1
# Copyright (C) 2011 Canonical Ltd
3
# UTextWrapper._handle_long_word, UTextWrapper._wrap_chunks,
4
# UTextWrapper._fix_sentence_endings, wrap and fill is copied from Python's
5
# textwrap module (under PSF license) and modified for support CJK.
6
# Original Copyright for these functions:
8
# Copyright (C) 1999-2001 Gregory P. Ward.
9
# Copyright (C) 2002, 2003 Python Software Foundation.
11
# Written by Greg Ward <gward@python.net>
12
# This program is free software; you can redistribute it and/or modify
13
# it under the terms of the GNU General Public License as published by
14
# the Free Software Foundation; either version 2 of the License, or
15
# (at your option) any later version.
17
# This program is distributed in the hope that it will be useful,
18
# but WITHOUT ANY WARRANTY; without even the implied warranty of
19
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
20
# GNU General Public License for more details.
22
# You should have received a copy of the GNU General Public License
23
# along with this program; if not, write to the Free Software
24
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
26
from __future__ import absolute_import
30
from unicodedata import east_asian_width as _eawidth
34
__all__ = ["UTextWrapper", "fill", "wrap"]
36
class UTextWrapper(textwrap.TextWrapper):
38
Extend TextWrapper for Unicode.
40
This textwrapper handles east asian double width and split word
41
even if !break_long_words when word contains double width
44
:param ambiguous_width: (keyword argument) width for character when
45
unicodedata.east_asian_width(c) == 'A'
49
* expand_tabs doesn't fixed. It uses len() for calculating width
50
of string on left of TAB.
51
* Handles one codeunit as a single character having 1 or 2 width.
52
This is not correct when there are surrogate pairs, combined
53
characters or zero-width characters.
54
* Treats all asian character are line breakable. But it is not
55
true because line breaking is prohibited around some characters.
56
(For example, breaking before punctation mark is prohibited.)
57
See UAX # 14 "UNICODE LINE BREAKING ALGORITHM"
60
def __init__(self, width=None, **kwargs):
62
width = (osutils.terminal_width() or
63
osutils.default_terminal_width) - 1
65
ambi_width = kwargs.pop('ambiguous_width', 1)
67
self._east_asian_doublewidth = 'FW'
69
self._east_asian_doublewidth = 'FWA'
71
raise ValueError("ambiguous_width should be 1 or 2")
73
textwrap.TextWrapper.__init__(self, width, **kwargs)
75
def _unicode_char_width(self, uc):
76
"""Return width of character `uc`.
78
:param: uc Single unicode character.
80
# 'A' means width of the character is not be able to determine.
81
# We assume that it's width is 2 because longer wrap may over
82
# terminal width but shorter wrap may be acceptable.
83
return (_eawidth(uc) in self._east_asian_doublewidth and 2) or 1
86
"""Returns width for s.
88
When s is unicode, take care of east asian width.
89
When s is bytes, treat all byte is single width character.
91
charwidth = self._unicode_char_width
92
return sum(charwidth(c) for c in s)
94
def _cut(self, s, width):
95
"""Returns head and rest of s. (head+rest == s)
97
Head is large as long as _width(head) <= width.
100
charwidth = self._unicode_char_width
101
for pos, c in enumerate(s):
104
return s[:pos], s[pos:]
107
def _fix_sentence_endings(self, chunks):
108
"""_fix_sentence_endings(chunks : [string])
110
Correct for sentence endings buried in 'chunks'. Eg. when the
111
original text contains "... foo.\nBar ...", munge_whitespace()
112
and split() will convert that to [..., "foo.", " ", "Bar", ...]
113
which has one too few spaces; this method simply changes the one
116
Note: This function is copied from textwrap.TextWrap and modified
117
to use unicode always.
121
patsearch = self.sentence_end_re.search
123
if chunks[i+1] == u" " and patsearch(chunks[i]):
129
def _handle_long_word(self, chunks, cur_line, cur_len, width):
130
# Figure out when indent is larger than the specified width, and make
131
# sure at least one character is stripped off on every pass
133
space_left = chunks[-1] and self._width(chunks[-1][0]) or 1
135
space_left = width - cur_len
137
# If we're allowed to break long words, then do so: put as much
138
# of the next chunk onto the current line as will fit.
139
if self.break_long_words:
140
head, rest = self._cut(chunks[-1], space_left)
141
cur_line.append(head)
147
# Otherwise, we have to preserve the long word intact. Only add
148
# it to the current line if there's nothing already there --
149
# that minimizes how much we violate the width constraint.
151
cur_line.append(chunks.pop())
153
# If we're not allowed to break long words, and there's already
154
# text on the current line, do nothing. Next time through the
155
# main loop of _wrap_chunks(), we'll wind up here again, but
156
# cur_len will be zero, so the next line will be entirely
157
# devoted to the long word that we can't handle right now.
159
def _wrap_chunks(self, chunks):
162
raise ValueError("invalid width %r (must be > 0)" % self.width)
164
# Arrange in reverse order so items can be efficiently popped
165
# from a stack of chucks.
170
# Start the list of chunks that will make up the current line.
171
# cur_len is just the length of all the chunks in cur_line.
175
# Figure out which static string will prefix this line.
177
indent = self.subsequent_indent
179
indent = self.initial_indent
181
# Maximum width for this line.
182
width = self.width - len(indent)
184
# First chunk on line is whitespace -- drop it, unless this
185
# is the very beginning of the text (ie. no lines started yet).
186
if self.drop_whitespace and chunks[-1].strip() == '' and lines:
190
# Use _width instead of len for east asian width
191
l = self._width(chunks[-1])
193
# Can at least squeeze this chunk onto the current line.
194
if cur_len + l <= width:
195
cur_line.append(chunks.pop())
198
# Nope, this line is full.
202
# The current line is full, and the next chunk is too big to
203
# fit on *any* line (not just this one).
204
if chunks and self._width(chunks[-1]) > width:
205
self._handle_long_word(chunks, cur_line, cur_len, width)
207
# If the last chunk on this line is all whitespace, drop it.
208
if self.drop_whitespace and cur_line and not cur_line[-1].strip():
211
# Convert current line back to a string and store it in list
212
# of all lines (return value).
214
lines.append(indent + u''.join(cur_line))
218
def _split(self, text):
219
chunks = textwrap.TextWrapper._split(self, unicode(text))
220
cjk_split_chunks = []
223
for pos, char in enumerate(chunk):
224
if self._unicode_char_width(char) == 2:
226
cjk_split_chunks.append(chunk[prev_pos:pos])
227
cjk_split_chunks.append(char)
229
if prev_pos < len(chunk):
230
cjk_split_chunks.append(chunk[prev_pos:])
231
return cjk_split_chunks
233
def wrap(self, text):
234
# ensure text is unicode
235
return textwrap.TextWrapper.wrap(self, unicode(text))
237
# -- Convenience interface ---------------------------------------------
239
def wrap(text, width=None, **kwargs):
240
"""Wrap a single paragraph of text, returning a list of wrapped lines.
242
Reformat the single paragraph in 'text' so it fits in lines of no
243
more than 'width' columns, and return a list of wrapped lines. By
244
default, tabs in 'text' are expanded with string.expandtabs(), and
245
all other whitespace characters (including newline) are converted to
246
space. See TextWrapper class for available keyword args to customize
249
return UTextWrapper(width=width, **kwargs).wrap(text)
251
def fill(text, width=None, **kwargs):
252
"""Fill a single paragraph of text, returning a new string.
254
Reformat the single paragraph in 'text' to fit in lines of no more
255
than 'width' columns, and return a new string containing the entire
256
wrapped paragraph. As with wrap(), tabs are expanded and other
257
whitespace characters converted to space. See TextWrapper class for
258
available keyword args to customize wrapping behaviour.
260
return UTextWrapper(width=width, **kwargs).fill(text)
added
removed
breezy/utextwrap.py
