##// END OF EJS Templates
cleanup: use () to wrap long lines instead of \...
cleanup: use () to wrap long lines instead of \ This is a little less brittle, and often helps indentation. In a surprising number of cases the entire cleanup was deleting the \, as the expression was *already* parenthesized in a workable way. Differential Revision: https://phab.mercurial-scm.org/D5993

File last commit:

r41925:aaad36b8 default
r41925:aaad36b8 default
Show More
minirst.py
827 lines | 28.6 KiB | text/x-python | PythonLexer
Martin Geisler
minimal reStructuredText parser
r9156 # minirst.py - minimal reStructuredText parser
#
Martin Geisler
minirst: support containers...
r10443 # Copyright 2009, 2010 Matt Mackall <mpm@selenic.com> and others
Martin Geisler
minimal reStructuredText parser
r9156 #
# This software may be used and distributed according to the terms of the
Matt Mackall
Update license to GPLv2+
r10263 # GNU General Public License version 2 or any later version.
Martin Geisler
minimal reStructuredText parser
r9156
"""simplified reStructuredText parser.
This parser knows just enough about reStructuredText to parse the
Mercurial docstrings.
It cheats in a major way: nested blocks are not really nested. They
are just indented blocks that look like they are nested. This relies
on the user to keep the right indentation for the blocks.
Matt Mackall
urls: bulk-change primary website URLs
r26421 Remember to update https://mercurial-scm.org/wiki/HelpStyleGuide
Martin Geisler
minirst: link to HelpStyleGuide in docstring
r12958 when adding support for new constructs.
Martin Geisler
minimal reStructuredText parser
r9156 """
Gregory Szorc
minirst: use absolute_import
r25960 from __future__ import absolute_import
Erik Zielke
minirst: Support for admonitions...
r12388
Gregory Szorc
minirst: use absolute_import
r25960 import re
from .i18n import _
from . import (
encoding,
Pulkit Goyal
minirst: make encoding.encoding unicodes to pass into encode() and decode()
r31318 pycompat,
Augie Fackler
python3: use our bytes-only version of cgi.escape everywhere...
r34696 url,
Yuya Nishihara
stringutil: bulk-replace call sites to point to new module...
r37102 )
from .utils import (
stringutil,
Gregory Szorc
minirst: use absolute_import
r25960 )
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750
Dan Villiom Podlaski Christiansen
help: use a full header for topic titles...
r18748 def section(s):
return "%s\n%s\n\n" % (s, "\"" * encoding.colwidth(s))
def subsection(s):
return "%s\n%s\n\n" % (s, '=' * encoding.colwidth(s))
def subsubsection(s):
return "%s\n%s\n\n" % (s, "-" * encoding.colwidth(s))
def subsubsubsection(s):
return "%s\n%s\n\n" % (s, "." * encoding.colwidth(s))
FUJIWARA Katsunori
minirst: use unicode string as intermediate form for replacement...
r11464 def replace(text, substs):
FUJIWARA Katsunori
minirst: use unicode string as intermediate form for replacement...
r15393 '''
Apply a list of (find, replace) pairs to a text.
Yuya Nishihara
doctest: bulk-replace string literals with b'' for Python 3...
r34133 >>> replace(b"foo bar", [(b'f', b'F'), (b'b', b'B')])
FUJIWARA Katsunori
minirst: use unicode string as intermediate form for replacement...
r15393 'Foo Bar'
Yuya Nishihara
doctest: bulk-replace string literals with b'' for Python 3...
r34133 >>> encoding.encoding = b'latin1'
>>> replace(b'\\x81\\\\', [(b'\\\\', b'/')])
FUJIWARA Katsunori
minirst: use unicode string as intermediate form for replacement...
r15393 '\\x81/'
Yuya Nishihara
doctest: bulk-replace string literals with b'' for Python 3...
r34133 >>> encoding.encoding = b'shiftjis'
>>> replace(b'\\x81\\\\', [(b'\\\\', b'/')])
FUJIWARA Katsunori
minirst: use unicode string as intermediate form for replacement...
r15393 '\\x81\\\\'
'''
# some character encodings (cp932 for Japanese, at least) use
# ASCII characters other than control/alphabet/digit as a part of
# multi-bytes characters, so direct replacing with such characters
# on strings in local encoding causes invalid byte sequences.
Pulkit Goyal
minirst: make encoding.encoding unicodes to pass into encode() and decode()
r31318 utext = text.decode(pycompat.sysstr(encoding.encoding))
FUJIWARA Katsunori
minirst: use unicode string as intermediate form for replacement...
r11464 for f, t in substs:
Matt Mackall
minirst: explicitly decode substitutions...
r21745 utext = utext.replace(f.decode("ascii"), t.decode("ascii"))
Pulkit Goyal
minirst: make encoding.encoding unicodes to pass into encode() and decode()
r31318 return utext.encode(pycompat.sysstr(encoding.encoding))
Martin Geisler
minirst: refactor/simplify findblocks
r12651
Pulkit Goyal
minirst: make regular expressions bytes
r31317 _blockre = re.compile(br"\n(?:\s*\n)+")
Martin Geisler
minirst: refactor/simplify findblocks
r12651
Martin Geisler
minimal reStructuredText parser
r9156 def findblocks(text):
"""Find continuous blocks of lines in text.
Returns a list of dictionaries representing the blocks. Each block
has an 'indent' field and a 'lines' field.
"""
Martin Geisler
minirst: refactor/simplify findblocks
r12651 blocks = []
Matt Mackall
minirst: only strip leading newlines, not indentation
r15036 for b in _blockre.split(text.lstrip('\n').rstrip()):
Martin Geisler
minirst: refactor/simplify findblocks
r12651 lines = b.splitlines()
Matt Mackall
minirst: don't choke on empty text
r15123 if lines:
indent = min((len(l) - len(l.lstrip())) for l in lines)
lines = [l[indent:] for l in lines]
Augie Fackler
minirst: move from dict() construction to {} literals...
r20682 blocks.append({'indent': indent, 'lines': lines})
Martin Geisler
minimal reStructuredText parser
r9156 return blocks
def findliteralblocks(blocks):
"""Finds literal blocks and adds a 'type' field to the blocks.
Literal blocks are given the type 'literal', all other blocks are
given type the 'paragraph'.
"""
i = 0
while i < len(blocks):
# Searching for a block that looks like this:
#
# +------------------------------+
# | paragraph |
# | (ends with "::") |
# +------------------------------+
# +---------------------------+
# | indented literal block |
# +---------------------------+
blocks[i]['type'] = 'paragraph'
Matt Mackall
many, many trivial check-code fixups
r10282 if blocks[i]['lines'][-1].endswith('::') and i + 1 < len(blocks):
Martin Geisler
minimal reStructuredText parser
r9156 indent = blocks[i]['indent']
Matt Mackall
many, many trivial check-code fixups
r10282 adjustment = blocks[i + 1]['indent'] - indent
Martin Geisler
minimal reStructuredText parser
r9156
if blocks[i]['lines'] == ['::']:
# Expanded form: remove block
del blocks[i]
i -= 1
elif blocks[i]['lines'][-1].endswith(' ::'):
# Partially minimized form: remove space and both
# colons.
blocks[i]['lines'][-1] = blocks[i]['lines'][-1][:-3]
Augie Fackler
cleanup: use () to wrap long lines instead of \...
r41925 elif (len(blocks[i]['lines']) == 1 and
blocks[i]['lines'][0].lstrip(' ').startswith('.. ') and
blocks[i]['lines'][0].find(' ', 3) == -1):
Mads Kiilerich
comments: fix minor spelling issues found with spell checker
r20549 # directive on its own line, not a literal block
Simon Heimberg
minirst: do not interpret a directive as a literal block...
r19992 i += 1
continue
Martin Geisler
minimal reStructuredText parser
r9156 else:
# Fully minimized form: remove just one colon.
blocks[i]['lines'][-1] = blocks[i]['lines'][-1][:-1]
# List items are formatted with a hanging indent. We must
# correct for this here while we still have the original
# information on the indentation of the subsequent literal
# blocks available.
Martin Geisler
minirst: prepare for general types of bullet lists...
r9738 m = _bulletre.match(blocks[i]['lines'][0])
if m:
indent += m.end()
adjustment -= m.end()
Martin Geisler
minimal reStructuredText parser
r9156
# Mark the following indented blocks.
Matt Mackall
many, many trivial check-code fixups
r10282 while i + 1 < len(blocks) and blocks[i + 1]['indent'] > indent:
blocks[i + 1]['type'] = 'literal'
blocks[i + 1]['indent'] -= adjustment
Martin Geisler
minimal reStructuredText parser
r9156 i += 1
i += 1
return blocks
Pulkit Goyal
minirst: make regular expressions bytes
r31317 _bulletre = re.compile(br'(\*|-|[0-9A-Za-z]+\.|\(?[0-9A-Za-z]+\)|\|) ')
_optionre = re.compile(br'^(-([a-zA-Z0-9]), )?(--[a-z0-9-]+)'
br'((.*) +)(.*)$')
_fieldre = re.compile(br':(?![: ])([^:]*)(?<! ):[ ]+(.*)')
_definitionre = re.compile(br'[^ ]')
_tablere = re.compile(br'(=+\s+)*=+')
Martin Geisler
minirst: combine list parsing in one function...
r9737
def splitparagraphs(blocks):
"""Split paragraphs into lists."""
# Tuples with (list type, item regexp, single line items?). Order
# matters: definition lists has the least specific regexp and must
# come last.
listtypes = [('bullet', _bulletre, True),
('option', _optionre, True),
('field', _fieldre, True),
('definition', _definitionre, False)]
def match(lines, i, itemre, singleline):
"""Does itemre match an item at line i?
Mads Kiilerich
fix trivial spelling errors
r17424 A list item can be followed by an indented line or another list
Martin Geisler
minirst: combine list parsing in one function...
r9737 item (but only if singleline is True).
"""
line1 = lines[i]
Matt Mackall
many, many trivial check-code fixups
r10282 line2 = i + 1 < len(lines) and lines[i + 1] or ''
Martin Geisler
minirst: combine list parsing in one function...
r9737 if not itemre.match(line1):
return False
if singleline:
Augie Fackler
minirst: fix bytes slicing defect on Python 3...
r37900 return line2 == '' or line2[0:1] == ' ' or itemre.match(line2)
Martin Geisler
minirst: combine list parsing in one function...
r9737 else:
return line2.startswith(' ')
i = 0
while i < len(blocks):
if blocks[i]['type'] == 'paragraph':
lines = blocks[i]['lines']
for type, itemre, singleline in listtypes:
if match(lines, 0, itemre, singleline):
items = []
for j, line in enumerate(lines):
if match(lines, j, itemre, singleline):
Augie Fackler
minirst: move from dict() construction to {} literals...
r20682 items.append({'type': type, 'lines': [],
'indent': blocks[i]['indent']})
Martin Geisler
minirst: combine list parsing in one function...
r9737 items[-1]['lines'].append(line)
Matt Mackall
many, many trivial check-code fixups
r10282 blocks[i:i + 1] = items
Martin Geisler
minirst: combine list parsing in one function...
r9737 break
i += 1
return blocks
Olav Reinert
minirst: simplify and standardize field list formatting...
r15861 _fieldwidth = 14
Martin Geisler
minirst: improve layout of field lists...
r10065
def updatefieldlists(blocks):
Olav Reinert
minirst: simplify and standardize field list formatting...
r15861 """Find key for field lists."""
Martin Geisler
minirst: improve layout of field lists...
r10065 i = 0
while i < len(blocks):
if blocks[i]['type'] != 'field':
i += 1
continue
j = i
while j < len(blocks) and blocks[j]['type'] == 'field':
m = _fieldre.match(blocks[j]['lines'][0])
key, rest = m.groups()
blocks[j]['lines'][0] = rest
blocks[j]['key'] = key
j += 1
i = j + 1
return blocks
Erik Zielke
minirst: improved support for option lists....
r13011 def updateoptionlists(blocks):
i = 0
while i < len(blocks):
if blocks[i]['type'] != 'option':
i += 1
continue
optstrwidth = 0
j = i
while j < len(blocks) and blocks[j]['type'] == 'option':
m = _optionre.match(blocks[j]['lines'][0])
shortoption = m.group(2)
group3 = m.group(3)
longoption = group3[2:].strip()
desc = m.group(6).strip()
longoptionarg = m.group(5).strip()
blocks[j]['lines'][0] = desc
noshortop = ''
if not shortoption:
noshortop = ' '
opt = "%s%s" % (shortoption and "-%s " % shortoption or '',
("%s--%s %s") % (noshortop, longoption,
longoptionarg))
opt = opt.rstrip()
blocks[j]['optstr'] = opt
optstrwidth = max(optstrwidth, encoding.colwidth(opt))
j += 1
for block in blocks[i:j]:
block['optstrwidth'] = optstrwidth
i = j + 1
return blocks
Martin Geisler
minirst: support containers...
r10443 def prunecontainers(blocks, keep):
"""Prune unwanted containers.
The blocks must have a 'type' field, i.e., they should have been
run through findliteralblocks first.
"""
Martin Geisler
minirst: report pruned container types
r10444 pruned = []
Martin Geisler
minirst: support containers...
r10443 i = 0
while i + 1 < len(blocks):
# Searching for a block that looks like this:
#
# +-------+---------------------------+
# | ".. container ::" type |
# +---+ |
# | blocks |
# +-------------------------------+
if (blocks[i]['type'] == 'paragraph' and
blocks[i]['lines'][0].startswith('.. container::')):
indent = blocks[i]['indent']
adjustment = blocks[i + 1]['indent'] - indent
containertype = blocks[i]['lines'][0][15:]
Matt Mackall
minirst: allow multiple container types for prune...
r22584 prune = True
for c in keep:
if c in containertype.split('.'):
prune = False
Martin Geisler
minirst: report pruned container types
r10444 if prune:
pruned.append(containertype)
Martin Geisler
minirst: support containers...
r10443
# Always delete "..container:: type" block
del blocks[i]
j = i
Matt Mackall
minirst: fix container stripping logic
r15102 i -= 1
Martin Geisler
minirst: support containers...
r10443 while j < len(blocks) and blocks[j]['indent'] > indent:
if prune:
del blocks[j]
else:
blocks[j]['indent'] -= adjustment
j += 1
i += 1
Martin Geisler
minirst: report pruned container types
r10444 return blocks, pruned
Martin Geisler
minirst: support containers...
r10443
Pulkit Goyal
minirst: make regular expressions bytes
r31317 _sectionre = re.compile(br"""^([-=`:.'"~^_*+#])\1+$""")
Martin Geisler
minirst: support all recommended title adornments
r10984
Matt Mackall
minirst: add simple table support...
r15037 def findtables(blocks):
'''Find simple tables
Only simple one-line table elements are supported
'''
for block in blocks:
# Searching for a block that looks like this:
#
# === ==== ===
# A B C
# === ==== === <- optional
# 1 2 3
# x y z
# === ==== ===
if (block['type'] == 'paragraph' and
Matt Mackall
rst: fix detection of single-row tables...
r15192 len(block['lines']) > 2 and
Matt Mackall
minirst: add simple table support...
r15037 _tablere.match(block['lines'][0]) and
block['lines'][0] == block['lines'][-1]):
block['type'] = 'table'
block['header'] = False
div = block['lines'][0]
Matt Mackall
minirst: fix column handling for simple tables
r15144
# column markers are ASCII so we can calculate column
# position in bytes
Gregory Szorc
global: use pycompat.xrange()...
r38806 columns = [x for x in pycompat.xrange(len(div))
Augie Fackler
minirst: look for column delimiters using slices instead of indicies...
r32526 if div[x:x + 1] == '=' and (x == 0 or
div[x - 1:x] == ' ')]
Matt Mackall
minirst: add simple table support...
r15037 rows = []
for l in block['lines'][1:-1]:
if l == div:
block['header'] = True
continue
row = []
Matt Mackall
minirst: fix column handling for simple tables
r15144 # we measure columns not in bytes or characters but in
# colwidth which makes things tricky
pos = columns[0] # leading whitespace is bytes
Matt Mackall
minirst: add simple table support...
r15037 for n, start in enumerate(columns):
if n + 1 < len(columns):
Matt Mackall
minirst: fix column handling for simple tables
r15144 width = columns[n + 1] - start
v = encoding.getcols(l, pos, width) # gather columns
pos += len(v) # calculate byte position of end
row.append(v.strip())
Matt Mackall
minirst: add simple table support...
r15037 else:
Matt Mackall
minirst: fix column handling for simple tables
r15144 row.append(l[pos:].strip())
Matt Mackall
minirst: add simple table support...
r15037 rows.append(row)
Matt Mackall
minirst: fix column handling for simple tables
r15144
Matt Mackall
minirst: add simple table support...
r15037 block['table'] = rows
return blocks
Martin Geisler
minimal reStructuredText parser
r9156 def findsections(blocks):
"""Finds sections.
The blocks must have a 'type' field, i.e., they should have been
run through findliteralblocks first.
"""
for block in blocks:
# Searching for a block that looks like this:
#
# +------------------------------+
# | Section title |
# | ------------- |
# +------------------------------+
if (block['type'] == 'paragraph' and
len(block['lines']) == 2 and
Matt Mackall
minirst: use colwidth to match title lengths (issue2455)
r12867 encoding.colwidth(block['lines'][0]) == len(block['lines'][1]) and
Martin Geisler
minirst: support all recommended title adornments
r10984 _sectionre.match(block['lines'][1])):
Augie Fackler
minirst: grab a byte, not an int, for the underline style
r32525 block['underline'] = block['lines'][1][0:1]
Martin Geisler
minimal reStructuredText parser
r9156 block['type'] = 'section'
Martin Geisler
minirst: correctly format sections containing inline markup...
r10983 del block['lines'][1]
Martin Geisler
minimal reStructuredText parser
r9156 return blocks
Martin Geisler
minirst: convert ``foo`` into "foo" upon display...
r9623 def inlineliterals(blocks):
FUJIWARA Katsunori
minirst: use unicode string as intermediate form for replacement...
r11464 substs = [('``', '"')]
Martin Geisler
minirst: convert ``foo`` into "foo" upon display...
r9623 for b in blocks:
Martin Geisler
minirst: correctly format sections containing inline markup...
r10983 if b['type'] in ('paragraph', 'section'):
FUJIWARA Katsunori
minirst: use unicode string as intermediate form for replacement...
r11464 b['lines'] = [replace(l, substs) for l in b['lines']]
Martin Geisler
minirst: convert ``foo`` into "foo" upon display...
r9623 return blocks
Martin Geisler
doc, minirst: support hg interpreted text role
r10972 def hgrole(blocks):
timeless
minirst: change hgrole to use single quotes...
r27729 substs = [(':hg:`', "'hg "), ('`', "'")]
Martin Geisler
doc, minirst: support hg interpreted text role
r10972 for b in blocks:
Martin Geisler
minirst: correctly format sections containing inline markup...
r10983 if b['type'] in ('paragraph', 'section'):
Martin Geisler
minirst: handle line breaks in hg role
r11192 # Turn :hg:`command` into "hg command". This also works
# when there is a line break in the command and relies on
# the fact that we have no stray back-quotes in the input
# (run the blocks through inlineliterals first).
FUJIWARA Katsunori
minirst: use unicode string as intermediate form for replacement...
r11464 b['lines'] = [replace(l, substs) for l in b['lines']]
Martin Geisler
doc, minirst: support hg interpreted text role
r10972 return blocks
Martin Geisler
minimal reStructuredText parser
r9156 def addmargins(blocks):
"""Adds empty blocks for vertical spacing.
This groups bullets, options, and definitions together with no vertical
space between them, and adds an empty block between all other blocks.
"""
i = 1
while i < len(blocks):
Matt Mackall
many, many trivial check-code fixups
r10282 if (blocks[i]['type'] == blocks[i - 1]['type'] and
Martin Geisler
minirst: add margin around definition items...
r10936 blocks[i]['type'] in ('bullet', 'option', 'field')):
Martin Geisler
minimal reStructuredText parser
r9156 i += 1
Simon Heimberg
minirst: do not add a 2nd empty paragraph...
r19995 elif not blocks[i - 1]['lines']:
Mads Kiilerich
comments: fix minor spelling issues found with spell checker
r20549 # no lines in previous block, do not separate
Simon Heimberg
minirst: do not add a 2nd empty paragraph...
r19995 i += 1
Martin Geisler
minimal reStructuredText parser
r9156 else:
Augie Fackler
minirst: move from dict() construction to {} literals...
r20682 blocks.insert(i, {'lines': [''], 'indent': 0, 'type': 'margin'})
Martin Geisler
minimal reStructuredText parser
r9156 i += 2
return blocks
Martin Geisler
minirst: ignore comments
r12819 def prunecomments(blocks):
"""Remove comments."""
i = 0
while i < len(blocks):
b = blocks[i]
Erik Zielke
minirst: modified minirst to also recognize empty comments....
r13009 if b['type'] == 'paragraph' and (b['lines'][0].startswith('.. ') or
b['lines'] == ['..']):
Martin Geisler
minirst: ignore comments
r12819 del blocks[i]
Martin Geisler
minirst: better interaction between comments and margins...
r13003 if i < len(blocks) and blocks[i]['type'] == 'margin':
del blocks[i]
Martin Geisler
minirst: ignore comments
r12819 else:
i += 1
return blocks
Gregory Szorc
minirst: dynamically compile admonitions regexp...
r31131
Gregory Szorc
minirst: support passing admonitions into findadmonitions() and parse()...
r31132 def findadmonitions(blocks, admonitions=None):
Erik Zielke
minirst: Support for admonitions...
r12388 """
Makes the type of the block an admonition block if
the first line is an admonition directive
"""
Gregory Szorc
minirst: remove redundant _admonitions set...
r31714 admonitions = admonitions or _admonitiontitles.keys()
Gregory Szorc
minirst: support passing admonitions into findadmonitions() and parse()...
r31132
Pulkit Goyal
minirst: make regular expressions bytes
r31317 admonitionre = re.compile(br'\.\. (%s)::' % '|'.join(sorted(admonitions)),
Gregory Szorc
minirst: dynamically compile admonitions regexp...
r31131 flags=re.IGNORECASE)
Erik Zielke
minirst: Support for admonitions...
r12388 i = 0
while i < len(blocks):
Gregory Szorc
minirst: dynamically compile admonitions regexp...
r31131 m = admonitionre.match(blocks[i]['lines'][0])
Erik Zielke
minirst: Support for admonitions...
r12388 if m:
blocks[i]['type'] = 'admonition'
admonitiontitle = blocks[i]['lines'][0][3:m.end() - 2].lower()
firstline = blocks[i]['lines'][0][m.end() + 1:]
Martin Geisler
minirst: small code cleanup
r12620 if firstline:
blocks[i]['lines'].insert(1, ' ' + firstline)
Erik Zielke
minirst: Support for admonitions...
r12388
blocks[i]['admonitiontitle'] = admonitiontitle
del blocks[i]['lines'][0]
i = i + 1
return blocks
Martin Geisler
minimal reStructuredText parser
r9156
Gregory Szorc
minirst: reindent _admonitiontitles...
r31712 _admonitiontitles = {
'attention': _('Attention:'),
'caution': _('Caution:'),
'danger': _('!Danger!'),
'error': _('Error:'),
'hint': _('Hint:'),
'important': _('Important:'),
'note': _('Note:'),
'tip': _('Tip:'),
'warning': _('Warning!'),
}
Martin Geisler
minirst: pull admonition titles out formatblock function
r12652
Erik Zielke
minirst: improved support for option lists....
r13011 def formatoption(block, width):
Augie Fackler
minirst: use bytes.strip instead of str.strip
r32524 desc = ' '.join(map(bytes.strip, block['lines']))
Erik Zielke
minirst: improved support for option lists....
r13011 colwidth = encoding.colwidth(block['optstr'])
usablewidth = width - 1
hanging = block['optstrwidth']
initindent = '%s%s ' % (block['optstr'], ' ' * ((hanging - colwidth)))
hangindent = ' ' * (encoding.colwidth(initindent) + 1)
Yuya Nishihara
stringutil: bulk-replace call sites to point to new module...
r37102 return ' %s\n' % (stringutil.wrap(desc, usablewidth,
initindent=initindent,
hangindent=hangindent))
Erik Zielke
minirst: improved support for option lists....
r13011
Martin Geisler
minimal reStructuredText parser
r9156 def formatblock(block, width):
"""Format a block according to width."""
Martin Geisler
util, minirst: do not crash with COLUMNS=0
r9417 if width <= 0:
width = 78
Martin Geisler
minimal reStructuredText parser
r9156 indent = ' ' * block['indent']
Erik Zielke
minirst: Support for admonitions...
r12388 if block['type'] == 'admonition':
Martin Geisler
minirst: pull admonition titles out formatblock function
r12652 admonition = _admonitiontitles[block['admonitiontitle']]
Simon Heimberg
minirst: do not fail on an empty admonition block
r19993 if not block['lines']:
return indent + admonition + '\n'
Erik Zielke
minirst: Support for admonitions...
r12388 hang = len(block['lines'][-1]) - len(block['lines'][-1].lstrip())
defindent = indent + hang * ' '
Augie Fackler
minirst: use bytes.strip instead of str.strip
r32524 text = ' '.join(map(bytes.strip, block['lines']))
Matt Mackall
minirst: end all blocks with newlines...
r15125 return '%s\n%s\n' % (indent + admonition,
Yuya Nishihara
stringutil: bulk-replace call sites to point to new module...
r37102 stringutil.wrap(text, width=width,
initindent=defindent,
hangindent=defindent))
Martin Geisler
minimal reStructuredText parser
r9156 if block['type'] == 'margin':
Matt Mackall
minirst: end all blocks with newlines...
r15125 return '\n'
Martin Geisler
minirst: remove unnecessary "elif:" statements
r9735 if block['type'] == 'literal':
Martin Geisler
minirst: indent literal blocks with two spaces...
r9291 indent += ' '
Matt Mackall
minirst: end all blocks with newlines...
r15125 return indent + ('\n' + indent).join(block['lines']) + '\n'
Martin Geisler
minirst: remove unnecessary "elif:" statements
r9735 if block['type'] == 'section':
Matt Mackall
minirst: use colwidth to match title lengths (issue2455)
r12867 underline = encoding.colwidth(block['lines'][0]) * block['underline']
Matt Mackall
minirst: end all blocks with newlines...
r15125 return "%s%s\n%s%s\n" % (indent, block['lines'][0],indent, underline)
Matt Mackall
minirst: add simple table support...
r15037 if block['type'] == 'table':
table = block['table']
# compute column widths
widths = [max([encoding.colwidth(e) for e in c]) for c in zip(*table)]
text = ''
span = sum(widths) + len(widths) - 1
indent = ' ' * block['indent']
hang = ' ' * (len(indent) + span - widths[-1])
for row in table:
Matt Mackall
minirst: fix column handling for simple tables
r15144 l = []
for w, v in zip(widths, row):
pad = ' ' * (w - encoding.colwidth(v))
l.append(v + pad)
l = ' '.join(l)
Yuya Nishihara
stringutil: bulk-replace call sites to point to new module...
r37102 l = stringutil.wrap(l, width=width,
initindent=indent,
hangindent=hang)
Matt Mackall
minirst: add simple table support...
r15037 if not text and block['header']:
text = l + '\n' + indent + '-' * (min(width, span)) + '\n'
else:
text += l + "\n"
return text
Martin Geisler
minirst: remove unnecessary "elif:" statements
r9735 if block['type'] == 'definition':
Martin Geisler
minimal reStructuredText parser
r9156 term = indent + block['lines'][0]
Martin Geisler
minirst: combine list parsing in one function...
r9737 hang = len(block['lines'][-1]) - len(block['lines'][-1].lstrip())
defindent = indent + hang * ' '
Augie Fackler
minirst: use bytes.strip instead of str.strip
r32524 text = ' '.join(map(bytes.strip, block['lines'][1:]))
Yuya Nishihara
stringutil: bulk-replace call sites to point to new module...
r37102 return '%s\n%s\n' % (term, stringutil.wrap(text, width=width,
initindent=defindent,
hangindent=defindent))
Martin Geisler
minirst: removed unnecessary initindent variable
r10937 subindent = indent
Martin Geisler
minirst: remove unnecessary "elif:" statements
r9735 if block['type'] == 'bullet':
Martin Geisler
minirst: support line blocks
r10447 if block['lines'][0].startswith('| '):
# Remove bullet for line blocks and add no extra
Mads Kiilerich
spelling: trivial spell checking
r26781 # indentation.
Martin Geisler
minirst: support line blocks
r10447 block['lines'][0] = block['lines'][0][2:]
else:
m = _bulletre.match(block['lines'][0])
subindent = indent + m.end() * ' '
Martin Geisler
minirst: combine list parsing in one function...
r9737 elif block['type'] == 'field':
Martin Geisler
minirst: improve layout of field lists...
r10065 key = block['key']
subindent = indent + _fieldwidth * ' '
if len(key) + 2 > _fieldwidth:
# key too large, use full line width
key = key.ljust(width)
else:
Olav Reinert
minirst: simplify and standardize field list formatting...
r15861 # key fits within field width
Martin Geisler
minirst: improve layout of field lists...
r10065 key = key.ljust(_fieldwidth)
block['lines'][0] = key + block['lines'][0]
Martin Geisler
minirst: combine list parsing in one function...
r9737 elif block['type'] == 'option':
Erik Zielke
minirst: improved support for option lists....
r13011 return formatoption(block, width)
Martin Geisler
minimal reStructuredText parser
r9156
Pulkit Goyal
minirst: use bytes.strip instead of str.strip...
r31340 text = ' '.join(map(bytes.strip, block['lines']))
Yuya Nishihara
stringutil: bulk-replace call sites to point to new module...
r37102 return stringutil.wrap(text, width=width,
initindent=indent,
hangindent=subindent) + '\n'
Martin Geisler
minimal reStructuredText parser
r9156
Matt Mackall
minirst: add basic HTML formatting support
r15261 def formathtml(blocks):
"""Format RST blocks as HTML"""
out = []
headernest = ''
listnest = []
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750 def escape(s):
Augie Fackler
python3: use our bytes-only version of cgi.escape everywhere...
r34696 return url.escape(s, True)
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750
Matt Mackall
minirst: add basic HTML formatting support
r15261 def openlist(start, level):
if not listnest or listnest[-1][0] != start:
listnest.append((start, level))
out.append('<%s>\n' % start)
blocks = [b for b in blocks if b['type'] != 'margin']
for pos, b in enumerate(blocks):
btype = b['type']
level = b['indent']
lines = b['lines']
if btype == 'admonition':
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750 admonition = escape(_admonitiontitles[b['admonitiontitle']])
Augie Fackler
minirst: use bytes.strip instead of str.strip
r32524 text = escape(' '.join(map(bytes.strip, lines)))
Matt Mackall
minirst: add basic HTML formatting support
r15261 out.append('<p>\n<b>%s</b> %s\n</p>\n' % (admonition, text))
elif btype == 'paragraph':
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750 out.append('<p>\n%s\n</p>\n' % escape('\n'.join(lines)))
Matt Mackall
minirst: add basic HTML formatting support
r15261 elif btype == 'margin':
pass
elif btype == 'literal':
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750 out.append('<pre>\n%s\n</pre>\n' % escape('\n'.join(lines)))
Matt Mackall
minirst: add basic HTML formatting support
r15261 elif btype == 'section':
i = b['underline']
if i not in headernest:
headernest += i
level = headernest.index(i) + 1
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750 out.append('<h%d>%s</h%d>\n' % (level, escape(lines[0]), level))
Matt Mackall
minirst: add basic HTML formatting support
r15261 elif btype == 'table':
table = b['table']
Dan Villiom Podlaski Christiansen
minirst: optimize HTML table generation a bit...
r18752 out.append('<table>\n')
Matt Mackall
minirst: add basic HTML formatting support
r15261 for row in table:
Dan Villiom Podlaski Christiansen
minirst: optimize HTML table generation a bit...
r18752 out.append('<tr>')
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750 for v in row:
Dan Villiom Podlaski Christiansen
minirst: optimize HTML table generation a bit...
r18752 out.append('<td>')
out.append(escape(v))
out.append('</td>')
out.append('\n')
out.pop()
out.append('</tr>\n')
out.append('</table>\n')
Matt Mackall
minirst: add basic HTML formatting support
r15261 elif btype == 'definition':
openlist('dl', level)
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750 term = escape(lines[0])
Augie Fackler
minirst: use bytes.strip instead of str.strip
r32524 text = escape(' '.join(map(bytes.strip, lines[1:])))
Matt Mackall
minirst: add basic HTML formatting support
r15261 out.append(' <dt>%s\n <dd>%s\n' % (term, text))
elif btype == 'bullet':
bullet, head = lines[0].split(' ', 1)
Gregory Szorc
minirst: detect bullet lists using asterisks...
r31130 if bullet in ('*', '-'):
Matt Mackall
minirst: add basic HTML formatting support
r15261 openlist('ul', level)
else:
openlist('ol', level)
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750 out.append(' <li> %s\n' % escape(' '.join([head] + lines[1:])))
Matt Mackall
minirst: add basic HTML formatting support
r15261 elif btype == 'field':
openlist('dl', level)
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750 key = escape(b['key'])
Augie Fackler
minirst: use bytes.strip instead of str.strip
r32524 text = escape(' '.join(map(bytes.strip, lines)))
Matt Mackall
minirst: add basic HTML formatting support
r15261 out.append(' <dt>%s\n <dd>%s\n' % (key, text))
elif btype == 'option':
openlist('dl', level)
Dan Villiom Podlaski Christiansen
minirst: CGI escape strings prior to embedding it in the HTML
r18750 opt = escape(b['optstr'])
Augie Fackler
minirst: use bytes.strip instead of str.strip
r32524 desc = escape(' '.join(map(bytes.strip, lines)))
Matt Mackall
minirst: add basic HTML formatting support
r15261 out.append(' <dt>%s\n <dd>%s\n' % (opt, desc))
# close lists if indent level of next block is lower
if listnest:
start, level = listnest[-1]
if pos == len(blocks) - 1:
out.append('</%s>\n' % start)
listnest.pop()
else:
nb = blocks[pos + 1]
ni = nb['indent']
if (ni < level or
(ni == level and
nb['type'] not in 'definition bullet field option')):
out.append('</%s>\n' % start)
listnest.pop()
return ''.join(out)
Gregory Szorc
minirst: support passing admonitions into findadmonitions() and parse()...
r31132 def parse(text, indent=0, keep=None, admonitions=None):
Matt Mackall
minirst: add parse method to get document structure
r15012 """Parse text into a list of blocks"""
Martin Geisler
minimal reStructuredText parser
r9156 blocks = findblocks(text)
Martin Geisler
help: un-indent help topics...
r9540 for b in blocks:
b['indent'] += indent
Martin Geisler
minimal reStructuredText parser
r9156 blocks = findliteralblocks(blocks)
Matt Mackall
minirst: add simple table support...
r15037 blocks = findtables(blocks)
Martin Geisler
minirst: report pruned container types
r10444 blocks, pruned = prunecontainers(blocks, keep or [])
Martin Geisler
minirst: correctly format sections containing inline markup...
r10983 blocks = findsections(blocks)
Martin Geisler
minirst: convert ``foo`` into "foo" upon display...
r9623 blocks = inlineliterals(blocks)
Martin Geisler
doc, minirst: support hg interpreted text role
r10972 blocks = hgrole(blocks)
Martin Geisler
minirst: combine list parsing in one function...
r9737 blocks = splitparagraphs(blocks)
Martin Geisler
minirst: improve layout of field lists...
r10065 blocks = updatefieldlists(blocks)
Erik Zielke
minirst: improved support for option lists....
r13011 blocks = updateoptionlists(blocks)
Gregory Szorc
minirst: support passing admonitions into findadmonitions() and parse()...
r31132 blocks = findadmonitions(blocks, admonitions=admonitions)
Martin Geisler
minirst: better interaction between comments and margins...
r13003 blocks = addmargins(blocks)
Martin Geisler
minirst: ignore comments
r12819 blocks = prunecomments(blocks)
Matt Mackall
minirst: add parse method to get document structure
r15012 return blocks, pruned
Matt Mackall
minirst: add formatblocks helper
r15013 def formatblocks(blocks, width):
Matt Mackall
minirst: end all blocks with newlines...
r15125 text = ''.join(formatblock(b, width) for b in blocks)
Matt Mackall
minirst: add formatblocks helper
r15013 return text
Yuya Nishihara
minirst: extract function that formats parsed blocks as plain text
r39343 def formatplain(blocks, width):
"""Format parsed blocks as plain text"""
return ''.join(formatblock(b, width) for b in blocks)
Matt Mackall
help: basic support for showing only specified topic sections...
r22587 def format(text, width=80, indent=0, keep=None, style='plain', section=None):
Matt Mackall
minirst: add parse method to get document structure
r15012 """Parse and format the text according to width."""
blocks, pruned = parse(text, indent, keep or [])
Yuya Nishihara
minirst: extract function that filters parsed blocks by section name...
r39341 if section:
blocks = filtersections(blocks, section)
if style == 'html':
Yuya Nishihara
minirst: make format() simply return a formatted text...
r39346 return formathtml(blocks)
Yuya Nishihara
minirst: extract function that filters parsed blocks by section name...
r39341 else:
Yuya Nishihara
minirst: make format() simply return a formatted text...
r39346 return formatplain(blocks, width=width)
Yuya Nishihara
minirst: extract function that filters parsed blocks by section name...
r39341
def filtersections(blocks, section):
Yuya Nishihara
minirst: filter blocks by full path to section
r39377 """Select parsed blocks under the specified section
The section name is separated by a dot, and matches the suffix of the
full section path.
"""
timeless@mozdev.org
help: distinguish sections when multiple match (issue4802)
r26113 parents = []
Yuya Nishihara
minirst: mark getsections() as an internal helper
r39376 sections = _getsections(blocks)
Yuya Nishihara
minirst: unindent "if True" block in filtersections()
r39342 blocks = []
i = 0
lastparents = []
synthetic = []
collapse = True
while i < len(sections):
Yuya Nishihara
minirst: filter blocks by full path to section
r39377 path, nest, b = sections[i]
Yuya Nishihara
minirst: unindent "if True" block in filtersections()
r39342 del parents[nest:]
parents.append(i)
Yuya Nishihara
minirst: filter blocks by full path to section
r39377 if path == section or path.endswith('.' + section):
Yuya Nishihara
minirst: unindent "if True" block in filtersections()
r39342 if lastparents != parents:
llen = len(lastparents)
plen = len(parents)
if llen and llen != plen:
collapse = False
s = []
for j in pycompat.xrange(3, plen - 1):
parent = parents[j]
if (j >= llen or
lastparents[j] != parent):
s.append(len(blocks))
sec = sections[parent][2]
blocks.append(sec[0])
blocks.append(sec[-1])
if s:
synthetic.append(s)
timeless
help: include section heading if section depth changes...
r27614
Yuya Nishihara
minirst: unindent "if True" block in filtersections()
r39342 lastparents = parents[:]
blocks.extend(b)
Jordi Gutiérrez Hermoso
help: show all nested subsections of a section with `hg help foo.section`...
r22770
Yuya Nishihara
minirst: unindent "if True" block in filtersections()
r39342 ## Also show all subnested sections
while i + 1 < len(sections) and sections[i + 1][1] > nest:
i += 1
blocks.extend(sections[i][2])
i += 1
if collapse:
synthetic.reverse()
for s in synthetic:
path = [blocks[syn]['lines'][0] for syn in s]
real = s[-1] + 2
realline = blocks[real]['lines']
realline[0] = ('"%s"' %
'.'.join(path + [realline[0]]).replace('"', ''))
del blocks[s[0]:real]
Jordi Gutiérrez Hermoso
help: show all nested subsections of a section with `hg help foo.section`...
r22770
Yuya Nishihara
minirst: extract function that filters parsed blocks by section name...
r39341 return blocks
Martin Geisler
minimal reStructuredText parser
r9156
Yuya Nishihara
minirst: mark getsections() as an internal helper
r39376 def _getsections(blocks):
Yuya Nishihara
minirst: filter blocks by full path to section
r39377 '''return a list of (section path, nesting level, blocks) tuples'''
Matt Mackall
minirst: add getsections helper
r15014 nest = ""
Yuya Nishihara
minirst: filter blocks by full path to section
r39377 names = ()
Matt Mackall
minirst: add getsections helper
r15014 secs = []
Matt Mackall
help: basic support for showing only specified topic sections...
r22587
def getname(b):
Matt Harbison
help: support 'hg help template.somekeyword'...
r25723 if b['type'] == 'field':
x = b['key']
else:
x = b['lines'][0]
FUJIWARA Katsunori
help: search section of help topic by translated section name correctly...
r29155 x = encoding.lower(x).strip('"')
Matt Mackall
help: basic support for showing only specified topic sections...
r22587 if '(' in x:
x = x.split('(')[0]
return x
Matt Mackall
minirst: add getsections helper
r15014 for b in blocks:
if b['type'] == 'section':
i = b['underline']
if i not in nest:
nest += i
level = nest.index(i) + 1
nest = nest[:level]
Yuya Nishihara
minirst: filter blocks by full path to section
r39377 names = names[:level] + (getname(b),)
secs.append(('.'.join(names), level, [b]))
Matt Harbison
help: support 'hg help template.somekeyword'...
r25723 elif b['type'] in ('definition', 'field'):
Matt Mackall
help: basic support for showing only specified topic sections...
r22587 i = ' '
if i not in nest:
nest += i
level = nest.index(i) + 1
nest = nest[:level]
timeless@mozdev.org
minirst: establish leveling for nested definitions
r26237 for i in range(1, len(secs) + 1):
sec = secs[-i]
if sec[1] < level:
break
siblings = [a for a in sec[2] if a['type'] == 'definition']
if siblings:
siblingindent = siblings[-1]['indent']
indent = b['indent']
if siblingindent < indent:
level += 1
break
elif siblingindent == indent:
level = sec[1]
break
Yuya Nishihara
minirst: filter blocks by full path to section
r39377 names = names[:level] + (getname(b),)
secs.append(('.'.join(names), level, [b]))
Matt Mackall
minirst: add getsections helper
r15014 else:
if not secs:
# add an initial empty section
secs = [('', 0, [])]
timeless@mozdev.org
minirst: don't treat top level item as children of last item (issue4803)...
r26157 if b['type'] != 'margin':
pointer = 1
bindent = b['indent']
while pointer < len(secs):
section = secs[-pointer][2][0]
if section['type'] != 'margin':
sindent = section['indent']
if len(section['lines']) > 1:
Augie Fackler
cleanup: use () to wrap long lines instead of \...
r41925 sindent += (len(section['lines'][1]) -
len(section['lines'][1].lstrip(' ')))
timeless@mozdev.org
minirst: don't treat top level item as children of last item (issue4803)...
r26157 if bindent >= sindent:
break
pointer += 1
if pointer > 1:
timeless@mozdev.org
minirst: handle edge in hunting for parents...
r26170 blevel = secs[-pointer][1]
if section['type'] != b['type']:
blevel += 1
secs.append(('', blevel, []))
Matt Mackall
minirst: add getsections helper
r15014 secs[-1][2].append(b)
return secs
Martin Geisler
minimal reStructuredText parser
r9156
Matt Mackall
minirst: add a helper function to build an RST table from an array
r15039 def maketable(data, indent=0, header=False):
Olav Reinert
minirst: generate tables as a list of joined lines
r16815 '''Generate an RST table for the given table data as a list of lines'''
Matt Mackall
minirst: add a helper function to build an RST table from an array
r15039
widths = [max(encoding.colwidth(e) for e in c) for c in zip(*data)]
indent = ' ' * indent
div = indent + ' '.join('=' * w for w in widths) + '\n'
out = [div]
for row in data:
Matt Mackall
minirst: fix column handling for simple tables
r15144 l = []
for w, v in zip(widths, row):
Simon Heimberg
minirst: create valid output when table data contains a newline...
r20654 if '\n' in v:
# only remove line breaks and indentation, long lines are
# handled by the next tool
v = ' '.join(e.lstrip() for e in v.split('\n'))
Matt Mackall
minirst: fix column handling for simple tables
r15144 pad = ' ' * (w - encoding.colwidth(v))
l.append(v + pad)
out.append(indent + ' '.join(l) + "\n")
Matt Mackall
minirst: add a helper function to build an RST table from an array
r15039 if header and len(data) > 1:
out.insert(2, div)
out.append(div)
Olav Reinert
minirst: generate tables as a list of joined lines
r16815 return out