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
27
from unicodedata import east_asian_width as _eawidth
31
__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
self.max_lines = kwargs.get('max_lines', None)
72
textwrap.TextWrapper.__init__(self, width, **kwargs)
74
def _unicode_char_width(self, uc):
75
"""Return width of character `uc`.
77
:param: uc Single unicode character.
79
# 'A' means width of the character is not be able to determine.
80
# We assume that it's width is 2 because longer wrap may over
81
# terminal width but shorter wrap may be acceptable.
82
return (_eawidth(uc) in self._east_asian_doublewidth and 2) or 1
85
"""Returns width for s.
87
When s is unicode, take care of east asian width.
88
When s is bytes, treat all byte is single width character.
90
charwidth = self._unicode_char_width
91
return sum(charwidth(c) for c in s)
93
def _cut(self, s, width):
94
"""Returns head and rest of s. (head+rest == s)
96
Head is large as long as _width(head) <= width.
99
charwidth = self._unicode_char_width
100
for pos, c in enumerate(s):
103
return s[:pos], s[pos:]
106
def _fix_sentence_endings(self, chunks):
107
"""_fix_sentence_endings(chunks : [string])
109
Correct for sentence endings buried in 'chunks'. Eg. when the
110
original text contains "... foo.\nBar ...", munge_whitespace()
111
and split() will convert that to [..., "foo.", " ", "Bar", ...]
112
which has one too few spaces; this method simply changes the one
115
Note: This function is copied from textwrap.TextWrap and modified
116
to use unicode always.
120
patsearch = self.sentence_end_re.search
122
if chunks[i + 1] == u" " and patsearch(chunks[i]):
128
def _handle_long_word(self, chunks, cur_line, cur_len, width):
129
# Figure out when indent is larger than the specified width, and make
130
# sure at least one character is stripped off on every pass
132
space_left = chunks[-1] and self._width(chunks[-1][0]) or 1
134
space_left = width - cur_len
136
# If we're allowed to break long words, then do so: put as much
137
# of the next chunk onto the current line as will fit.
138
if self.break_long_words:
139
head, rest = self._cut(chunks[-1], space_left)
140
cur_line.append(head)
146
# Otherwise, we have to preserve the long word intact. Only add
147
# it to the current line if there's nothing already there --
148
# that minimizes how much we violate the width constraint.
150
cur_line.append(chunks.pop())
152
# If we're not allowed to break long words, and there's already
153
# text on the current line, do nothing. Next time through the
154
# main loop of _wrap_chunks(), we'll wind up here again, but
155
# cur_len will be zero, so the next line will be entirely
156
# devoted to the long word that we can't handle right now.
158
def _wrap_chunks(self, chunks):
161
raise ValueError("invalid width %r (must be > 0)" % self.width)
162
if self.max_lines is not None:
163
if self.max_lines > 1:
164
indent = self.subsequent_indent
166
indent = self.initial_indent
167
if self._width(indent) + self._width(self.placeholder.lstrip()) > self.width:
168
raise ValueError("placeholder too large for max width")
170
# Arrange in reverse order so items can be efficiently popped
171
# from a stack of chucks.
176
# Start the list of chunks that will make up the current line.
177
# cur_len is just the length of all the chunks in cur_line.
181
# Figure out which static string will prefix this line.
183
indent = self.subsequent_indent
185
indent = self.initial_indent
187
# Maximum width for this line.
188
width = self.width - len(indent)
190
# First chunk on line is whitespace -- drop it, unless this
191
# is the very beginning of the text (ie. no lines started yet).
192
if self.drop_whitespace and chunks[-1].strip() == '' and lines:
196
# Use _width instead of len for east asian width
197
l = self._width(chunks[-1])
199
# Can at least squeeze this chunk onto the current line.
200
if cur_len + l <= width:
201
cur_line.append(chunks.pop())
204
# Nope, this line is full.
208
# The current line is full, and the next chunk is too big to
209
# fit on *any* line (not just this one).
210
if chunks and self._width(chunks[-1]) > width:
211
self._handle_long_word(chunks, cur_line, cur_len, width)
212
cur_len = sum(map(len, cur_line))
214
# If the last chunk on this line is all whitespace, drop it.
215
if self.drop_whitespace and cur_line and not cur_line[-1].strip():
218
# Convert current line back to a string and store it in list
219
# of all lines (return value).
221
if (self.max_lines is None or
222
len(lines) + 1 < self.max_lines or
224
self.drop_whitespace and
226
not chunks[0].strip()) and cur_len <= width):
227
# Convert current line back to a string and store it in
228
# list of all lines (return value).
229
lines.append(indent + u''.join(cur_line))
232
if (cur_line[-1].strip() and
233
cur_len + self._width(self.placeholder) <= width):
234
cur_line.append(self.placeholder)
235
lines.append(indent + ''.join(cur_line))
237
cur_len -= self._width(cur_line[-1])
241
prev_line = lines[-1].rstrip()
242
if (self._width(prev_line) + self._width(self.placeholder) <=
244
lines[-1] = prev_line + self.placeholder
246
lines.append(indent + self.placeholder.lstrip())
251
def _split(self, text):
252
chunks = textwrap.TextWrapper._split(self, osutils.safe_unicode(text))
253
cjk_split_chunks = []
256
for pos, char in enumerate(chunk):
257
if self._unicode_char_width(char) == 2:
259
cjk_split_chunks.append(chunk[prev_pos:pos])
260
cjk_split_chunks.append(char)
262
if prev_pos < len(chunk):
263
cjk_split_chunks.append(chunk[prev_pos:])
264
return cjk_split_chunks
266
def wrap(self, text):
267
# ensure text is unicode
268
return textwrap.TextWrapper.wrap(self, osutils.safe_unicode(text))
270
# -- Convenience interface ---------------------------------------------
273
def wrap(text, width=None, **kwargs):
274
"""Wrap a single paragraph of text, returning a list of wrapped lines.
276
Reformat the single paragraph in 'text' so it fits in lines of no
277
more than 'width' columns, and return a list of wrapped lines. By
278
default, tabs in 'text' are expanded with string.expandtabs(), and
279
all other whitespace characters (including newline) are converted to
280
space. See TextWrapper class for available keyword args to customize
283
return UTextWrapper(width=width, **kwargs).wrap(text)
286
def fill(text, width=None, **kwargs):
287
"""Fill a single paragraph of text, returning a new string.
289
Reformat the single paragraph in 'text' to fit in lines of no more
290
than 'width' columns, and return a new string containing the entire
291
wrapped paragraph. As with wrap(), tabs are expanded and other
292
whitespace characters converted to space. See TextWrapper class for
293
available keyword args to customize wrapping behaviour.
295
return UTextWrapper(width=width, **kwargs).fill(text)