1
# Copyright (C) 2011 Canonical Ltd
3
# UTextWrapper._handle_long_word, UTextWrapper._wrap_chunks,
4
# wrap and fill is copied from Python's textwrap module
5
# (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
28
from unicodedata import east_asian_width as _eawidth
30
from bzrlib import osutils
32
__all__ = ["UTextWrapper", "fill", "wrap"]
34
class UTextWrapper(textwrap.TextWrapper):
36
Extend TextWrapper for Unicode.
38
This textwrapper handles east asian double width and split word
39
even if !break_long_words when word contains double width
42
:param ambiguous_width: (keyword argument) width for character when
43
unicodedata.east_asian_width(c) == 'A'
47
* expand_tabs doesn't fixed. It uses len() for calculating width
48
of string on left of TAB.
49
* Handles one codeunit as a single character having 1 or 2 width.
50
This is not correct when there are surrogate pairs, combined
51
characters or zero-width characters.
52
* Treats all asian character are line breakable. But it is not
53
true because line breaking is prohibited around some characters.
54
(For example, breaking before punctation mark is prohibited.)
55
See UAX # 14 "UNICODE LINE BREAKING ALGORITHM"
58
def __init__(self, width=None, **kwargs):
60
width = (osutils.terminal_width() or
61
osutils.default_terminal_width) - 1
63
ambi_width = kwargs.pop('ambiguous_width', 1)
65
self._east_asian_doublewidth = 'FW'
67
self._east_asian_doublewidth = 'FWA'
69
raise ValueError("ambiguous_width should be 1 or 2")
71
# No drop_whitespace param before Python 2.6 it was always dropped
72
if sys.version_info < (2, 6):
73
self.drop_whitespace = kwargs.pop("drop_whitespace", True)
74
if not self.drop_whitespace:
75
raise ValueError("TextWrapper version must drop whitespace")
76
textwrap.TextWrapper.__init__(self, width, **kwargs)
78
def _unicode_char_width(self, uc):
79
"""Return width of character `uc`.
81
:param: uc Single unicode character.
83
# 'A' means width of the character is not be able to determine.
84
# We assume that it's width is 2 because longer wrap may over
85
# terminal width but shorter wrap may be acceptable.
86
return (_eawidth(uc) in self._east_asian_doublewidth and 2) or 1
89
"""Returns width for s.
91
When s is unicode, take care of east asian width.
92
When s is bytes, treat all byte is single width character.
94
assert isinstance(s, unicode)
95
charwidth = self._unicode_char_width
96
return sum(charwidth(c) for c in s)
98
def _cut(self, s, width):
99
"""Returns head and rest of s. (head+rest == s)
101
Head is large as long as _width(head) <= width.
103
assert isinstance(s, unicode)
105
charwidth = self._unicode_char_width
106
for pos, c in enumerate(s):
109
return s[:pos], s[pos:]
112
def _handle_long_word(self, chunks, cur_line, cur_len, width):
113
# Figure out when indent is larger than the specified width, and make
114
# sure at least one character is stripped off on every pass
116
space_left = chunks[-1] and self._width(chunks[-1][0]) or 1
118
space_left = width - cur_len
120
# If we're allowed to break long words, then do so: put as much
121
# of the next chunk onto the current line as will fit.
122
if self.break_long_words:
123
head, rest = self._cut(chunks[-1], space_left)
124
cur_line.append(head)
130
# Otherwise, we have to preserve the long word intact. Only add
131
# it to the current line if there's nothing already there --
132
# that minimizes how much we violate the width constraint.
134
cur_line.append(chunks.pop())
136
# If we're not allowed to break long words, and there's already
137
# text on the current line, do nothing. Next time through the
138
# main loop of _wrap_chunks(), we'll wind up here again, but
139
# cur_len will be zero, so the next line will be entirely
140
# devoted to the long word that we can't handle right now.
142
def _wrap_chunks(self, chunks):
145
raise ValueError("invalid width %r (must be > 0)" % self.width)
147
# Arrange in reverse order so items can be efficiently popped
148
# from a stack of chucks.
153
# Start the list of chunks that will make up the current line.
154
# cur_len is just the length of all the chunks in cur_line.
158
# Figure out which static string will prefix this line.
160
indent = self.subsequent_indent
162
indent = self.initial_indent
164
# Maximum width for this line.
165
width = self.width - len(indent)
167
# First chunk on line is whitespace -- drop it, unless this
168
# is the very beginning of the text (ie. no lines started yet).
169
if self.drop_whitespace and chunks[-1].strip() == '' and lines:
173
# Use _width instead of len for east asian width
174
l = self._width(chunks[-1])
176
# Can at least squeeze this chunk onto the current line.
177
if cur_len + l <= width:
178
cur_line.append(chunks.pop())
181
# Nope, this line is full.
185
# The current line is full, and the next chunk is too big to
186
# fit on *any* line (not just this one).
187
if chunks and self._width(chunks[-1]) > width:
188
self._handle_long_word(chunks, cur_line, cur_len, width)
190
# If the last chunk on this line is all whitespace, drop it.
191
if self.drop_whitespace and cur_line and cur_line[-1].strip() == '':
194
# Convert current line back to a string and store it in list
195
# of all lines (return value).
197
lines.append(indent + ''.join(cur_line))
201
def _split(self, text):
202
chunks = textwrap.TextWrapper._split(self, unicode(text))
203
cjk_split_chunks = []
205
assert chunk # TextWrapper._split removes empty chunk
207
for pos, char in enumerate(chunk):
208
if _eawidth(char) in 'FWA':
210
cjk_split_chunks.append(chunk[prev_pos:pos])
211
cjk_split_chunks.append(char)
213
if prev_pos < len(chunk):
214
cjk_split_chunks.append(chunk[prev_pos:])
215
return cjk_split_chunks
217
def wrap(self, text):
218
# ensure text is unicode
219
return textwrap.TextWrapper.wrap(self, unicode(text))
221
# -- Convenience interface ---------------------------------------------
223
def wrap(text, width=None, **kwargs):
224
"""Wrap a single paragraph of text, returning a list of wrapped lines.
226
Reformat the single paragraph in 'text' so it fits in lines of no
227
more than 'width' columns, and return a list of wrapped lines. By
228
default, tabs in 'text' are expanded with string.expandtabs(), and
229
all other whitespace characters (including newline) are converted to
230
space. See TextWrapper class for available keyword args to customize
233
return UTextWrapper(width=width, **kwargs).wrap(text)
235
def fill(text, width=None, **kwargs):
236
"""Fill a single paragraph of text, returning a new string.
238
Reformat the single paragraph in 'text' to fit in lines of no more
239
than 'width' columns, and return a new string containing the entire
240
wrapped paragraph. As with wrap(), tabs are expanded and other
241
whitespace characters converted to space. See TextWrapper class for
242
available keyword args to customize wrapping behaviour.
244
return UTextWrapper(width=width, **kwargs).fill(text)
added
removed
bzrlib/utextwrap.py
