailto link, with each character 2397 # of the address encoded as either a decimal or hex entity, in 2398 # the hopes of foiling most address harvesting spam bots. E.g.: 2399 # 2400 # <a href="mailto:foo@e 2401 # xample.com">foo 2402 # @example.com</a> 2403 # 2404 # Based on a filter by Matthew Wickline, posted to the BBEdit-Talk 2405 # mailing list: <http://tinyurl.com/yu7ue> 2406 chars = [_xml_encode_email_char_at_random(ch) 2407 for ch in "mailto:" + addr] 2408 # Strip the mailto: from the visible part. 2409 addr = '<a href="%s">%s</a>' \ 2410 % (''.join(chars), ''.join(chars[7:])) 2411 return addr 2412 2413 def _unescape_special_chars(self, text: str) -> str: 2414 # Swap back in all the special characters we've hidden. 2415 hashmap = tuple(self._escape_table.items()) + tuple(self._code_table.items()) 2416 # html_blocks table is in format {hash: item} compared to usual {item: hash} 2417 hashmap += tuple(tuple(reversed(i)) for i in self.html_blocks.items()) 2418 while True: 2419 orig_text = text 2420 for ch, hash in hashmap: 2421 text = text.replace(hash, ch) 2422 if text == orig_text: 2423 break 2424 return text 2425 2426 def _outdent(self, text: str) -> str: 2427 # Remove one level of line-leading tabs or spaces 2428 return self._outdent_re.sub('', text) 2429 2430 @staticmethod 2431 def _uniform_outdent( 2432 text: str, 2433 min_outdent: Optional[str] = None, 2434 max_outdent: Optional[str] = None 2435 ) -> Tuple[str, str]: 2436 ''' 2437 Removes the smallest common leading indentation from each (non empty) 2438 line of `text` and returns said indent along with the outdented text. 2439 2440 Args: 2441 min_outdent: make sure the smallest common whitespace is at least this size 2442 max_outdent: the maximum amount a line can be outdented by 2443 ''' 2444 2445 # find the leading whitespace for every line 2446 whitespace: List[Union[str, None]] = [ 2447 re.findall(r'^[ \t]*', line)[0] if line else None 2448 for line in text.splitlines() 2449 ] 2450 whitespace_not_empty = [i for i in whitespace if i is not None] 2451 2452 # if no whitespace detected (ie: no lines in code block, issue #505) 2453 if not whitespace_not_empty: 2454 return '', text 2455 2456 # get minimum common whitespace 2457 outdent = min(whitespace_not_empty) 2458 # adjust min common ws to be within bounds 2459 if min_outdent is not None: 2460 outdent = min([i for i in whitespace_not_empty if i >= min_outdent] or [min_outdent]) 2461 if max_outdent is not None: 2462 outdent = min(outdent, max_outdent) 2463 2464 outdented = [] 2465 for line_ws, line in zip(whitespace, text.splitlines(True)): 2466 if line.startswith(outdent): 2467 # if line starts with smallest common ws, dedent it 2468 outdented.append(line.replace(outdent, '', 1)) 2469 elif line_ws is not None and line_ws < outdent: 2470 # if less indented than min common whitespace then outdent as much as possible 2471 outdented.append(line.replace(line_ws, '', 1)) 2472 else: 2473 outdented.append(line) 2474 2475 return outdent, ''.join(outdented) 2476 2477 @staticmethod 2478 def _uniform_indent( 2479 text: str, 2480 indent: str, 2481 include_empty_lines: bool = False, 2482 indent_empty_lines: bool = False 2483 ) -> str: 2484 ''' 2485 Uniformly indent a block of text by a fixed amount 2486 2487 Args: 2488 text: the text to indent 2489 indent: a string containing the indent to apply 2490 include_empty_lines: don't remove whitespace only lines 2491 indent_empty_lines: indent whitespace only lines with the rest of the text 2492 ''' 2493 blocks = [] 2494 for line in text.splitlines(True): 2495 if line.strip() or indent_empty_lines: 2496 blocks.append(indent + line) 2497 elif include_empty_lines: 2498 blocks.append(line) 2499 else: 2500 blocks.append('') 2501 return ''.join(blocks) 2502 2503 @staticmethod 2504 def _match_overlaps_substr(text, match: re.Match, substr: str) -> bool: 2505 ''' 2506 Checks if a regex match overlaps with a substring in the given text. 2507 ''' 2508 for instance in re.finditer(re.escape(substr), text): 2509 start, end = instance.span() 2510 if start <= match.start() <= end: 2511 return True 2512 if start <= match.end() <= end: 2513 return True 2514 return False 2515 2516 2517class MarkdownWithExtras(Markdown): 2518 """A markdowner class that enables most extras: 2519 2520 - footnotes 2521 - fenced-code-blocks (only highlights code if 'pygments' Python module on path) 2522 2523 These are not included: 2524 - pyshell (specific to Python-related documenting) 2525 - code-friendly (because it *disables* part of the syntax) 2526 - link-patterns (because you need to specify some actual 2527 link-patterns anyway) 2528 """ 2529 extras = ["footnotes", "fenced-code-blocks"] # type: ignore 2530 2531 2532# ---------------------------------------------------------- 2533# Extras 2534# ---------------------------------------------------------- 2535 2536# Base classes 2537# ---------------------------------------------------------- 2538 2539class Extra(ABC): 2540 _registry: Dict[str, Type['Extra']] = {} 2541 _exec_order: Dict[Stage, Tuple[List[Type['Extra']], List[Type['Extra']]]] = {} 2542 2543 name: str 2544 ''' 2545 An identifiable name that users can use to invoke the extra 2546 in the Markdown class 2547 ''' 2548 order: Tuple[Collection[Union[Stage, Type['Extra']]], Collection[Union[Stage, Type['Extra']]]] 2549 ''' 2550 Tuple of two iterables containing the stages/extras this extra will run before and 2551 after, respectively 2552 ''' 2553 2554 def __init__(self, md: Markdown, options: Optional[dict]): 2555 ''' 2556 Args: 2557 md: An instance of `Markdown` 2558 options: a dict of settings to alter the extra's behaviour 2559 ''' 2560 self.md = md 2561 self.options = options if options is not None else {} 2562 2563 @classmethod 2564 def deregister(cls): 2565 ''' 2566 Removes the class from the extras registry and unsets its execution order. 2567 ''' 2568 if cls.name in cls._registry: 2569 del cls._registry[cls.name] 2570 2571 for exec_order in Extra._exec_order.values(): 2572 # find everywhere this extra is mentioned and remove it 2573 for section in exec_order: 2574 while cls in section: 2575 section.remove(cls) 2576 2577 @classmethod 2578 def register(cls): 2579 ''' 2580 Registers the class for use with `Markdown` and calculates its execution order based on 2581 the `order` class attribute. 2582 ''' 2583 cls._registry[cls.name] = cls 2584 2585 for index, item in enumerate((*cls.order[0], *cls.order[1])): 2586 before = index < len(cls.order[0]) 2587 if not isinstance(item, Stage) and issubclass(item, Extra): 2588 # eg: FencedCodeBlocks 2589 for exec_orders in Extra._exec_order.values(): 2590 # insert this extra everywhere the other one is mentioned 2591 for section in exec_orders: 2592 if item in section: 2593 to_index = section.index(item) 2594 if not before: 2595 to_index += 1 2596 section.insert(to_index, cls) 2597 else: 2598 # eg: Stage.PREPROCESS 2599 Extra._exec_order.setdefault(item, ([], [])) 2600 if cls in Extra._exec_order[item][0 if before else 1]: 2601 # extra is already runnig after this stage. Don't duplicate that effort 2602 continue 2603 if before: 2604 Extra._exec_order[item][0].insert(0, cls) 2605 else: 2606 Extra._exec_order[item][1].append(cls) 2607 2608 @abstractmethod 2609 def run(self, text: str) -> str: 2610 ''' 2611 Run the extra against the given text. 2612 2613 Returns: 2614 The new text after being modified by the extra 2615 ''' 2616 ... 2617 2618 def test(self, text: str) -> bool: 2619 ''' 2620 Check a section of markdown to see if this extra should be run upon it. 2621 The default implementation will always return True but it's recommended to override 2622 this behaviour to improve performance. 2623 ''' 2624 return True 2625 2626 2627class ItalicAndBoldProcessor(Extra): 2628 ''' 2629 An ABC that provides hooks for dealing with italics and bold syntax. 2630 This class is set to trigger both before AND after the italics and bold stage. 2631 This allows any child classes to intercept instances of bold or italic syntax and 2632 change the output or hash it to prevent it from being processed. 2633 2634 After the I&B stage any hashes in the `hash_tables` instance variable are replaced. 2635 ''' 2636 name = 'italic-and-bold-processor' 2637 order = (Stage.ITALIC_AND_BOLD,), (Stage.ITALIC_AND_BOLD,) 2638 2639 strong_re = Markdown._strong_re 2640 em_re = Markdown._em_re 2641 2642 def __init__(self, md: Markdown, options: dict): 2643 super().__init__(md, options) 2644 self.hash_table = {} 2645 2646 def run(self, text): 2647 if self.md.order < Stage.ITALIC_AND_BOLD: 2648 text = self.strong_re.sub(self.sub, text) 2649 text = self.em_re.sub(self.sub, text) 2650 else: 2651 # push any hashed values back, using a while loop to deal with recursive hashes 2652 orig_text = '' 2653 while orig_text != text: 2654 orig_text = text 2655 for key, substr in self.hash_table.items(): 2656 text = text.replace(key, substr) 2657 return text 2658 2659 @abstractmethod 2660 def sub(self, match: re.Match) -> str: 2661 # do nothing. Let `Markdown._do_italics_and_bold` do its thing later 2662 return match.string[match.start(): match.end()] 2663 2664 def sub_hash(self, match: re.Match) -> str: 2665 substr = match.string[match.start(): match.end()] 2666 key = _hash_text(substr) 2667 self.hash_table[key] = substr 2668 return key 2669 2670 def test(self, text): 2671 if self.md.order < Stage.ITALIC_AND_BOLD: 2672 return '*' in text or '_' in text 2673 return self.hash_table and re.search(r'md5-[0-9a-z]{32}', text) 2674 2675# User facing extras 2676# ---------------------------------------------------------- 2677 2678 2679class Admonitions(Extra): 2680 ''' 2681 Enable parsing of RST admonitions 2682 ''' 2683 2684 name = 'admonitions' 2685 order = (Stage.BLOCK_GAMUT, Stage.LINK_DEFS), () 2686 2687 admonitions = r'admonition|attention|caution|danger|error|hint|important|note|tip|warning' 2688 2689 admonitions_re = re.compile(r''' 2690 ^(\ *)\.\.\ (%s)::\ * # $1 leading indent, $2 the admonition 2691 (.*)? # $3 admonition title 2692 ((?:\s*\n\1\ {3,}.*)+?) # $4 admonition body (required) 2693 (?=\s*(?:\Z|\n{4,}|\n\1?\ {0,2}\S)) # until EOF, 3 blank lines or something less indented 2694 ''' % admonitions, 2695 re.IGNORECASE | re.MULTILINE | re.VERBOSE 2696 ) 2697 2698 def test(self, text): 2699 return self.admonitions_re.search(text) is not None 2700 2701 def sub(self, match: re.Match) -> str: 2702 lead_indent, admonition_name, title, body = match.groups() 2703 2704 admonition_type = '<strong>%s</strong>' % admonition_name 2705 2706 # figure out the class names to assign the block 2707 if admonition_name.lower() == 'admonition': 2708 admonition_class = 'admonition' 2709 else: 2710 admonition_class = 'admonition %s' % admonition_name.lower() 2711 2712 # titles are generally optional 2713 if title: 2714 title = '<em>%s</em>' % title 2715 2716 # process the admonition body like regular markdown 2717 body = self.md._run_block_gamut("\n%s\n" % self.md._uniform_outdent(body)[1]) 2718 2719 # indent the body before placing inside the aside block 2720 admonition = self.md._uniform_indent( 2721 '%s\n%s\n\n%s\n' % (admonition_type, title, body), 2722 self.md.tab, False 2723 ) 2724 # wrap it in an aside 2725 admonition = '<aside class="%s">\n%s</aside>' % (admonition_class, admonition) 2726 # now indent the whole admonition back to where it started 2727 return self.md._uniform_indent(admonition, lead_indent, False) 2728 2729 def run(self, text): 2730 return self.admonitions_re.sub(self.sub, text) 2731 2732 2733class Alerts(Extra): 2734 ''' 2735 Markdown Alerts as per 2736 https://docs.github.com/en/get-started/writing-on-github/getting-started-with-writing-and-formatting-on-github/basic-writing-and-formatting-syntax#alerts 2737 ''' 2738 2739 name = 'alerts' 2740 order = (), (Stage.BLOCK_QUOTES, ) 2741 2742 alert_re = re.compile(r''' 2743 <blockquote>\s* 2744 <p> 2745 \[!(?P<type>NOTE|TIP|IMPORTANT|WARNING|CAUTION)\] 2746 (?P<closing_tag></p>[ \t]*\n?)? 2747 (?P<contents>[\s\S]+?) 2748 </blockquote> 2749 ''', re.X 2750 ) 2751 2752 def test(self, text): 2753 return "<blockquote>" in text 2754 2755 def sub(self, match: re.Match) -> str: 2756 typ = match["type"].lower() 2757 heading = f"<em>{match['type'].title()}</em>" 2758 contents = match["contents"].strip() 2759 if match["closing_tag"]: 2760 return f'<div class="alert {typ}">\n{heading}\n{contents}\n</div>' 2761 else: 2762 return f'<div class="alert {typ}">\n{heading}\n<p>{contents}\n</div>' 2763 2764 def run(self, text): 2765 return self.alert_re.sub(self.sub, text) 2766 2767 2768class _BreaksExtraOpts(TypedDict, total=False): 2769 '''Options for the `Breaks` extra''' 2770 on_backslash: bool 2771 '''Replace backslashes at the end of a line with <br>''' 2772 on_newline: bool 2773 '''Replace single new line characters with <br> when True''' 2774 2775 2776class Breaks(Extra): 2777 name = 'breaks' 2778 order = (), (Stage.ITALIC_AND_BOLD,) 2779 options: _BreaksExtraOpts 2780 2781 def run(self, text): 2782 on_backslash = self.options.get('on_backslash', False) 2783 on_newline = self.options.get('on_newline', False) 2784 2785 if on_backslash and on_newline: 2786 pattern = r' *\\?' 2787 elif on_backslash: 2788 pattern = r'(?: *\\| {2,})' 2789 elif on_newline: 2790 pattern = r' *' 2791 else: 2792 pattern = r' {2,}' 2793 2794 break_tag = "<br%s\n" % self.md.empty_element_suffix 2795 text = re.sub(pattern + r"\n(?!\<(?:\/?(ul|ol|li))\>)", break_tag, text) 2796 2797 return text 2798 2799 2800class CodeFriendly(ItalicAndBoldProcessor): 2801 ''' 2802 Disable _ and __ for em and strong. 2803 ''' 2804 name = 'code-friendly' 2805 2806 def sub(self, match: re.Match) -> str: 2807 syntax = match.group(1) 2808 text: str = match.string[match.start(): match.end()] 2809 if '_' in syntax: 2810 # if using _this_ syntax, hash the whole thing so that it doesn't get processed 2811 key = _hash_text(text) 2812 self.hash_table[key] = text 2813 return key 2814 elif '_' in text: 2815 # if the text within the bold/em markers contains '_' then hash those contents to protect them from em_re 2816 text = text[len(syntax): -len(syntax)] 2817 key = _hash_text(text) 2818 self.hash_table[key] = text 2819 return syntax + key + syntax 2820 # if no underscores are present, the text is fine and we can just leave it alone 2821 return super().sub(match) 2822 2823 2824class FencedCodeBlocks(Extra): 2825 ''' 2826 Allows a code block to not have to be indented 2827 by fencing it with '```' on a line before and after. Based on 2828 <http://github.github.com/github-flavored-markdown/> with support for 2829 syntax highlighting. 2830 ''' 2831 2832 name = 'fenced-code-blocks' 2833 order = (Stage.LINK_DEFS, Stage.BLOCK_GAMUT), (Stage.PREPROCESS,) 2834 2835 fenced_code_block_re = re.compile(r''' 2836 (?:\n+|\A\n?|(?<=\n)) 2837 (^[ \t]*`{3,})\s{0,99}?([\w+-]+)?\s{0,99}?\n # $1 = opening fence (captured for back-referencing), $2 = optional lang 2838 (.*?) # $3 = code block content 2839 \1[ \t]*\n # closing fence 2840 ''', re.M | re.X | re.S) 2841 2842 def test(self, text): 2843 if '```' not in text: 2844 return False 2845 if self.md.stage == Stage.PREPROCESS and not self.md.safe_mode: 2846 return True 2847 if self.md.stage == Stage.LINK_DEFS and self.md.safe_mode: 2848 return True 2849 return self.md.stage == Stage.BLOCK_GAMUT 2850 2851 def _code_block_with_lexer_sub( 2852 self, 2853 codeblock: str, 2854 leading_indent: str, 2855 lexer 2856 ) -> str: 2857 ''' 2858 Args: 2859 codeblock: the codeblock to format 2860 leading_indent: the indentation to prefix the block with 2861 lexer (pygments.Lexer): the lexer to use 2862 ''' 2863 formatter_opts = self.md.extras['fenced-code-blocks'] or {} 2864 2865 def unhash_code(codeblock): 2866 for key, sanitized in list(self.md.html_spans.items()): 2867 codeblock = codeblock.replace(key, sanitized) 2868 replacements = [ 2869 ("&", "&"), 2870 ("<", "<"), 2871 (">", ">") 2872 ] 2873 for old, new in replacements: 2874 codeblock = codeblock.replace(old, new) 2875 return codeblock 2876 # remove leading indent from code block 2877 _, codeblock = self.md._uniform_outdent(codeblock, max_outdent=leading_indent) 2878 2879 codeblock = unhash_code(codeblock) 2880 colored = self.md._color_with_pygments(codeblock, lexer, 2881 **formatter_opts) 2882 2883 # add back the indent to all lines 2884 return "\n%s\n" % self.md._uniform_indent(colored, leading_indent, True) 2885 2886 def tags(self, lexer_name: str) -> Tuple[str, str]: 2887 ''' 2888 Returns the tags that the encoded code block will be wrapped in, based 2889 upon the lexer name. 2890 2891 This function can be overridden by subclasses to piggy-back off of the 2892 fenced code blocks syntax (see `Mermaid` extra). 2893 2894 Returns: 2895 The opening and closing tags, as strings within a tuple 2896 ''' 2897 pre_class = self.md._html_class_str_from_tag('pre') 2898 if "highlightjs-lang" in self.md.extras and lexer_name: 2899 code_class = ' class="%s language-%s"' % (lexer_name, lexer_name) 2900 else: 2901 code_class = self.md._html_class_str_from_tag('code') 2902 return ('<pre%s><code%s>' % (pre_class, code_class), '</code></pre>') 2903 2904 def sub(self, match: re.Match) -> str: 2905 lexer_name = match.group(2) 2906 codeblock = match.group(3) 2907 codeblock = codeblock[:-1] # drop one trailing newline 2908 2909 # Use pygments only if not using the highlightjs-lang extra 2910 if lexer_name and "highlightjs-lang" not in self.md.extras: 2911 lexer = self.md._get_pygments_lexer(lexer_name) 2912 if lexer: 2913 leading_indent = ' '*(len(match.group(1)) - len(match.group(1).lstrip())) 2914 return self._code_block_with_lexer_sub(codeblock, leading_indent, lexer) 2915 2916 # Fenced code blocks need to be outdented before encoding, and then reapplied 2917 leading_indent = ' ' * (len(match.group(1)) - len(match.group(1).lstrip())) 2918 if codeblock: 2919 # only run the codeblock through the outdenter if not empty 2920 leading_indent, codeblock = self.md._uniform_outdent(codeblock, max_outdent=leading_indent) 2921 2922 codeblock = self.md._encode_code(codeblock) 2923 2924 tags = self.tags(lexer_name) 2925 2926 return "\n%s%s%s\n%s%s\n" % (leading_indent, tags[0], codeblock, leading_indent, tags[1]) 2927 2928 def run(self, text): 2929 return self.fenced_code_block_re.sub(self.sub, text) 2930 2931 2932class Latex(Extra): 2933 ''' 2934 Convert $ and $$ to <math> and </math> tags for inline and block math. 2935 ''' 2936 name = 'latex' 2937 order = (Stage.CODE_BLOCKS, FencedCodeBlocks), () 2938 2939 _single_dollar_re = re.compile(r'(?<!\$)\$(?!\$)(.*?)\$') 2940 _double_dollar_re = re.compile(r'\$\$(.*?)\$\$', re.DOTALL) 2941 2942 # Ways to escape 2943 _pre_code_block_re = re.compile(r"<pre>(.*?)</pre>", re.DOTALL) # Wraped in <pre> 2944 _triple_re = re.compile(r'```(.*?)```', re.DOTALL) # Wrapped in a code block ``` 2945 _single_re = re.compile(r'(?<!`)(`)(.*?)(?<!`)\1(?!`)') # Wrapped in a single ` 2946 2947 converter = None 2948 code_blocks = {} 2949 2950 def _convert_single_match(self, match): 2951 return self.converter.convert(match.group(1)) 2952 2953 def _convert_double_match(self, match): 2954 return self.converter.convert(match.group(1).replace(r"\n", ''), display="block") 2955 2956 def code_placeholder(self, match): 2957 placeholder = f"<!--CODE_BLOCK_{len(self.code_blocks)}-->" 2958 self.code_blocks[placeholder] = match.group(0) 2959 return placeholder 2960 2961 def run(self, text): 2962 try: 2963 import latex2mathml.converter 2964 self.converter = latex2mathml.converter 2965 except ImportError: 2966 raise ImportError('The "latex" extra requires the "latex2mathml" package to be installed.') 2967 2968 # Escape by replacing with a code block 2969 text = self._pre_code_block_re.sub(self.code_placeholder, text) 2970 text = self._single_re.sub(self.code_placeholder, text) 2971 text = self._triple_re.sub(self.code_placeholder, text) 2972 2973 text = self._single_dollar_re.sub(self._convert_single_match, text) 2974 text = self._double_dollar_re.sub(self._convert_double_match, text) 2975 2976 # Convert placeholder tag back to original code 2977 for placeholder, code_block in self.code_blocks.items(): 2978 text = text.replace(placeholder, code_block) 2979 2980 return text 2981 2982 2983class LinkPatterns(Extra): 2984 ''' 2985 Auto-link given regex patterns in text (e.g. bug number 2986 references, revision number references). 2987 ''' 2988 name = 'link-patterns' 2989 order = (Stage.LINKS,), () 2990 options: _link_patterns 2991 2992 _basic_link_re = re.compile(r'!?\[.*?\]\(.*?\)') 2993 2994 def run(self, text): 2995 link_from_hash = {} 2996 for regex, repl in self.options: 2997 replacements = [] 2998 for match in regex.finditer(text): 2999 if any(self.md._match_overlaps_substr(text, match, h) for h in link_from_hash): 3000 continue 3001 3002 if callable(repl): 3003 href = repl(match) 3004 else: 3005 href = match.expand(repl) 3006 replacements.append((match.span(), href)) 3007 for (start, end), href in reversed(replacements): 3008 3009 # Do not match against links inside brackets. 3010 if text[start - 1:start] == '[' and text[end:end + 1] == ']': 3011 continue 3012 3013 # Do not match against links in the standard markdown syntax. 3014 if text[start - 2:start] == '](' or text[end:end + 2] == '")': 3015 continue 3016 3017 # Do not match against links which are escaped. 3018 if text[start - 3:start] == '"""' and text[end:end + 3] == '"""': 3019 text = text[:start - 3] + text[start:end] + text[end + 3:] 3020 continue 3021 3022 # search the text for anything that looks like a link 3023 is_inside_link = False 3024 for link_re in (self.md._auto_link_re, self._basic_link_re): 3025 for match in link_re.finditer(text): 3026 if any((r[0] <= start and end <= r[1]) for r in match.regs): 3027 # if the link pattern start and end pos is within the bounds of 3028 # something that looks like a link, then don't process it 3029 is_inside_link = True 3030 break 3031 else: 3032 continue 3033 break 3034 3035 if is_inside_link: 3036 continue 3037 3038 escaped_href = ( 3039 href.replace('"', '"') # b/c of attr quote 3040 # To avoid markdown <em> and <strong>: 3041 .replace('*', self.md._escape_table['*']) 3042 .replace('_', self.md._escape_table['_'])) 3043 link = '<a href="%s">%s</a>' % (escaped_href, text[start:end]) 3044 hash = _hash_text(link) 3045 link_from_hash[hash] = link 3046 text = text[:start] + hash + text[end:] 3047 for hash, link in list(link_from_hash.items()): 3048 text = text.replace(hash, link) 3049 return text 3050 3051 def test(self, text): 3052 return True 3053 3054 3055class MarkdownInHTML(Extra): 3056 ''' 3057 Allow the use of `markdown="1"` in a block HTML tag to 3058 have markdown processing be done on its contents. Similar to 3059 <http://michelf.com/projects/php-markdown/extra/#markdown-attr> but with 3060 some limitations. 3061 ''' 3062 name = 'markdown-in-html' 3063 order = (), (Stage.HASH_HTML,) 3064 3065 def run(self, text): 3066 def callback(block): 3067 indent, block = self.md._uniform_outdent(block) 3068 block = self.md._hash_html_block_sub(block) 3069 block = self.md._uniform_indent(block, indent, include_empty_lines=True, indent_empty_lines=False) 3070 return block 3071 3072 return self.md._strict_tag_block_sub(text, self.md._block_tags_a, callback, True) 3073 3074 def test(self, text): 3075 return True 3076 3077 3078class Mermaid(FencedCodeBlocks): 3079 name = 'mermaid' 3080 order = (FencedCodeBlocks,), () 3081 3082 def tags(self, lexer_name): 3083 if lexer_name == 'mermaid': 3084 return ('<pre class="mermaid-pre"><div class="mermaid">', '</div></pre>') 3085 return super().tags(lexer_name) 3086 3087 3088class MiddleWordEm(ItalicAndBoldProcessor): 3089 ''' 3090 Allows or disallows emphasis syntax in the middle of words, 3091 defaulting to allow. Disabling this means that `this_text_here` will not be 3092 converted to `this<em>text</em>here`. 3093 ''' 3094 name = 'middle-word-em' 3095 order = (CodeFriendly,), (Stage.ITALIC_AND_BOLD,) 3096 3097 def __init__(self, md: Markdown, options: Union[dict, bool]): 3098 ''' 3099 Args: 3100 md: the markdown instance 3101 options: can be bool for backwards compatibility but will be converted to a dict 3102 in the constructor. All options are: 3103 - allowed (bool): whether to allow emphasis in the middle of a word. 3104 If `options` is a bool it will be placed under this key. 3105 ''' 3106 if isinstance(options, bool): 3107 options = {'allowed': options} 3108 options.setdefault('allowed', True) 3109 super().__init__(md, options) 3110 3111 self.liberal_em_re = self.em_re 3112 if not options['allowed']: 3113 self.em_re = re.compile(r'(?<=\b)%s(?=\b)' % self.liberal_em_re.pattern, self.liberal_em_re.flags) 3114 3115 def run(self, text): 3116 # run strong and whatnot first 3117 # this also will process all strict ems 3118 text = super().run(text) 3119 if self.md.order < self.md.stage: 3120 # hash all non-valid ems 3121 text = self.liberal_em_re.sub(self.sub_hash, text) 3122 return text 3123 3124 def sub(self, match: re.Match) -> str: 3125 syntax = match.group(1) 3126 if len(syntax) != 1: 3127 # strong syntax 3128 return super().sub(match) 3129 return '<em>%s</em>' % match.group(2) 3130 3131 3132class Numbering(Extra): 3133 ''' 3134 Support of generic counters. Non standard extension to 3135 allow sequential numbering of figures, tables, equations, exhibits etc. 3136 ''' 3137 3138 name = 'numbering' 3139 order = (Stage.LINK_DEFS,), () 3140 3141 def run(self, text): 3142 # First pass to define all the references 3143 regex_defns = re.compile(r''' 3144 \[\#(\w+) # the counter. Open square plus hash plus a word \1 3145 ([^@]*) # Some optional characters, that aren't an @. \2 3146 @(\w+) # the id. Should this be normed? \3 3147 ([^\]]*)\] # The rest of the text up to the terminating ] \4 3148 ''', re.VERBOSE) 3149 regex_subs = re.compile(r"\[@(\w+)\s*\]") # [@ref_id] 3150 counters = {} 3151 references = {} 3152 replacements = [] 3153 definition_html = '<figcaption class="{}" id="counter-ref-{}">{}{}{}</figcaption>' 3154 reference_html = '<a class="{}" href="#counter-ref-{}">{}</a>' 3155 for match in regex_defns.finditer(text): 3156 # We must have four match groups otherwise this isn't a numbering reference 3157 if len(match.groups()) != 4: 3158 continue 3159 counter = match.group(1) 3160 text_before = match.group(2).strip() 3161 ref_id = match.group(3) 3162 text_after = match.group(4) 3163 number = counters.get(counter, 1) 3164 references[ref_id] = (number, counter) 3165 replacements.append((match.start(0), 3166 definition_html.format(counter, 3167 ref_id, 3168 text_before, 3169 number, 3170 text_after), 3171 match.end(0))) 3172 counters[counter] = number + 1 3173 for repl in reversed(replacements): 3174 text = text[:repl[0]] + repl[1] + text[repl[2]:] 3175 3176 # Second pass to replace the references with the right 3177 # value of the counter 3178 # Fwiw, it's vaguely annoying to have to turn the iterator into 3179 # a list and then reverse it but I can't think of a better thing to do. 3180 for match in reversed(list(regex_subs.finditer(text))): 3181 number, counter = references.get(match.group(1), (None, None)) 3182 if number is not None: 3183 repl = reference_html.format(counter, 3184 match.group(1), 3185 number) 3186 else: 3187 repl = reference_html.format(match.group(1), 3188 'countererror', 3189 '?' + match.group(1) + '?') 3190 if "smarty-pants" in self.md.extras: 3191 repl = repl.replace('"', self.md._escape_table['"']) 3192 3193 text = text[:match.start()] + repl + text[match.end():] 3194 return text 3195 3196 3197class PyShell(Extra): 3198 ''' 3199 Treats unindented Python interactive shell sessions as <code> 3200 blocks. 3201 ''' 3202 3203 name = 'pyshell' 3204 order = (), (Stage.LISTS,) 3205 3206 def test(self, text): 3207 return ">>>" in text 3208 3209 def sub(self, match: re.Match) -> str: 3210 if "fenced-code-blocks" in self.md.extras: 3211 dedented = _dedent(match.group(0)) 3212 return self.md.extra_classes['fenced-code-blocks'].run("```pycon\n" + dedented + "```\n") 3213 3214 lines = match.group(0).splitlines(0) 3215 _dedentlines(lines) 3216 indent = ' ' * self.md.tab_width 3217 s = ('\n' # separate from possible cuddled paragraph 3218 + indent + ('\n'+indent).join(lines) 3219 + '\n') 3220 return s 3221 3222 def run(self, text): 3223 less_than_tab = self.md.tab_width - 1 3224 _pyshell_block_re = re.compile(r""" 3225 ^([ ]{0,%d})>>>[ ].*\n # first line 3226 ^(\1[^\S\n]*\S.*\n)* # any number of subsequent lines with at least one character 3227 (?=^\1?\n|\Z) # ends with a blank line or end of document 3228 """ % less_than_tab, re.M | re.X) 3229 3230 return _pyshell_block_re.sub(self.sub, text) 3231 3232 3233class SmartyPants(Extra): 3234 ''' 3235 Replaces ' and " with curly quotation marks or curly 3236 apostrophes. Replaces --, ---, ..., and . . . with en dashes, em dashes, 3237 and ellipses. 3238 ''' 3239 name = 'smarty-pants' 3240 order = (), (Stage.SPAN_GAMUT,) 3241 3242 _opening_single_quote_re = re.compile(r"(?<!\S)'(?=\S)") 3243 _opening_double_quote_re = re.compile(r'(?<!\S)"(?=\S)') 3244 _closing_single_quote_re = re.compile(r"(?<=\S)'") 3245 _closing_double_quote_re = re.compile(r'(?<=\S)"(?=(\s|,|;|\.|\?|!|$))') 3246 # "smarty-pants" extra: Very liberal in interpreting a single prime as an 3247 # apostrophe; e.g. ignores the fact that "round", "bout", "twer", and 3248 # "twixt" can be written without an initial apostrophe. This is fine because 3249 # using scare quotes (single quotation marks) is rare. 3250 _apostrophe_year_re = re.compile(r"'(\d\d)(?=(\s|,|;|\.|\?|!|$))") 3251 _contractions = ["tis", "twas", "twer", "neath", "o", "n", 3252 "round", "bout", "twixt", "nuff", "fraid", "sup"] 3253 3254 3255 def contractions(self, text: str) -> str: 3256 text = self._apostrophe_year_re.sub(r"’\1", text) 3257 for c in self._contractions: 3258 text = text.replace("'%s" % c, "’%s" % c) 3259 text = text.replace("'%s" % c.capitalize(), 3260 "’%s" % c.capitalize()) 3261 return text 3262 3263 def run(self, text): 3264 """Fancifies 'single quotes', "double quotes", and apostrophes. 3265 Converts --, ---, and ... into en dashes, em dashes, and ellipses. 3266 3267 Inspiration is: <http://daringfireball.net/projects/smartypants/> 3268 See "test/tm-cases/smarty_pants.text" for a full discussion of the 3269 support here and 3270 <http://code.google.com/p/python-markdown2/issues/detail?id=42> for a 3271 discussion of some diversion from the original SmartyPants. 3272 """ 3273 if "'" in text: # guard for perf 3274 text = self.contractions(text) 3275 text = self._opening_single_quote_re.sub("‘", text) 3276 text = self._closing_single_quote_re.sub("’", text) 3277 3278 if '"' in text: # guard for perf 3279 text = self._opening_double_quote_re.sub("“", text) 3280 text = self._closing_double_quote_re.sub("”", text) 3281 3282 text = text.replace("---", "—") 3283 text = text.replace("--", "–") 3284 text = text.replace("...", "…") 3285 text = text.replace(" . . . ", "…") 3286 text = text.replace(". . .", "…") 3287 3288 # TODO: Temporary hack to fix https://github.com/trentm/python-markdown2/issues/150 3289 if "footnotes" in self.md.extras and "footnote-ref" in text: 3290 # Quotes in the footnote back ref get converted to "smart" quotes 3291 # Change them back here to ensure they work. 3292 text = text.replace('class="footnote-ref”', 'class="footnote-ref"') 3293 3294 return text 3295 3296 def test(self, text): 3297 return "'" in text or '"' in text 3298 3299 3300class Strike(Extra): 3301 ''' 3302 Text inside of double tilde is ~~strikethrough~~ 3303 ''' 3304 name = 'strike' 3305 order = (Stage.ITALIC_AND_BOLD,), () 3306 3307 _strike_re = re.compile(r"~~(?=\S)(.+?)(?<=\S)~~", re.S) 3308 3309 def run(self, text): 3310 return self._strike_re.sub(r"<s>\1</s>", text) 3311 3312 def test(self, text): 3313 return '~~' in text 3314 3315 3316class Tables(Extra): 3317 ''' 3318 Tables using the same format as GFM 3319 <https://help.github.com/articles/github-flavored-markdown#tables> and 3320 PHP-Markdown Extra <https://michelf.ca/projects/php-markdown/extra/#table>. 3321 ''' 3322 name = 'tables' 3323 order = (), (Stage.LISTS,) 3324 3325 def run(self, text): 3326 """Copying PHP-Markdown and GFM table syntax. Some regex borrowed from 3327 https://github.com/michelf/php-markdown/blob/lib/Michelf/Markdown.php#L2538 3328 """ 3329 less_than_tab = self.md.tab_width - 1 3330 table_re = re.compile(r''' 3331 (?:(?<=\n)|\A\n?) # leading blank line 3332 3333 ^[ ]{0,%d} # allowed whitespace 3334 (.*[|].*)[ ]*\n # $1: header row (at least one pipe) 3335 3336 ^[ ]{0,%d} # allowed whitespace 3337 ( # $2: underline row 3338 # underline row with leading bar 3339 (?: \|\ *:?-+:?\ * )+ \|? \s?[ ]*\n 3340 | 3341 # or, underline row without leading bar 3342 (?: \ *:?-+:?\ *\| )+ (?: \ *:?-+:?\ * )? \s?[ ]*\n 3343 ) 3344 3345 ( # $3: data rows 3346 (?: 3347 ^[ ]{0,%d}(?!\ ) # ensure line begins with 0 to less_than_tab spaces 3348 .*\|.*[ ]*\n 3349 )+ 3350 ) 3351 ''' % (less_than_tab, less_than_tab, less_than_tab), re.M | re.X) 3352 return table_re.sub(self.sub, text) 3353 3354 def sub(self, match: re.Match) -> str: 3355 trim_space_re = '^[ \t\n]+|[ \t\n]+$' 3356 trim_bar_re = r'^\||\|$' 3357 split_bar_re = r'^\||(?<![\`\\])\|' 3358 escape_bar_re = r'\\\|' 3359 3360 head, underline, body = match.groups() 3361 3362 # Determine aligns for columns. 3363 cols = [re.sub(escape_bar_re, '|', cell.strip()) for cell in re.split(split_bar_re, re.sub(trim_bar_re, "", re.sub(trim_space_re, "", underline)))] 3364 align_from_col_idx = {} 3365 for col_idx, col in enumerate(cols): 3366 if col[0] == ':' and col[-1] == ':': 3367 align_from_col_idx[col_idx] = ' style="text-align:center;"' 3368 elif col[0] == ':': 3369 align_from_col_idx[col_idx] = ' style="text-align:left;"' 3370 elif col[-1] == ':': 3371 align_from_col_idx[col_idx] = ' style="text-align:right;"' 3372 3373 # thead 3374 hlines = ['<table%s>' % self.md._html_class_str_from_tag('table'), '<thead%s>' % self.md._html_class_str_from_tag('thead'), '<tr>'] 3375 cols = [re.sub(escape_bar_re, '|', cell.strip()) for cell in re.split(split_bar_re, re.sub(trim_bar_re, "", re.sub(trim_space_re, "", head)))] 3376 for col_idx, col in enumerate(cols): 3377 hlines.append(' <th%s>%s</th>' % ( 3378 align_from_col_idx.get(col_idx, ''), 3379 self.md._run_span_gamut(col) 3380 )) 3381 hlines.append('</tr>') 3382 hlines.append('</thead>') 3383 3384 # tbody 3385 hlines.append('<tbody>') 3386 for line in body.strip('\n').split('\n'): 3387 hlines.append('<tr>') 3388 cols = [re.sub(escape_bar_re, '|', cell.strip()) for cell in re.split(split_bar_re, re.sub(trim_bar_re, "", re.sub(trim_space_re, "", line)))] 3389 for col_idx, col in enumerate(cols): 3390 hlines.append(' <td%s>%s</td>' % ( 3391 align_from_col_idx.get(col_idx, ''), 3392 self.md._run_span_gamut(col) 3393 )) 3394 hlines.append('</tr>') 3395 hlines.append('</tbody>') 3396 hlines.append('</table>') 3397 3398 return '\n'.join(hlines) + '\n' 3399 3400 3401class TelegramSpoiler(Extra): 3402 name = 'tg-spoiler' 3403 order = (), (Stage.ITALIC_AND_BOLD,) 3404 3405 _tg_spoiler_re = re.compile(r"\|\|\s?(.+?)\s?\|\|", re.S) 3406 3407 def run(self, text): 3408 return self._tg_spoiler_re.sub(r"<tg-spoiler>\1</tg-spoiler>", text) 3409 3410 def test(self, text): 3411 return '||' in text 3412 3413 3414class Underline(Extra): 3415 ''' 3416 Text inside of double dash is --underlined--. 3417 ''' 3418 name = 'underline' 3419 order = (Stage.ITALIC_AND_BOLD,), () 3420 3421 _underline_re = re.compile(r"(?<!<!)--(?!>)(?=\S)(.+?)(?<=\S)(?<!<!)--(?!>)", re.S) 3422 3423 def run(self, text): 3424 return self._underline_re.sub(r"<u>\1</u>", text) 3425 3426 def test(self, text): 3427 return '--' in text 3428 3429 3430class _WavedromExtraOpts(TypedDict, total=False): 3431 '''Options for the `Wavedrom` extra''' 3432 prefer_embed_svg: bool 3433 ''' 3434 Use the `wavedrom` library to convert diagrams to SVGs and embed them directly. 3435 This will only work if the `wavedrom` library has been installed. 3436 3437 Defaults to `True` 3438 ''' 3439 3440 3441class Wavedrom(Extra): 3442 ''' 3443 Support for generating Wavedrom digital timing diagrams 3444 ''' 3445 name = 'wavedrom' 3446 order = (Stage.CODE_BLOCKS, FencedCodeBlocks), () 3447 options: _WavedromExtraOpts 3448 3449 def test(self, text): 3450 match = FencedCodeBlocks.fenced_code_block_re.search(text) 3451 return match is None or match.group(2) == 'wavedrom' 3452 3453 def sub(self, match: re.Match) -> str: 3454 # dedent the block for processing 3455 lead_indent, waves = self.md._uniform_outdent(match.group(3)) 3456 # default tags to wrap the wavedrom block in 3457 open_tag, close_tag = '<script type="WaveDrom">\n', '</script>' 3458 3459 # check if the user would prefer to have the SVG embedded directly 3460 embed_svg = self.options.get('prefer_embed_svg', True) 3461 3462 if embed_svg: 3463 try: 3464 import wavedrom 3465 waves = wavedrom.render(waves).tostring() 3466 open_tag, close_tag = '<div>', '\n</div>' 3467 except ImportError: 3468 pass 3469 3470 # hash SVG to prevent <> chars being messed with 3471 self.md._escape_table[waves] = _hash_text(waves) 3472 3473 return self.md._uniform_indent( 3474 '\n%s%s%s\n' % (open_tag, self.md._escape_table[waves], close_tag), 3475 lead_indent, include_empty_lines=True 3476 ) 3477 3478 def run(self, text): 3479 return FencedCodeBlocks.fenced_code_block_re.sub(self.sub, text) 3480 3481 3482class WikiTables(Extra): 3483 ''' 3484 Google Code Wiki-style tables. See 3485 <http://code.google.com/p/support/wiki/WikiSyntax#Tables>. 3486 ''' 3487 name = 'wiki-tables' 3488 order = (Tables,), () 3489 3490 def run(self, text): 3491 less_than_tab = self.md.tab_width - 1 3492 wiki_table_re = re.compile(r''' 3493 (?:(?<=\n\n)|\A\n?) # leading blank line 3494 ^([ ]{0,%d})\|\|.+?\|\|[ ]*\n # first line 3495 (^\1\|\|.+?\|\|\n)* # any number of subsequent lines 3496 ''' % less_than_tab, re.M | re.X) 3497 return wiki_table_re.sub(self.sub, text) 3498 3499 def sub(self, match: re.Match) -> str: 3500 ttext = match.group(0).strip() 3501 rows = [] 3502 for line in ttext.splitlines(0): 3503 line = line.strip()[2:-2].strip() 3504 row = [c.strip() for c in re.split(r'(?<!\\)\|\|', line)] 3505 rows.append(row) 3506 3507 hlines = [] 3508 3509 def add_hline(line, indents=0): 3510 hlines.append((self.md.tab * indents) + line) 3511 3512 def format_cell(text): 3513 return self.md._run_span_gamut(re.sub(r"^\s*~", "", cell).strip(" ")) 3514 3515 add_hline('<table%s>' % self.md._html_class_str_from_tag('table')) 3516 # Check if first cell of first row is a header cell. If so, assume the whole row is a header row. 3517 if rows and rows[0] and re.match(r"^\s*~", rows[0][0]): 3518 add_hline('<thead%s>' % self.md._html_class_str_from_tag('thead'), 1) 3519 add_hline('<tr>', 2) 3520 for cell in rows[0]: 3521 add_hline("<th>{}</th>".format(format_cell(cell)), 3) 3522 add_hline('</tr>', 2) 3523 add_hline('</thead>', 1) 3524 # Only one header row allowed. 3525 rows = rows[1:] 3526 # If no more rows, don't create a tbody. 3527 if rows: 3528 add_hline('<tbody>', 1) 3529 for row in rows: 3530 add_hline('<tr>', 2) 3531 for cell in row: 3532 add_hline('<td>{}</td>'.format(format_cell(cell)), 3) 3533 add_hline('</tr>', 2) 3534 add_hline('</tbody>', 1) 3535 add_hline('</table>') 3536 return '\n'.join(hlines) + '\n' 3537 3538 def test(self, text): 3539 return '||' in text 3540 3541 3542# Register extras 3543Admonitions.register() 3544Alerts.register() 3545Breaks.register() 3546CodeFriendly.register() 3547FencedCodeBlocks.register() 3548Latex.register() 3549LinkPatterns.register() 3550MarkdownInHTML.register() 3551MiddleWordEm.register() 3552Mermaid.register() 3553Numbering.register() 3554PyShell.register() 3555SmartyPants.register() 3556Strike.register() 3557Tables.register() 3558TelegramSpoiler.register() 3559Underline.register() 3560Wavedrom.register() 3561WikiTables.register() 3562 3563 3564# ---------------------------------------------------------- 3565 3566 3567# ---- internal support functions 3568 3569 3570def calculate_toc_html(toc: Union[List[Tuple[int, str, str]], None]) -> Optional[str]: 3571 """Return the HTML for the current TOC. 3572 3573 This expects the `_toc` attribute to have been set on this instance. 3574 """ 3575 if toc is None: 3576 return None 3577 3578 def indent(): 3579 return ' ' * (len(h_stack) - 1) 3580 lines = [] 3581 h_stack = [0] # stack of header-level numbers 3582 for level, id, name in toc: 3583 if level > h_stack[-1]: 3584 lines.append("%s<ul>" % indent()) 3585 h_stack.append(level) 3586 elif level == h_stack[-1]: 3587 lines[-1] += "</li>" 3588 else: 3589 while level < h_stack[-1]: 3590 h_stack.pop() 3591 if not lines[-1].endswith("</li>"): 3592 lines[-1] += "</li>" 3593 lines.append("%s</ul></li>" % indent()) 3594 lines.append('%s<li><a href="#%s">%s</a>' % ( 3595 indent(), id, name)) 3596 while len(h_stack) > 1: 3597 h_stack.pop() 3598 if not lines[-1].endswith("</li>"): 3599 lines[-1] += "</li>" 3600 lines.append("%s</ul>" % indent()) 3601 return '\n'.join(lines) + '\n' 3602 3603 3604class UnicodeWithAttrs(str): 3605 """A subclass of unicode used for the return value of conversion to 3606 possibly attach some attributes. E.g. the "toc_html" attribute when 3607 the "toc" extra is used. 3608 """ 3609 metadata: Optional[Dict[str, str]] = None 3610 toc_html: Optional[str] = None 3611 3612## {{{ http://code.activestate.com/recipes/577257/ (r1) 3613_slugify_strip_re = re.compile(r'[^\w\s-]') 3614_slugify_hyphenate_re = re.compile(r'[-\s]+') 3615def _slugify(value: str) -> str: 3616 """ 3617 Normalizes string, converts to lowercase, removes non-alpha characters, 3618 and converts spaces to hyphens. 3619 3620 From Django's "django/template/defaultfilters.py". 3621 """ 3622 import unicodedata 3623 value = unicodedata.normalize('NFKD', value).encode('utf-8', 'ignore').decode() 3624 value = _slugify_strip_re.sub('', value).strip().lower() 3625 return _slugify_hyphenate_re.sub('-', value) 3626## end of http://code.activestate.com/recipes/577257/ }}} 3627 3628 3629# From http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52549 3630def _curry(function: Callable, *args, **kwargs) -> Callable: 3631 def result(*rest, **kwrest): 3632 combined = kwargs.copy() 3633 combined.update(kwrest) 3634 return function(*args + rest, **combined) 3635 return result 3636 3637 3638# Recipe: regex_from_encoded_pattern (1.0) 3639def _regex_from_encoded_pattern(s: str) -> re.Pattern: 3640 """'foo' -> re.compile(re.escape('foo')) 3641 '/foo/' -> re.compile('foo') 3642 '/foo/i' -> re.compile('foo', re.I) 3643 """ 3644 if s.startswith('/') and s.rfind('/') != 0: 3645 # Parse it: /PATTERN/FLAGS 3646 idx = s.rfind('/') 3647 _, flags_str = s[1:idx], s[idx+1:] 3648 flag_from_char = { 3649 "i": re.IGNORECASE, 3650 "l": re.LOCALE, 3651 "s": re.DOTALL, 3652 "m": re.MULTILINE, 3653 "u": re.UNICODE, 3654 } 3655 flags = 0 3656 for char in flags_str: 3657 try: 3658 flags |= flag_from_char[char] 3659 except KeyError: 3660 raise ValueError("unsupported regex flag: '%s' in '%s' " 3661 "(must be one of '%s')" 3662 % (char, s, ''.join(list(flag_from_char.keys())))) 3663 return re.compile(s[1:idx], flags) 3664 else: # not an encoded regex 3665 return re.compile(re.escape(s)) 3666 3667 3668# Recipe: dedent (0.1.2) 3669def _dedentlines(lines: List[str], tabsize: int = 8, skip_first_line: bool = False) -> List[str]: 3670 """_dedentlines(lines, tabsize=8, skip_first_line=False) -> dedented lines 3671 3672 "lines" is a list of lines to dedent. 3673 "tabsize" is the tab width to use for indent width calculations. 3674 "skip_first_line" is a boolean indicating if the first line should 3675 be skipped for calculating the indent width and for dedenting. 3676 This is sometimes useful for docstrings and similar. 3677 3678 Same as dedent() except operates on a sequence of lines. Note: the 3679 lines list is modified **in-place**. 3680 """ 3681 DEBUG = False 3682 if DEBUG: 3683 print("dedent: dedent(..., tabsize=%d, skip_first_line=%r)"\ 3684 % (tabsize, skip_first_line)) 3685 margin = None 3686 for i, line in enumerate(lines): 3687 if i == 0 and skip_first_line: 3688 continue 3689 indent = 0 3690 for ch in line: 3691 if ch == ' ': 3692 indent += 1 3693 elif ch == '\t': 3694 indent += tabsize - (indent % tabsize) 3695 elif ch in '\r\n': 3696 continue # skip all-whitespace lines 3697 else: 3698 break 3699 else: 3700 continue # skip all-whitespace lines 3701 if DEBUG: 3702 print("dedent: indent=%d: %r" % (indent, line)) 3703 if margin is None: 3704 margin = indent 3705 else: 3706 margin = min(margin, indent) 3707 if DEBUG: 3708 print("dedent: margin=%r" % margin) 3709 3710 if margin is not None and margin > 0: 3711 for i, line in enumerate(lines): 3712 if i == 0 and skip_first_line: 3713 continue 3714 removed = 0 3715 for j, ch in enumerate(line): 3716 if ch == ' ': 3717 removed += 1 3718 elif ch == '\t': 3719 removed += tabsize - (removed % tabsize) 3720 elif ch in '\r\n': 3721 if DEBUG: 3722 print("dedent: %r: EOL -> strip up to EOL" % line) 3723 lines[i] = lines[i][j:] 3724 break 3725 else: 3726 raise ValueError("unexpected non-whitespace char %r in " 3727 "line %r while removing %d-space margin" 3728 % (ch, line, margin)) 3729 if DEBUG: 3730 print("dedent: %r: %r -> removed %d/%d"\ 3731 % (line, ch, removed, margin)) 3732 if removed == margin: 3733 lines[i] = lines[i][j+1:] 3734 break 3735 elif removed > margin: 3736 lines[i] = ' '*(removed-margin) + lines[i][j+1:] 3737 break 3738 else: 3739 if removed: 3740 lines[i] = lines[i][removed:] 3741 return lines 3742 3743 3744def _dedent(text: str, tabsize: int = 8, skip_first_line: bool = False) -> str: 3745 """_dedent(text, tabsize=8, skip_first_line=False) -> dedented text 3746 3747 "text" is the text to dedent. 3748 "tabsize" is the tab width to use for indent width calculations. 3749 "skip_first_line" is a boolean indicating if the first line should 3750 be skipped for calculating the indent width and for dedenting. 3751 This is sometimes useful for docstrings and similar. 3752 3753 textwrap.dedent(s), but don't expand tabs to spaces 3754 """ 3755 lines = text.splitlines(True) 3756 _dedentlines(lines, tabsize=tabsize, skip_first_line=skip_first_line) 3757 return ''.join(lines) 3758 3759 3760class _memoized(object): 3761 """Decorator that caches a function's return value each time it is called. 3762 If called later with the same arguments, the cached value is returned, and 3763 not re-evaluated. 3764 3765 http://wiki.python.org/moin/PythonDecoratorLibrary 3766 """ 3767 def __init__(self, func): 3768 self.func = func 3769 self.cache = {} 3770 3771 def __call__(self, *args): 3772 try: 3773 return self.cache[args] 3774 except KeyError: 3775 self.cache[args] = value = self.func(*args) 3776 return value 3777 except TypeError: 3778 # uncachable -- for instance, passing a list as an argument. 3779 # Better to not cache than to blow up entirely. 3780 return self.func(*args) 3781 3782 def __repr__(self): 3783 """Return the function's docstring.""" 3784 return self.func.__doc__ 3785 3786 3787def _xml_oneliner_re_from_tab_width(tab_width: int) -> re.Pattern: 3788 """Standalone XML processing instruction regex.""" 3789 return re.compile(r""" 3790 (?: 3791 (?<=\n\n) # Starting after a blank line 3792 | # or 3793 \A\n? # the beginning of the doc 3794 ) 3795 ( # save in $1 3796 [ ]{0,%d} 3797 (?: 3798 <\?\w+\b\s+.*?\?> # XML processing instruction 3799 | 3800 <\w+:\w+\b\s+.*?/> # namespaced single tag 3801 ) 3802 [ \t]* 3803 (?=\n{2,}|\Z) # followed by a blank line or end of document 3804 ) 3805 """ % (tab_width - 1), re.X) 3806_xml_oneliner_re_from_tab_width = _memoized(_xml_oneliner_re_from_tab_width) 3807 3808 3809def _hr_tag_re_from_tab_width(tab_width: int) -> re.Pattern: 3810 return re.compile(r""" 3811 (?: 3812 (?<=\n\n) # Starting after a blank line 3813 | # or 3814 \A\n? # the beginning of the doc 3815 ) 3816 ( # save in \1 3817 [ ]{0,%d} 3818 <(hr) # start tag = \2 3819 \b # word break 3820 ([^<>])*? # 3821 /?> # the matching end tag 3822 [ \t]* 3823 (?=\n{2,}|\Z) # followed by a blank line or end of document 3824 ) 3825 """ % (tab_width - 1), re.X) 3826_hr_tag_re_from_tab_width = _memoized(_hr_tag_re_from_tab_width) 3827 3828 3829def _xml_escape_attr(attr: str, skip_single_quote: bool = True) -> str: 3830 """Escape the given string for use in an HTML/XML tag attribute. 3831 3832 By default this doesn't bother with escaping `'` to `'`, presuming that 3833 the tag attribute is surrounded by double quotes. 3834 """ 3835 escaped = _AMPERSAND_RE.sub('&', attr) 3836 3837 escaped = (attr 3838 .replace('"', '"') 3839 .replace('<', '<') 3840 .replace('>', '>')) 3841 if not skip_single_quote: 3842 escaped = escaped.replace("'", "'") 3843 return escaped 3844 3845 3846def _xml_encode_email_char_at_random(ch: str) -> str: 3847 r = random() 3848 # Roughly 10% raw, 45% hex, 45% dec. 3849 # '@' *must* be encoded. I [John Gruber] insist. 3850 # Issue 26: '_' must be encoded. 3851 if r > 0.9 and ch not in "@_": 3852 return ch 3853 elif r < 0.45: 3854 # The [1:] is to drop leading '0': 0x63 -> x63 3855 return '&#%s;' % hex(ord(ch))[1:] 3856 else: 3857 return '&#%s;' % ord(ch) 3858 3859 3860def _html_escape_url( 3861 attr: str, 3862 safe_mode: Union[_safe_mode, bool, None] = False, 3863 charset: Optional[str] = None 3864): 3865 """ 3866 Replace special characters that are potentially malicious in url string. 3867 3868 Args: 3869 charset: don't escape characters from this charset. Currently the only 3870 exception is for '+' when charset=='base64' 3871 """ 3872 escaped = (attr 3873 .replace('"', '"') 3874 .replace('<', '<') 3875 .replace('>', '>')) 3876 if safe_mode: 3877 if charset != 'base64': 3878 escaped = escaped.replace('+', ' ') 3879 escaped = escaped.replace("'", "'") 3880 return escaped 3881 3882 3883# ---- mainline 3884 3885class _NoReflowFormatter(argparse.RawDescriptionHelpFormatter): 3886 """An argparse formatter that does NOT reflow the description.""" 3887 def format_description(self, description): 3888 return description or "" 3889 3890 3891def _test(): 3892 import doctest 3893 doctest.testmod() 3894 3895 3896def main(argv=None): 3897 if argv is None: 3898 argv = sys.argv 3899 if not logging.root.handlers: 3900 logging.basicConfig() 3901 3902 parser = argparse.ArgumentParser( 3903 prog="markdown2", description=cmdln_desc, usage='%(prog)s [PATHS...]', 3904 formatter_class=_NoReflowFormatter 3905 ) 3906 parser.add_argument('--version', action='version', 3907 version='%(prog)s {version}'.format(version=__version__)) 3908 parser.add_argument('paths', nargs='*', 3909 help=( 3910 'optional list of files to convert.' 3911 'If none are given, stdin will be used' 3912 )) 3913 parser.add_argument("-v", "--verbose", dest="log_level", 3914 action="store_const", const=logging.DEBUG, 3915 help="more verbose output") 3916 parser.add_argument("--encoding", 3917 help="specify encoding of text content") 3918 parser.add_argument("--html4tags", action="store_true", default=False, 3919 help="use HTML 4 style for empty element tags") 3920 parser.add_argument("-s", "--safe", metavar="MODE", dest="safe_mode", 3921 help="sanitize literal HTML: 'escape' escapes " 3922 "HTML meta chars, 'replace' replaces with an " 3923 "[HTML_REMOVED] note") 3924 parser.add_argument("-x", "--extras", action="append", 3925 help="Turn on specific extra features (not part of " 3926 "the core Markdown spec). See above.") 3927 parser.add_argument("--use-file-vars", 3928 help="Look for and use Emacs-style 'markdown-extras' " 3929 "file var to turn on extras. See " 3930 "<https://github.com/trentm/python-markdown2/wiki/Extras>") 3931 parser.add_argument("--link-patterns-file", 3932 help="path to a link pattern file") 3933 parser.add_argument("--self-test", action="store_true", 3934 help="run internal self-tests (some doctests)") 3935 parser.add_argument("--compare", action="store_true", 3936 help="run against Markdown.pl as well (for testing)") 3937 parser.set_defaults(log_level=logging.INFO, compare=False, 3938 encoding="utf-8", safe_mode=None, use_file_vars=False) 3939 opts = parser.parse_args() 3940 paths = opts.paths 3941 log.setLevel(opts.log_level) 3942 3943 if opts.self_test: 3944 return _test() 3945 3946 if opts.extras: 3947 extras = {} 3948 for s in opts.extras: 3949 splitter = re.compile("[,;: ]+") 3950 for e in splitter.split(s): 3951 if '=' in e: 3952 ename, earg = e.split('=', 1) 3953 try: 3954 earg = int(earg) 3955 except ValueError: 3956 pass 3957 else: 3958 ename, earg = e, None 3959 extras[ename] = earg 3960 else: 3961 extras = None 3962 3963 if opts.link_patterns_file: 3964 link_patterns = [] 3965 f = open(opts.link_patterns_file) 3966 try: 3967 for i, line in enumerate(f.readlines()): 3968 if not line.strip(): 3969 continue 3970 if line.lstrip().startswith("#"): 3971 continue 3972 try: 3973 pat, href = line.rstrip().rsplit(None, 1) 3974 except ValueError: 3975 raise MarkdownError("%s:%d: invalid link pattern line: %r" 3976 % (opts.link_patterns_file, i+1, line)) 3977 link_patterns.append( 3978 (_regex_from_encoded_pattern(pat), href)) 3979 finally: 3980 f.close() 3981 else: 3982 link_patterns = None 3983 3984 from os.path import abspath, dirname, exists, join 3985 markdown_pl = join(dirname(dirname(abspath(__file__))), "test", 3986 "Markdown.pl") 3987 if not paths: 3988 paths = ['-'] 3989 for path in paths: 3990 if path == '-': 3991 text = sys.stdin.read() 3992 else: 3993 fp = codecs.open(path, 'r', opts.encoding) 3994 text = fp.read() 3995 fp.close() 3996 if opts.compare: 3997 from subprocess import PIPE, Popen 3998 print("==== Markdown.pl ====") 3999 p = Popen('perl %s' % markdown_pl, shell=True, stdin=PIPE, stdout=PIPE, close_fds=True) 4000 p.stdin.write(text.encode('utf-8')) 4001 p.stdin.close() 4002 perl_html = p.stdout.read().decode('utf-8') 4003 sys.stdout.write(perl_html) 4004 print("==== markdown2.py ====") 4005 html = markdown(text, 4006 html4tags=opts.html4tags, 4007 safe_mode=opts.safe_mode, 4008 extras=extras, link_patterns=link_patterns, 4009 use_file_vars=opts.use_file_vars, 4010 cli=True) 4011 sys.stdout.write(html) 4012 if extras and "toc" in extras: 4013 log.debug("toc_html: " + 4014 str(html.toc_html.encode(sys.stdout.encoding or "utf-8", 'xmlcharrefreplace'))) 4015 if opts.compare: 4016 test_dir = join(dirname(dirname(abspath(__file__))), "test") 4017 if exists(join(test_dir, "test_markdown2.py")): 4018 sys.path.insert(0, test_dir) 4019 from test_markdown2 import norm_html_from_html 4020 norm_html = norm_html_from_html(html) 4021 norm_perl_html = norm_html_from_html(perl_html) 4022 else: 4023 norm_html = html 4024 norm_perl_html = perl_html 4025 print("==== match? %r ====" % (norm_perl_html == norm_html)) 4026 4027 4028if __name__ == "__main__": 4029 sys.exit(main(sys.argv))(builtins.Exception): Common base class for all non-exit exceptions.
Inherited Members
- builtins.Exception
- Exception
- builtins.BaseException
- with_traceback
- add_note
- args
(enum.IntEnum): 216class Stage(IntEnum): 217 PREPROCESS = auto() 218 HASH_HTML = auto() 219 LINK_DEFS = auto() 220 221 BLOCK_GAMUT = auto() 222 HEADERS = auto() 223 LISTS = auto() 224 CODE_BLOCKS = auto() 225 BLOCK_QUOTES = auto() 226 PARAGRAPHS = auto() 227 228 SPAN_GAMUT = auto() 229 CODE_SPANS = auto() 230 ESCAPE_SPECIAL = auto() 231 LINKS = auto() # and auto links 232 ITALIC_AND_BOLD = auto() 233 234 POSTPROCESS = auto() 235 UNHASH_HTML = auto()PREPROCESS = <Stage.PREPROCESS: 1>HASH_HTML = <Stage.HASH_HTML: 2>LINK_DEFS = <Stage.LINK_DEFS: 3>BLOCK_GAMUT = <Stage.BLOCK_GAMUT: 4>HEADERS = <Stage.HEADERS: 5>LISTS = <Stage.LISTS: 6>CODE_BLOCKS = <Stage.CODE_BLOCKS: 7>BLOCK_QUOTES = <Stage.BLOCK_QUOTES: 8>PARAGRAPHS = <Stage.PARAGRAPHS: 9>SPAN_GAMUT = <Stage.SPAN_GAMUT: 10>CODE_SPANS = <Stage.CODE_SPANS: 11>ESCAPE_SPECIAL = <Stage.ESCAPE_SPECIAL: 12>LINKS = <Stage.LINKS: 13>ITALIC_AND_BOLD = <Stage.ITALIC_AND_BOLD: 14>POSTPROCESS = <Stage.POSTPROCESS: 15>UNHASH_HTML = <Stage.UNHASH_HTML: 16>Inherited Members
- enum.Enum
- name
- value
- builtins.int
- conjugate
- bit_length
- bit_count
- to_bytes
- from_bytes
- as_integer_ratio
- is_integer
- real
- imag
- numerator
- denominator
Convert the given text.
@mark_stage(Stage.POSTPROCESS)def postprocess(self, text: str) -> str:571 @mark_stage(Stage.POSTPROCESS) 572 def postprocess(self, text: str) -> str: 573 """A hook for subclasses to do some postprocessing of the html, if 574 desired. This is called before unescaping of special chars and 575 unhashing of raw HTML spans. 576 """ 577 return textA hook for subclasses to do some postprocessing of the html, if desired. This is called before unescaping of special chars and unhashing of raw HTML spans.
@mark_stage(Stage.PREPROCESS)def preprocess(self, text: str) -> str:579 @mark_stage(Stage.PREPROCESS) 580 def preprocess(self, text: str) -> str: 581 """A hook for subclasses to do some preprocessing of the Markdown, if 582 desired. This is called after basic formatting of the text, but prior 583 to any extras, safe mode, etc. processing. 584 """ 585 return textA hook for subclasses to do some preprocessing of the Markdown, if desired. This is called after basic formatting of the text, but prior to any extras, safe mode, etc. processing.
def header_id_from_text(self, text: str, prefix: str, n: Optional[int] = None) -> str:1701 def header_id_from_text(self, 1702 text: str, 1703 prefix: str, 1704 n: Optional[int] = None 1705 ) -> str: 1706 """Generate a header id attribute value from the given header 1707 HTML content. 1708 1709 This is only called if the "header-ids" extra is enabled. 1710 Subclasses may override this for different header ids. 1711 1712 @param text {str} The text of the header tag 1713 @param prefix {str} The requested prefix for header ids. This is the 1714 value of the "header-ids" extra key, if any. Otherwise, None. 1715 @param n {int} (unused) The <hN> tag number, i.e. `1` for an <h1> tag. 1716 @returns {str} The value for the header tag's "id" attribute. Return 1717 None to not have an id attribute and to exclude this header from 1718 the TOC (if the "toc" extra is specified). 1719 """ 1720 header_id = _slugify(text) 1721 if prefix and isinstance(prefix, str): 1722 header_id = prefix + '-' + header_id 1723 1724 self._count_from_header_id[header_id] += 1 1725 if 0 == len(header_id) or self._count_from_header_id[header_id] > 1: 1726 header_id += '-%s' % self._count_from_header_id[header_id] 1727 1728 return header_idGenerate a header id attribute value from the given header HTML content.
This is only called if the "header-ids" extra is enabled. Subclasses may override this for different header ids.
@param text {str} The text of the header tag @param prefix {str} The requested prefix for header ids. This is the value of the "header-ids" extra key, if any. Otherwise, None. @param n {int} (unused) The
tag number, i.e. 1for antag. @returns {str} The value for the header tag's "id" attribute. Return None to not have an id attribute and to exclude this header from the TOC (if the "toc" extra is specified).
A markdowner class that enables most extras:
- footnotes
- fenced-code-blocks (only highlights code if 'pygments' Python module on path)
These are not included:
- pyshell (specific to Python-related documenting)
- code-friendly (because it disables part of the syntax)
- link-patterns (because you need to specify some actual link-patterns anyway)
Inherited Members
(abc.ABC): 2540class Extra(ABC): 2541 _registry: Dict[str, Type['Extra']] = {} 2542 _exec_order: Dict[Stage, Tuple[List[Type['Extra']], List[Type['Extra']]]] = {} 2543 2544 name: str 2545 ''' 2546 An identifiable name that users can use to invoke the extra 2547 in the Markdown class 2548 ''' 2549 order: Tuple[Collection[Union[Stage, Type['Extra']]], Collection[Union[Stage, Type['Extra']]]] 2550 ''' 2551 Tuple of two iterables containing the stages/extras this extra will run before and 2552 after, respectively 2553 ''' 2554 2555 def __init__(self, md: Markdown, options: Optional[dict]): 2556 ''' 2557 Args: 2558 md: An instance of `Markdown` 2559 options: a dict of settings to alter the extra's behaviour 2560 ''' 2561 self.md = md 2562 self.options = options if options is not None else {} 2563 2564 @classmethod 2565 def deregister(cls): 2566 ''' 2567 Removes the class from the extras registry and unsets its execution order. 2568 ''' 2569 if cls.name in cls._registry: 2570 del cls._registry[cls.name] 2571 2572 for exec_order in Extra._exec_order.values(): 2573 # find everywhere this extra is mentioned and remove it 2574 for section in exec_order: 2575 while cls in section: 2576 section.remove(cls) 2577 2578 @classmethod 2579 def register(cls): 2580 ''' 2581 Registers the class for use with `Markdown` and calculates its execution order based on 2582 the `order` class attribute. 2583 ''' 2584 cls._registry[cls.name] = cls 2585 2586 for index, item in enumerate((*cls.order[0], *cls.order[1])): 2587 before = index < len(cls.order[0]) 2588 if not isinstance(item, Stage) and issubclass(item, Extra): 2589 # eg: FencedCodeBlocks 2590 for exec_orders in Extra._exec_order.values(): 2591 # insert this extra everywhere the other one is mentioned 2592 for section in exec_orders: 2593 if item in section: 2594 to_index = section.index(item) 2595 if not before: 2596 to_index += 1 2597 section.insert(to_index, cls) 2598 else: 2599 # eg: Stage.PREPROCESS 2600 Extra._exec_order.setdefault(item, ([], [])) 2601 if cls in Extra._exec_order[item][0 if before else 1]: 2602 # extra is already runnig after this stage. Don't duplicate that effort 2603 continue 2604 if before: 2605 Extra._exec_order[item][0].insert(0, cls) 2606 else: 2607 Extra._exec_order[item][1].append(cls) 2608 2609 @abstractmethod 2610 def run(self, text: str) -> str: 2611 ''' 2612 Run the extra against the given text. 2613 2614 Returns: 2615 The new text after being modified by the extra 2616 ''' 2617 ... 2618 2619 def test(self, text: str) -> bool: 2620 ''' 2621 Check a section of markdown to see if this extra should be run upon it. 2622 The default implementation will always return True but it's recommended to override 2623 this behaviour to improve performance. 2624 ''' 2625 return TrueHelper class that provides a standard way to create an ABC using inheritance.
Extra(md: Markdown, options: Optional[dict])2555 def __init__(self, md: Markdown, options: Optional[dict]): 2556 ''' 2557 Args: 2558 md: An instance of `Markdown` 2559 options: a dict of settings to alter the extra's behaviour 2560 ''' 2561 self.md = md 2562 self.options = options if options is not None else {}Args: md: An instance of
Markdownoptions: a dict of settings to alter the extra's behaviourTuple of two iterables containing the stages/extras this extra will run before and after, respectively
@classmethoddef deregister(cls):2564 @classmethod 2565 def deregister(cls): 2566 ''' 2567 Removes the class from the extras registry and unsets its execution order. 2568 ''' 2569 if cls.name in cls._registry: 2570 del cls._registry[cls.name] 2571 2572 for exec_order in Extra._exec_order.values(): 2573 # find everywhere this extra is mentioned and remove it 2574 for section in exec_order: 2575 while cls in section: 2576 section.remove(cls)Removes the class from the extras registry and unsets its execution order.
@classmethoddef register(cls):2578 @classmethod 2579 def register(cls): 2580 ''' 2581 Registers the class for use with `Markdown` and calculates its execution order based on 2582 the `order` class attribute. 2583 ''' 2584 cls._registry[cls.name] = cls 2585 2586 for index, item in enumerate((*cls.order[0], *cls.order[1])): 2587 before = index < len(cls.order[0]) 2588 if not isinstance(item, Stage) and issubclass(item, Extra): 2589 # eg: FencedCodeBlocks 2590 for exec_orders in Extra._exec_order.values(): 2591 # insert this extra everywhere the other one is mentioned 2592 for section in exec_orders: 2593 if item in section: 2594 to_index = section.index(item) 2595 if not before: 2596 to_index += 1 2597 section.insert(to_index, cls) 2598 else: 2599 # eg: Stage.PREPROCESS 2600 Extra._exec_order.setdefault(item, ([], [])) 2601 if cls in Extra._exec_order[item][0 if before else 1]: 2602 # extra is already runnig after this stage. Don't duplicate that effort 2603 continue 2604 if before: 2605 Extra._exec_order[item][0].insert(0, cls) 2606 else: 2607 Extra._exec_order[item][1].append(cls)@abstractmethoddef run(self, text: str) -> str:2609 @abstractmethod 2610 def run(self, text: str) -> str: 2611 ''' 2612 Run the extra against the given text. 2613 2614 Returns: 2615 The new text after being modified by the extra 2616 ''' 2617 ...Run the extra against the given text.
Returns: The new text after being modified by the extra
def test(self, text: str) -> bool:2619 def test(self, text: str) -> bool: 2620 ''' 2621 Check a section of markdown to see if this extra should be run upon it. 2622 The default implementation will always return True but it's recommended to override 2623 this behaviour to improve performance. 2624 ''' 2625 return TrueCheck a section of markdown to see if this extra should be run upon it. The default implementation will always return True but it's recommended to override this behaviour to improve performance.
An ABC that provides hooks for dealing with italics and bold syntax. This class is set to trigger both before AND after the italics and bold stage. This allows any child classes to intercept instances of bold or italic syntax and change the output or hash it to prevent it from being processed.
After the I&B stage any hashes in the
hash_tablesinstance variable are replaced.name = 'italic-and-bold-processor'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((<Stage.ITALIC_AND_BOLD: 14>,), (<Stage.ITALIC_AND_BOLD: 14>,))Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):2647 def run(self, text): 2648 if self.md.order < Stage.ITALIC_AND_BOLD: 2649 text = self.strong_re.sub(self.sub, text) 2650 text = self.em_re.sub(self.sub, text) 2651 else: 2652 # push any hashed values back, using a while loop to deal with recursive hashes 2653 orig_text = '' 2654 while orig_text != text: 2655 orig_text = text 2656 for key, substr in self.hash_table.items(): 2657 text = text.replace(key, substr) 2658 return textRun the extra against the given text.
Returns: The new text after being modified by the extra
def test(self, text):2671 def test(self, text): 2672 if self.md.order < Stage.ITALIC_AND_BOLD: 2673 return '*' in text or '_' in text 2674 return self.hash_table and re.search(r'md5-[0-9a-z]{32}', text)Check a section of markdown to see if this extra should be run upon it. The default implementation will always return True but it's recommended to override this behaviour to improve performance.
Inherited Members
Enable parsing of RST admonitions
name = 'admonitions'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((<Stage.BLOCK_GAMUT: 4>, <Stage.LINK_DEFS: 3>), ())Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
admonitions_re = re.compile('\n ^(\\ *)\\.\\.\\ (admonition|attention|caution|danger|error|hint|important|note|tip|warning)::\\ * # $1 leading indent, $2 the admonition\n (.*)? , re.IGNORECASE|re.MULTILINE|re.VERBOSE)def test(self, text):Check a section of markdown to see if this extra should be run upon it. The default implementation will always return True but it's recommended to override this behaviour to improve performance.
def sub(self, match: re.Match) -> str:2702 def sub(self, match: re.Match) -> str: 2703 lead_indent, admonition_name, title, body = match.groups() 2704 2705 admonition_type = '<strong>%s</strong>' % admonition_name 2706 2707 # figure out the class names to assign the block 2708 if admonition_name.lower() == 'admonition': 2709 admonition_class = 'admonition' 2710 else: 2711 admonition_class = 'admonition %s' % admonition_name.lower() 2712 2713 # titles are generally optional 2714 if title: 2715 title = '<em>%s</em>' % title 2716 2717 # process the admonition body like regular markdown 2718 body = self.md._run_block_gamut("\n%s\n" % self.md._uniform_outdent(body)[1]) 2719 2720 # indent the body before placing inside the aside block 2721 admonition = self.md._uniform_indent( 2722 '%s\n%s\n\n%s\n' % (admonition_type, title, body), 2723 self.md.tab, False 2724 ) 2725 # wrap it in an aside 2726 admonition = '<aside class="%s">\n%s</aside>' % (admonition_class, admonition) 2727 # now indent the whole admonition back to where it started 2728 return self.md._uniform_indent(admonition, lead_indent, False)Markdown Alerts as per https://docs.github.com/en/get-started/writing-on-github/getting-started-with-writing-and-formatting-on-github/basic-writing-and-formatting-syntax#alerts
order = ((), (<Stage.BLOCK_QUOTES: 8>,))Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
alert_re = re.compile('\n <blockquote>\\s*\n <p>\n \\[!(?P<type>NOTE|TIP|IMPORTANT|WARNING|CAUTION)\\]\n (?P<closing_tag></p>[ \\t]*\\n?)?\n (?P<contents>[\\s\\S]+?)\n </blockquote, re.VERBOSE)def test(self, text):Check a section of markdown to see if this extra should be run upon it. The default implementation will always return True but it's recommended to override this behaviour to improve performance.
def sub(self, match: re.Match) -> str:2756 def sub(self, match: re.Match) -> str: 2757 typ = match["type"].lower() 2758 heading = f"<em>{match['type'].title()}</em>" 2759 contents = match["contents"].strip() 2760 if match["closing_tag"]: 2761 return f'<div class="alert {typ}">\n{heading}\n{contents}\n</div>' 2762 else: 2763 return f'<div class="alert {typ}">\n{heading}\n<p>{contents}\n</div>'Helper class that provides a standard way to create an ABC using inheritance.
order = ((), (<Stage.ITALIC_AND_BOLD: 14>,))Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):2782 def run(self, text): 2783 on_backslash = self.options.get('on_backslash', False) 2784 on_newline = self.options.get('on_newline', False) 2785 2786 if on_backslash and on_newline: 2787 pattern = r' *\\?' 2788 elif on_backslash: 2789 pattern = r'(?: *\\| {2,})' 2790 elif on_newline: 2791 pattern = r' *' 2792 else: 2793 pattern = r' {2,}' 2794 2795 break_tag = "<br%s\n" % self.md.empty_element_suffix 2796 text = re.sub(pattern + r"\n(?!\<(?:\/?(ul|ol|li))\>)", break_tag, text) 2797 2798 return textRun the extra against the given text.
Returns: The new text after being modified by the extra
Disable _ and __ for em and strong.
name = 'code-friendly'An identifiable name that users can use to invoke the extra in the Markdown class
def sub(self, match: re.Match) -> str:2807 def sub(self, match: re.Match) -> str: 2808 syntax = match.group(1) 2809 text: str = match.string[match.start(): match.end()] 2810 if '_' in syntax: 2811 # if using _this_ syntax, hash the whole thing so that it doesn't get processed 2812 key = _hash_text(text) 2813 self.hash_table[key] = text 2814 return key 2815 elif '_' in text: 2816 # if the text within the bold/em markers contains '_' then hash those contents to protect them from em_re 2817 text = text[len(syntax): -len(syntax)] 2818 key = _hash_text(text) 2819 self.hash_table[key] = text 2820 return syntax + key + syntax 2821 # if no underscores are present, the text is fine and we can just leave it alone 2822 return super().sub(match)Inherited Members
Allows a code block to not have to be indented by fencing it with '```' on a line before and after. Based on http://github.github.com/github-flavored-markdown/ with support for syntax highlighting.
name = 'fenced-code-blocks'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((<Stage.LINK_DEFS: 3>, <Stage.BLOCK_GAMUT: 4>), (<Stage.PREPROCESS: 1>,))Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
fenced_code_block_re = re.compile('\n (?:\\n+|\\A\\n?|(?<=\\n))\n (^[ \\t]*`{3,})\\s{0,99}?([\\w+-]+)?\\s{0,99}?\\n # $1 = opening fence (captured for back-referencing), $2 = optional lang\n (.*?) , re.MULTILINE|re.DOTALL|re.VERBOSE)def test(self, text):2843 def test(self, text): 2844 if '```' not in text: 2845 return False 2846 if self.md.stage == Stage.PREPROCESS and not self.md.safe_mode: 2847 return True 2848 if self.md.stage == Stage.LINK_DEFS and self.md.safe_mode: 2849 return True 2850 return self.md.stage == Stage.BLOCK_GAMUTCheck a section of markdown to see if this extra should be run upon it. The default implementation will always return True but it's recommended to override this behaviour to improve performance.
def sub(self, match: re.Match) -> str:2905 def sub(self, match: re.Match) -> str: 2906 lexer_name = match.group(2) 2907 codeblock = match.group(3) 2908 codeblock = codeblock[:-1] # drop one trailing newline 2909 2910 # Use pygments only if not using the highlightjs-lang extra 2911 if lexer_name and "highlightjs-lang" not in self.md.extras: 2912 lexer = self.md._get_pygments_lexer(lexer_name) 2913 if lexer: 2914 leading_indent = ' '*(len(match.group(1)) - len(match.group(1).lstrip())) 2915 return self._code_block_with_lexer_sub(codeblock, leading_indent, lexer) 2916 2917 # Fenced code blocks need to be outdented before encoding, and then reapplied 2918 leading_indent = ' ' * (len(match.group(1)) - len(match.group(1).lstrip())) 2919 if codeblock: 2920 # only run the codeblock through the outdenter if not empty 2921 leading_indent, codeblock = self.md._uniform_outdent(codeblock, max_outdent=leading_indent) 2922 2923 codeblock = self.md._encode_code(codeblock) 2924 2925 tags = self.tags(lexer_name) 2926 2927 return "\n%s%s%s\n%s%s\n" % (leading_indent, tags[0], codeblock, leading_indent, tags[1])Convert $ and $$ to tags for inline and block math.
order = ((<Stage.CODE_BLOCKS: 7>, <class 'FencedCodeBlocks'>), ())Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):2962 def run(self, text): 2963 try: 2964 import latex2mathml.converter 2965 self.converter = latex2mathml.converter 2966 except ImportError: 2967 raise ImportError('The "latex" extra requires the "latex2mathml" package to be installed.') 2968 2969 # Escape by replacing with a code block 2970 text = self._pre_code_block_re.sub(self.code_placeholder, text) 2971 text = self._single_re.sub(self.code_placeholder, text) 2972 text = self._triple_re.sub(self.code_placeholder, text) 2973 2974 text = self._single_dollar_re.sub(self._convert_single_match, text) 2975 text = self._double_dollar_re.sub(self._convert_double_match, text) 2976 2977 # Convert placeholder tag back to original code 2978 for placeholder, code_block in self.code_blocks.items(): 2979 text = text.replace(placeholder, code_block) 2980 2981 return textRun the extra against the given text.
Returns: The new text after being modified by the extra
Auto-link given regex patterns in text (e.g. bug number references, revision number references).
name = 'link-patterns'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((<Stage.LINKS: 13>,), ())Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
options: collections.abc.Iterable[typing.Tuple[re.Pattern, typing.Union[str, typing.Callable[[re.Match], str]]]]def run(self, text):2995 def run(self, text): 2996 link_from_hash = {} 2997 for regex, repl in self.options: 2998 replacements = [] 2999 for match in regex.finditer(text): 3000 if any(self.md._match_overlaps_substr(text, match, h) for h in link_from_hash): 3001 continue 3002 3003 if callable(repl): 3004 href = repl(match) 3005 else: 3006 href = match.expand(repl) 3007 replacements.append((match.span(), href)) 3008 for (start, end), href in reversed(replacements): 3009 3010 # Do not match against links inside brackets. 3011 if text[start - 1:start] == '[' and text[end:end + 1] == ']': 3012 continue 3013 3014 # Do not match against links in the standard markdown syntax. 3015 if text[start - 2:start] == '](' or text[end:end + 2] == '")': 3016 continue 3017 3018 # Do not match against links which are escaped. 3019 if text[start - 3:start] == '"""' and text[end:end + 3] == '"""': 3020 text = text[:start - 3] + text[start:end] + text[end + 3:] 3021 continue 3022 3023 # search the text for anything that looks like a link 3024 is_inside_link = False 3025 for link_re in (self.md._auto_link_re, self._basic_link_re): 3026 for match in link_re.finditer(text): 3027 if any((r[0] <= start and end <= r[1]) for r in match.regs): 3028 # if the link pattern start and end pos is within the bounds of 3029 # something that looks like a link, then don't process it 3030 is_inside_link = True 3031 break 3032 else: 3033 continue 3034 break 3035 3036 if is_inside_link: 3037 continue 3038 3039 escaped_href = ( 3040 href.replace('"', '"') # b/c of attr quote 3041 # To avoid markdown <em> and <strong>: 3042 .replace('*', self.md._escape_table['*']) 3043 .replace('_', self.md._escape_table['_'])) 3044 link = '<a href="%s">%s</a>' % (escaped_href, text[start:end]) 3045 hash = _hash_text(link) 3046 link_from_hash[hash] = link 3047 text = text[:start] + hash + text[end:] 3048 for hash, link in list(link_from_hash.items()): 3049 text = text.replace(hash, link) 3050 return textRun the extra against the given text.
Returns: The new text after being modified by the extra
def test(self, text):Check a section of markdown to see if this extra should be run upon it. The default implementation will always return True but it's recommended to override this behaviour to improve performance.
Inherited Members
Allow the use of
markdown="1"in a block HTML tag to have markdown processing be done on its contents. Similar to http://michelf.com/projects/php-markdown/extra/#markdown-attr but with some limitations.name = 'markdown-in-html'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((), (<Stage.HASH_HTML: 2>,))Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):3066 def run(self, text): 3067 def callback(block): 3068 indent, block = self.md._uniform_outdent(block) 3069 block = self.md._hash_html_block_sub(block) 3070 block = self.md._uniform_indent(block, indent, include_empty_lines=True, indent_empty_lines=False) 3071 return block 3072 3073 return self.md._strict_tag_block_sub(text, self.md._block_tags_a, callback, True)Run the extra against the given text.
Returns: The new text after being modified by the extra
Allows a code block to not have to be indented by fencing it with '```' on a line before and after. Based on http://github.github.com/github-flavored-markdown/ with support for syntax highlighting.
order = ((<class 'FencedCodeBlocks'>,), ())Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
Inherited Members
Allows or disallows emphasis syntax in the middle of words, defaulting to allow. Disabling this means that
this_text_herewill not be converted tothis<em>text</em>here.MiddleWordEm(md: Markdown, options: Union[dict, bool])3098 def __init__(self, md: Markdown, options: Union[dict, bool]): 3099 ''' 3100 Args: 3101 md: the markdown instance 3102 options: can be bool for backwards compatibility but will be converted to a dict 3103 in the constructor. All options are: 3104 - allowed (bool): whether to allow emphasis in the middle of a word. 3105 If `options` is a bool it will be placed under this key. 3106 ''' 3107 if isinstance(options, bool): 3108 options = {'allowed': options} 3109 options.setdefault('allowed', True) 3110 super().__init__(md, options) 3111 3112 self.liberal_em_re = self.em_re 3113 if not options['allowed']: 3114 self.em_re = re.compile(r'(?<=\b)%s(?=\b)' % self.liberal_em_re.pattern, self.liberal_em_re.flags)Args: md: the markdown instance options: can be bool for backwards compatibility but will be converted to a dict in the constructor. All options are: - allowed (bool): whether to allow emphasis in the middle of a word. If
optionsis a bool it will be placed under this key.name = 'middle-word-em'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((<class 'CodeFriendly'>,), (<Stage.ITALIC_AND_BOLD: 14>,))Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):3116 def run(self, text): 3117 # run strong and whatnot first 3118 # this also will process all strict ems 3119 text = super().run(text) 3120 if self.md.order < self.md.stage: 3121 # hash all non-valid ems 3122 text = self.liberal_em_re.sub(self.sub_hash, text) 3123 return textRun the extra against the given text.
Returns: The new text after being modified by the extra
Inherited Members
Support of generic counters. Non standard extension to allow sequential numbering of figures, tables, equations, exhibits etc.
name = 'numbering'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((<Stage.LINK_DEFS: 3>,), ())Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):3142 def run(self, text): 3143 # First pass to define all the references 3144 regex_defns = re.compile(r''' 3145 \[\#(\w+) # the counter. Open square plus hash plus a word \1 3146 ([^@]*) # Some optional characters, that aren't an @. \2 3147 @(\w+) # the id. Should this be normed? \3 3148 ([^\]]*)\] # The rest of the text up to the terminating ] \4 3149 ''', re.VERBOSE) 3150 regex_subs = re.compile(r"\[@(\w+)\s*\]") # [@ref_id] 3151 counters = {} 3152 references = {} 3153 replacements = [] 3154 definition_html = '<figcaption class="{}" id="counter-ref-{}">{}{}{}</figcaption>' 3155 reference_html = '<a class="{}" href="#counter-ref-{}">{}</a>' 3156 for match in regex_defns.finditer(text): 3157 # We must have four match groups otherwise this isn't a numbering reference 3158 if len(match.groups()) != 4: 3159 continue 3160 counter = match.group(1) 3161 text_before = match.group(2).strip() 3162 ref_id = match.group(3) 3163 text_after = match.group(4) 3164 number = counters.get(counter, 1) 3165 references[ref_id] = (number, counter) 3166 replacements.append((match.start(0), 3167 definition_html.format(counter, 3168 ref_id, 3169 text_before, 3170 number, 3171 text_after), 3172 match.end(0))) 3173 counters[counter] = number + 1 3174 for repl in reversed(replacements): 3175 text = text[:repl[0]] + repl[1] + text[repl[2]:] 3176 3177 # Second pass to replace the references with the right 3178 # value of the counter 3179 # Fwiw, it's vaguely annoying to have to turn the iterator into 3180 # a list and then reverse it but I can't think of a better thing to do. 3181 for match in reversed(list(regex_subs.finditer(text))): 3182 number, counter = references.get(match.group(1), (None, None)) 3183 if number is not None: 3184 repl = reference_html.format(counter, 3185 match.group(1), 3186 number) 3187 else: 3188 repl = reference_html.format(match.group(1), 3189 'countererror', 3190 '?' + match.group(1) + '?') 3191 if "smarty-pants" in self.md.extras: 3192 repl = repl.replace('"', self.md._escape_table['"']) 3193 3194 text = text[:match.start()] + repl + text[match.end():] 3195 return textRun the extra against the given text.
Returns: The new text after being modified by the extra
Treats unindented Python interactive shell sessions as
blocks.order = ((), (<Stage.LISTS: 6>,))Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def test(self, text):Check a section of markdown to see if this extra should be run upon it. The default implementation will always return True but it's recommended to override this behaviour to improve performance.
def sub(self, match: re.Match) -> str:3210 def sub(self, match: re.Match) -> str: 3211 if "fenced-code-blocks" in self.md.extras: 3212 dedented = _dedent(match.group(0)) 3213 return self.md.extra_classes['fenced-code-blocks'].run("```pycon\n" + dedented + "```\n") 3214 3215 lines = match.group(0).splitlines(0) 3216 _dedentlines(lines) 3217 indent = ' ' * self.md.tab_width 3218 s = ('\n' # separate from possible cuddled paragraph 3219 + indent + ('\n'+indent).join(lines) 3220 + '\n') 3221 return sdef run(self, text):3223 def run(self, text): 3224 less_than_tab = self.md.tab_width - 1 3225 _pyshell_block_re = re.compile(r""" 3226 ^([ ]{0,%d})>>>[ ].*\n # first line 3227 ^(\1[^\S\n]*\S.*\n)* # any number of subsequent lines with at least one character 3228 (?=^\1?\n|\Z) # ends with a blank line or end of document 3229 """ % less_than_tab, re.M | re.X) 3230 3231 return _pyshell_block_re.sub(self.sub, text)Run the extra against the given text.
Returns: The new text after being modified by the extra
Replaces ' and " with curly quotation marks or curly apostrophes. Replaces --, ---, ..., and . . . with en dashes, em dashes, and ellipses.
name = 'smarty-pants'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((), (<Stage.SPAN_GAMUT: 10>,))Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):3264 def run(self, text): 3265 """Fancifies 'single quotes', "double quotes", and apostrophes. 3266 Converts --, ---, and ... into en dashes, em dashes, and ellipses. 3267 3268 Inspiration is: <http://daringfireball.net/projects/smartypants/> 3269 See "test/tm-cases/smarty_pants.text" for a full discussion of the 3270 support here and 3271 <http://code.google.com/p/python-markdown2/issues/detail?id=42> for a 3272 discussion of some diversion from the original SmartyPants. 3273 """ 3274 if "'" in text: # guard for perf 3275 text = self.contractions(text) 3276 text = self._opening_single_quote_re.sub("‘", text) 3277 text = self._closing_single_quote_re.sub("’", text) 3278 3279 if '"' in text: # guard for perf 3280 text = self._opening_double_quote_re.sub("“", text) 3281 text = self._closing_double_quote_re.sub("”", text) 3282 3283 text = text.replace("---", "—") 3284 text = text.replace("--", "–") 3285 text = text.replace("...", "…") 3286 text = text.replace(" . . . ", "…") 3287 text = text.replace(". . .", "…") 3288 3289 # TODO: Temporary hack to fix https://github.com/trentm/python-markdown2/issues/150 3290 if "footnotes" in self.md.extras and "footnote-ref" in text: 3291 # Quotes in the footnote back ref get converted to "smart" quotes 3292 # Change them back here to ensure they work. 3293 text = text.replace('class="footnote-ref”', 'class="footnote-ref"') 3294 3295 return textFancifies 'single quotes', "double quotes", and apostrophes. Converts --, ---, and ... into en dashes, em dashes, and ellipses.
Inspiration is: http://daringfireball.net/projects/smartypants/ See "test/tm-cases/smarty_pants.text" for a full discussion of the support here and http://code.google.com/p/python-markdown2/issues/detail?id=42 for a discussion of some diversion from the original SmartyPants.
Text inside of double tilde is
strikethroughorder = ((<Stage.ITALIC_AND_BOLD: 14>,), ())Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):Run the extra against the given text.
Returns: The new text after being modified by the extra
Tables using the same format as GFM https://help.github.com/articles/github-flavored-markdown#tables and PHP-Markdown Extra https://michelf.ca/projects/php-markdown/extra/#table.
order = ((), (<Stage.LISTS: 6>,))Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):3326 def run(self, text): 3327 """Copying PHP-Markdown and GFM table syntax. Some regex borrowed from 3328 https://github.com/michelf/php-markdown/blob/lib/Michelf/Markdown.php#L2538 3329 """ 3330 less_than_tab = self.md.tab_width - 1 3331 table_re = re.compile(r''' 3332 (?:(?<=\n)|\A\n?) # leading blank line 3333 3334 ^[ ]{0,%d} # allowed whitespace 3335 (.*[|].*)[ ]*\n # $1: header row (at least one pipe) 3336 3337 ^[ ]{0,%d} # allowed whitespace 3338 ( # $2: underline row 3339 # underline row with leading bar 3340 (?: \|\ *:?-+:?\ * )+ \|? \s?[ ]*\n 3341 | 3342 # or, underline row without leading bar 3343 (?: \ *:?-+:?\ *\| )+ (?: \ *:?-+:?\ * )? \s?[ ]*\n 3344 ) 3345 3346 ( # $3: data rows 3347 (?: 3348 ^[ ]{0,%d}(?!\ ) # ensure line begins with 0 to less_than_tab spaces 3349 .*\|.*[ ]*\n 3350 )+ 3351 ) 3352 ''' % (less_than_tab, less_than_tab, less_than_tab), re.M | re.X) 3353 return table_re.sub(self.sub, text)Copying PHP-Markdown and GFM table syntax. Some regex borrowed from https://github.com/michelf/php-markdown/blob/lib/Michelf/Markdown.php#L2538
def sub(self, match: re.Match) -> str:3355 def sub(self, match: re.Match) -> str: 3356 trim_space_re = '^[ \t\n]+|[ \t\n]+$' 3357 trim_bar_re = r'^\||\|$' 3358 split_bar_re = r'^\||(?<![\`\\])\|' 3359 escape_bar_re = r'\\\|' 3360 3361 head, underline, body = match.groups() 3362 3363 # Determine aligns for columns. 3364 cols = [re.sub(escape_bar_re, '|', cell.strip()) for cell in re.split(split_bar_re, re.sub(trim_bar_re, "", re.sub(trim_space_re, "", underline)))] 3365 align_from_col_idx = {} 3366 for col_idx, col in enumerate(cols): 3367 if col[0] == ':' and col[-1] == ':': 3368 align_from_col_idx[col_idx] = ' style="text-align:center;"' 3369 elif col[0] == ':': 3370 align_from_col_idx[col_idx] = ' style="text-align:left;"' 3371 elif col[-1] == ':': 3372 align_from_col_idx[col_idx] = ' style="text-align:right;"' 3373 3374 # thead 3375 hlines = ['<table%s>' % self.md._html_class_str_from_tag('table'), '<thead%s>' % self.md._html_class_str_from_tag('thead'), '<tr>'] 3376 cols = [re.sub(escape_bar_re, '|', cell.strip()) for cell in re.split(split_bar_re, re.sub(trim_bar_re, "", re.sub(trim_space_re, "", head)))] 3377 for col_idx, col in enumerate(cols): 3378 hlines.append(' <th%s>%s</th>' % ( 3379 align_from_col_idx.get(col_idx, ''), 3380 self.md._run_span_gamut(col) 3381 )) 3382 hlines.append('</tr>') 3383 hlines.append('</thead>') 3384 3385 # tbody 3386 hlines.append('<tbody>') 3387 for line in body.strip('\n').split('\n'): 3388 hlines.append('<tr>') 3389 cols = [re.sub(escape_bar_re, '|', cell.strip()) for cell in re.split(split_bar_re, re.sub(trim_bar_re, "", re.sub(trim_space_re, "", line)))] 3390 for col_idx, col in enumerate(cols): 3391 hlines.append(' <td%s>%s</td>' % ( 3392 align_from_col_idx.get(col_idx, ''), 3393 self.md._run_span_gamut(col) 3394 )) 3395 hlines.append('</tr>') 3396 hlines.append('</tbody>') 3397 hlines.append('</table>') 3398 3399 return '\n'.join(hlines) + '\n'Helper class that provides a standard way to create an ABC using inheritance.
name = 'tg-spoiler'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((), (<Stage.ITALIC_AND_BOLD: 14>,))Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):Run the extra against the given text.
Returns: The new text after being modified by the extra
Text inside of double dash is --underlined--.
name = 'underline'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((<Stage.ITALIC_AND_BOLD: 14>,), ())Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):Run the extra against the given text.
Returns: The new text after being modified by the extra
Support for generating Wavedrom digital timing diagrams
order = ((<Stage.CODE_BLOCKS: 7>, <class 'FencedCodeBlocks'>), ())Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def test(self, text):3450 def test(self, text): 3451 match = FencedCodeBlocks.fenced_code_block_re.search(text) 3452 return match is None or match.group(2) == 'wavedrom'Check a section of markdown to see if this extra should be run upon it. The default implementation will always return True but it's recommended to override this behaviour to improve performance.
def sub(self, match: re.Match) -> str:3454 def sub(self, match: re.Match) -> str: 3455 # dedent the block for processing 3456 lead_indent, waves = self.md._uniform_outdent(match.group(3)) 3457 # default tags to wrap the wavedrom block in 3458 open_tag, close_tag = '<script type="WaveDrom">\n', '</script>' 3459 3460 # check if the user would prefer to have the SVG embedded directly 3461 embed_svg = self.options.get('prefer_embed_svg', True) 3462 3463 if embed_svg: 3464 try: 3465 import wavedrom 3466 waves = wavedrom.render(waves).tostring() 3467 open_tag, close_tag = '<div>', '\n</div>' 3468 except ImportError: 3469 pass 3470 3471 # hash SVG to prevent <> chars being messed with 3472 self.md._escape_table[waves] = _hash_text(waves) 3473 3474 return self.md._uniform_indent( 3475 '\n%s%s%s\n' % (open_tag, self.md._escape_table[waves], close_tag), 3476 lead_indent, include_empty_lines=True 3477 )def run(self, text):Run the extra against the given text.
Returns: The new text after being modified by the extra
Inherited Members
Google Code Wiki-style tables. See http://code.google.com/p/support/wiki/WikiSyntax#Tables.
name = 'wiki-tables'An identifiable name that users can use to invoke the extra in the Markdown class
order = ((<class 'Tables'>,), ())Tuple of two iterables containing the stages/extras this extra will run before and after, respectively
def run(self, text):3491 def run(self, text): 3492 less_than_tab = self.md.tab_width - 1 3493 wiki_table_re = re.compile(r''' 3494 (?:(?<=\n\n)|\A\n?) # leading blank line 3495 ^([ ]{0,%d})\|\|.+?\|\|[ ]*\n # first line 3496 (^\1\|\|.+?\|\|\n)* # any number of subsequent lines 3497 ''' % less_than_tab, re.M | re.X) 3498 return wiki_table_re.sub(self.sub, text)Run the extra against the given text.
Returns: The new text after being modified by the extra
def sub(self, match: re.Match) -> str:3500 def sub(self, match: re.Match) -> str: 3501 ttext = match.group(0).strip() 3502 rows = [] 3503 for line in ttext.splitlines(0): 3504 line = line.strip()[2:-2].strip() 3505 row = [c.strip() for c in re.split(r'(?<!\\)\|\|', line)] 3506 rows.append(row) 3507 3508 hlines = [] 3509 3510 def add_hline(line, indents=0): 3511 hlines.append((self.md.tab * indents) + line) 3512 3513 def format_cell(text): 3514 return self.md._run_span_gamut(re.sub(r"^\s*~", "", cell).strip(" ")) 3515 3516 add_hline('<table%s>' % self.md._html_class_str_from_tag('table')) 3517 # Check if first cell of first row is a header cell. If so, assume the whole row is a header row. 3518 if rows and rows[0] and re.match(r"^\s*~", rows[0][0]): 3519 add_hline('<thead%s>' % self.md._html_class_str_from_tag('thead'), 1) 3520 add_hline('<tr>', 2) 3521 for cell in rows[0]: 3522 add_hline("<th>{}</th>".format(format_cell(cell)), 3) 3523 add_hline('</tr>', 2) 3524 add_hline('</thead>', 1) 3525 # Only one header row allowed. 3526 rows = rows[1:] 3527 # If no more rows, don't create a tbody. 3528 if rows: 3529 add_hline('<tbody>', 1) 3530 for row in rows: 3531 add_hline('<tr>', 2) 3532 for cell in row: 3533 add_hline('<td>{}</td>'.format(format_cell(cell)), 3) 3534 add_hline('</tr>', 2) 3535 add_hline('</tbody>', 1) 3536 add_hline('</table>') 3537 return '\n'.join(hlines) + '\n'Return the HTML for the current TOC.
This expects the
_tocattribute to have been set on this instance.(builtins.str): 3605class UnicodeWithAttrs(str): 3606 """A subclass of unicode used for the return value of conversion to 3607 possibly attach some attributes. E.g. the "toc_html" attribute when 3608 the "toc" extra is used. 3609 """ 3610 metadata: Optional[Dict[str, str]] = None 3611 toc_html: Optional[str] = NoneA subclass of unicode used for the return value of conversion to possibly attach some attributes. E.g. the "toc_html" attribute when the "toc" extra is used.
Inherited Members
- builtins.str
- encode
- replace
- split
- rsplit
- join
- capitalize
- casefold
- title
- center
- count
- expandtabs
- find
- partition
- index
- ljust
- lower
- lstrip
- rfind
- rindex
- rjust
- rstrip
- rpartition
- splitlines
- strip
- swapcase
- translate
- upper
- startswith
- endswith
- removeprefix
- removesuffix
- isascii
- islower
- isupper
- istitle
- isspace
- isdecimal
- isdigit
- isnumeric
- isalpha
- isalnum
- isidentifier
- isprintable
- zfill
- format
- format_map
- maketrans