s around
- "paragraphs" that are wrapped in non-block-level tags, such as anchors,
- phrase emphasis, and spans. The list of tags we're looking for is
- hard-coded.
-
- @param raw {boolean} indicates if these are raw HTML blocks in
- the original source. It makes a difference in "safe" mode.
- """
- if '<' not in text:
- return text
-
- # Pass `raw` value into our calls to self._hash_html_block_sub.
- hash_html_block_sub = _curry(self._hash_html_block_sub, raw=raw)
-
- # First, look for nested blocks, e.g.:
- #
", start_idx) + 3
- except ValueError:
- break
-
- # Start position for next comment block search.
- start = end_idx
-
- # Validate whitespace before comment.
- if start_idx:
- # - Up to `tab_width - 1` spaces before start_idx.
- for i in range(self.tab_width - 1):
- if text[start_idx - 1] != ' ':
- break
- start_idx -= 1
- if start_idx == 0:
- break
- # - Must be preceded by 2 newlines or hit the start of
- # the document.
- if start_idx == 0:
- pass
- elif start_idx == 1 and text[0] == '\n':
- start_idx = 0 # to match minute detail of Markdown.pl regex
- elif text[start_idx-2:start_idx] == '\n\n':
- pass
- else:
- break
-
- # Validate whitespace after comment.
- # - Any number of spaces and tabs.
- while end_idx < len(text):
- if text[end_idx] not in ' \t':
- break
- end_idx += 1
- # - Must be following by 2 newlines or hit end of text.
- if text[end_idx:end_idx+2] not in ('', '\n', '\n\n'):
- continue
-
- # Escape and hash (must match `_hash_html_block_sub`).
- html = text[start_idx:end_idx]
- if raw and self.safe_mode:
- html = self._sanitize_html(html)
- key = _hash_text(html)
- self.html_blocks[key] = html
- text = text[:start_idx] + "\n\n" + key + "\n\n" + text[end_idx:]
-
- if "xml" in self.extras:
- # Treat XML processing instructions and namespaced one-liner
- # tags as if they were block HTML tags. E.g., if standalone
- # (i.e. are their own paragraph), the following do not get
- # wrapped in a tag:
- #
- #
- #
- _xml_oneliner_re = _xml_oneliner_re_from_tab_width(self.tab_width)
- text = _xml_oneliner_re.sub(hash_html_block_sub, text)
-
- return text
-
- def _strip_link_definitions(self, text):
- # Strips link definitions from text, stores the URLs and titles in
- # hash references.
- less_than_tab = self.tab_width - 1
-
- # Link defs are in the form:
- # [id]: url "optional title"
- _link_def_re = re.compile(r"""
- ^[ ]{0,%d}\[(.+)\]: # id = \1
- [ \t]*
- \n? # maybe *one* newline
- [ \t]*
- ? # url = \2
- [ \t]*
- (?:
- \n? # maybe one newline
- [ \t]*
- (?<=\s) # lookbehind for whitespace
- ['"(]
- ([^\n]*) # title = \3
- ['")]
- [ \t]*
- )? # title is optional
- (?:\n+|\Z)
- """ % less_than_tab, re.X | re.M | re.U)
- return _link_def_re.sub(self._extract_link_def_sub, text)
-
- def _extract_link_def_sub(self, match):
- id, url, title = match.groups()
- key = id.lower() # Link IDs are case-insensitive
- self.urls[key] = self._encode_amps_and_angles(url)
- if title:
- self.titles[key] = title
- return ""
-
- def _do_numbering(self, text):
- ''' We handle the special extension for generic numbering for
- tables, figures etc.
- '''
- # First pass to define all the references
- self.regex_defns = re.compile(r'''
- \[\#(\w+)\s* # the counter. Open square plus hash plus a word \1
- ([^@]*)\s* # Some optional characters, that aren't an @. \2
- @(\w+) # the id. Should this be normed? \3
- ([^\]]*)\] # The rest of the text up to the terminating ] \4
- ''', re.VERBOSE)
- self.regex_subs = re.compile(r"\[@(\w+)\s*\]") # [@ref_id]
- counters = {}
- references = {}
- replacements = []
- definition_html = '{}{}{}'
- reference_html = '{}'
- for match in self.regex_defns.finditer(text):
- # We must have four match groups otherwise this isn't a numbering reference
- if len(match.groups()) != 4:
- continue
- counter = match.group(1)
- text_before = match.group(2)
- ref_id = match.group(3)
- text_after = match.group(4)
- number = counters.get(counter, 1)
- references[ref_id] = (number, counter)
- replacements.append((match.start(0),
- definition_html.format(counter,
- ref_id,
- text_before,
- number,
- text_after),
- match.end(0)))
- counters[counter] = number + 1
- for repl in reversed(replacements):
- text = text[:repl[0]] + repl[1] + text[repl[2]:]
-
- # Second pass to replace the references with the right
- # value of the counter
- # Fwiw, it's vaguely annoying to have to turn the iterator into
- # a list and then reverse it but I can't think of a better thing to do.
- for match in reversed(list(self.regex_subs.finditer(text))):
- number, counter = references.get(match.group(1), (None, None))
- if number is not None:
- repl = reference_html.format(counter,
- match.group(1),
- number)
- else:
- repl = reference_html.format(match.group(1),
- 'countererror',
- '?' + match.group(1) + '?')
- if "smarty-pants" in self.extras:
- repl = repl.replace('"', self._escape_table['"'])
-
- text = text[:match.start()] + repl + text[match.end():]
- return text
-
- def _extract_footnote_def_sub(self, match):
- id, text = match.groups()
- text = _dedent(text, skip_first_line=not text.startswith('\n')).strip()
- normed_id = re.sub(r'\W', '-', id)
- # Ensure footnote text ends with a couple newlines (for some
- # block gamut matches).
- self.footnotes[normed_id] = text + "\n\n"
- return ""
-
- def _strip_footnote_definitions(self, text):
- """A footnote definition looks like this:
-
- [^note-id]: Text of the note.
-
- May include one or more indented paragraphs.
-
- Where,
- - The 'note-id' can be pretty much anything, though typically it
- is the number of the footnote.
- - The first paragraph may start on the next line, like so:
-
- [^note-id]:
- Text of the note.
- """
- less_than_tab = self.tab_width - 1
- footnote_def_re = re.compile(r'''
- ^[ ]{0,%d}\[\^(.+)\]: # id = \1
- [ \t]*
- ( # footnote text = \2
- # First line need not start with the spaces.
- (?:\s*.*\n+)
- (?:
- (?:[ ]{%d} | \t) # Subsequent lines must be indented.
- .*\n+
- )*
- )
- # Lookahead for non-space at line-start, or end of doc.
- (?:(?=^[ ]{0,%d}\S)|\Z)
- ''' % (less_than_tab, self.tab_width, self.tab_width),
- re.X | re.M)
- return footnote_def_re.sub(self._extract_footnote_def_sub, text)
-
- _hr_re = re.compile(r'^[ ]{0,3}([-_*][ ]{0,2}){3,}$', re.M)
-
- def _run_block_gamut(self, text):
- # These are all the transformations that form block-level
- # tags like paragraphs, headers, and list items.
-
- if "fenced-code-blocks" in self.extras:
- text = self._do_fenced_code_blocks(text)
-
- text = self._do_headers(text)
-
- # Do Horizontal Rules:
- # On the number of spaces in horizontal rules: The spec is fuzzy: "If
- # you wish, you may use spaces between the hyphens or asterisks."
- # Markdown.pl 1.0.1's hr regexes limit the number of spaces between the
- # hr chars to one or two. We'll reproduce that limit here.
- hr = "\n tags around block-level tags.
- text = self._hash_html_blocks(text)
-
- text = self._form_paragraphs(text)
-
- return text
-
- def _pyshell_block_sub(self, match):
- lines = match.group(0).splitlines(0)
- _dedentlines(lines)
- indent = ' ' * self.tab_width
- s = ('\n' # separate from possible cuddled paragraph
- + indent + ('\n'+indent).join(lines)
- + '\n\n')
- return s
-
- def _prepare_pyshell_blocks(self, text):
- """Ensure that Python interactive shell sessions are put in
- code blocks -- even if not properly indented.
- """
- if ">>>" not in text:
- return text
-
- less_than_tab = self.tab_width - 1
- _pyshell_block_re = re.compile(r"""
- ^([ ]{0,%d})>>>[ ].*\n # first line
- ^(\1.*\S+.*\n)* # any number of subsequent lines
- ^\n # ends with a blank line
- """ % less_than_tab, re.M | re.X)
-
- return _pyshell_block_re.sub(self._pyshell_block_sub, text)
-
- def _table_sub(self, match):
- trim_space_re = '^[ \t\n]+|[ \t\n]+$'
- trim_bar_re = '^\||\|$'
-
- head, underline, body = match.groups()
-
- # Determine aligns for columns.
- cols = [cell.strip() for cell in re.sub(trim_bar_re, "", re.sub(trim_space_re, "", underline)).split('|')]
- align_from_col_idx = {}
- for col_idx, col in enumerate(cols):
- if col[0] == ':' and col[-1] == ':':
- align_from_col_idx[col_idx] = ' align="center"'
- elif col[0] == ':':
- align_from_col_idx[col_idx] = ' align="left"'
- elif col[-1] == ':':
- align_from_col_idx[col_idx] = ' align="right"'
-
- # thead
- hlines = ['' % self._html_class_str_from_tag('table'), '', '']
- cols = [cell.strip() for cell in re.sub(trim_bar_re, "", re.sub(trim_space_re, "", head)).split('|')]
- for col_idx, col in enumerate(cols):
- hlines.append(' %s' % (
- align_from_col_idx.get(col_idx, ''),
- self._run_span_gamut(col)
- ))
- hlines.append('
')
- hlines.append('')
-
- # tbody
- hlines.append('')
- for line in body.strip('\n').split('\n'):
- hlines.append('')
- cols = [cell.strip() for cell in re.sub(trim_bar_re, "", re.sub(trim_space_re, "", line)).split('|')]
- for col_idx, col in enumerate(cols):
- hlines.append(' %s' % (
- align_from_col_idx.get(col_idx, ''),
- self._run_span_gamut(col)
- ))
- hlines.append('
')
- hlines.append('')
- hlines.append('