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
29
from unicodedata import east_asian_width as _eawidth
33
__all__ = ["UTextWrapper", "fill", "wrap"]
35
class UTextWrapper(textwrap.TextWrapper):
37
Extend TextWrapper for Unicode.
39
This textwrapper handles east asian double width and split word
40
even if !break_long_words when word contains double width
43
:param ambiguous_width: (keyword argument) width for character when
44
unicodedata.east_asian_width(c) == 'A'
48
* expand_tabs doesn't fixed. It uses len() for calculating width
49
of string on left of TAB.
50
* Handles one codeunit as a single character having 1 or 2 width.
51
This is not correct when there are surrogate pairs, combined
52
characters or zero-width characters.
53
* Treats all asian character are line breakable. But it is not
54
true because line breaking is prohibited around some characters.
55
(For example, breaking before punctation mark is prohibited.)
56
See UAX # 14 "UNICODE LINE BREAKING ALGORITHM"
59
def __init__(self, width=None, **kwargs):
61
width = (osutils.terminal_width() or
62
osutils.default_terminal_width) - 1
64
ambi_width = kwargs.pop('ambiguous_width', 1)
66
self._east_asian_doublewidth = 'FW'
68
self._east_asian_doublewidth = 'FWA'
70
raise ValueError("ambiguous_width should be 1 or 2")
72
self.max_lines = kwargs.get('max_lines', None)
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)
163
if self.max_lines is not None:
164
if self.max_lines > 1:
165
indent = self.subsequent_indent
167
indent = self.initial_indent
168
if self._width(indent) + self._width(self.placeholder.lstrip()) > self.width:
169
raise ValueError("placeholder too large for max width")
171
# Arrange in reverse order so items can be efficiently popped
172
# from a stack of chucks.
177
# Start the list of chunks that will make up the current line.
178
# cur_len is just the length of all the chunks in cur_line.
182
# Figure out which static string will prefix this line.
184
indent = self.subsequent_indent
186
indent = self.initial_indent
188
# Maximum width for this line.
189
width = self.width - len(indent)
191
# First chunk on line is whitespace -- drop it, unless this
192
# is the very beginning of the text (ie. no lines started yet).
193
if self.drop_whitespace and chunks[-1].strip() == '' and lines:
197
# Use _width instead of len for east asian width
198
l = self._width(chunks[-1])
200
# Can at least squeeze this chunk onto the current line.
201
if cur_len + l <= width:
202
cur_line.append(chunks.pop())
205
# Nope, this line is full.
209
# The current line is full, and the next chunk is too big to
210
# fit on *any* line (not just this one).
211
if chunks and self._width(chunks[-1]) > width:
212
self._handle_long_word(chunks, cur_line, cur_len, width)
213
cur_len = sum(map(len, cur_line))
215
# If the last chunk on this line is all whitespace, drop it.
216
if self.drop_whitespace and cur_line and not cur_line[-1].strip():
219
# Convert current line back to a string and store it in list
220
# of all lines (return value).
222
if (self.max_lines is None or
223
len(lines) + 1 < self.max_lines or
225
self.drop_whitespace and
227
not chunks[0].strip()) and cur_len <= width):
228
# Convert current line back to a string and store it in
229
# list of all lines (return value).
230
lines.append(indent + u''.join(cur_line))
233
if (cur_line[-1].strip() and
234
cur_len + self._width(self.placeholder) <= width):
235
cur_line.append(self.placeholder)
236
lines.append(indent + ''.join(cur_line))
238
cur_len -= self._width(cur_line[-1])
242
prev_line = lines[-1].rstrip()
243
if (self._width(prev_line) + self._width(self.placeholder) <=
245
lines[-1] = prev_line + self.placeholder
247
lines.append(indent + self.placeholder.lstrip())
252
def _split(self, text):
253
chunks = textwrap.TextWrapper._split(self, osutils.safe_unicode(text))
254
cjk_split_chunks = []
257
for pos, char in enumerate(chunk):
258
if self._unicode_char_width(char) == 2:
260
cjk_split_chunks.append(chunk[prev_pos:pos])
261
cjk_split_chunks.append(char)
263
if prev_pos < len(chunk):
264
cjk_split_chunks.append(chunk[prev_pos:])
265
return cjk_split_chunks
267
def wrap(self, text):
268
# ensure text is unicode
269
return textwrap.TextWrapper.wrap(self, osutils.safe_unicode(text))
271
# -- 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)
285
def fill(text, width=None, **kwargs):
286
"""Fill a single paragraph of text, returning a new string.
288
Reformat the single paragraph in 'text' to fit in lines of no more
289
than 'width' columns, and return a new string containing the entire
290
wrapped paragraph. As with wrap(), tabs are expanded and other
291
whitespace characters converted to space. See TextWrapper class for
292
available keyword args to customize wrapping behaviour.
294
return UTextWrapper(width=width, **kwargs).fill(text)