Mercurial > babel > mirror
annotate babel/messages/pofile.py @ 582:3fd7fb953633 trunk
fix handling of messages containing '\\n' (#171)
author | fschwarz |
---|---|
date | Fri, 03 Aug 2012 22:41:49 +0000 |
parents | 99706377c930 |
children | 5c9dba5dd311 |
rev | line source |
---|---|
1 | 1 # -*- coding: utf-8 -*- |
2 # | |
530 | 3 # Copyright (C) 2007-2011 Edgewall Software |
1 | 4 # All rights reserved. |
5 # | |
6 # This software is licensed as described in the file COPYING, which | |
7 # you should have received as part of this distribution. The terms | |
8 # are also available at http://babel.edgewall.org/wiki/License. | |
9 # | |
10 # This software consists of voluntary contributions made by many | |
11 # individuals. For the exact contribution history, see the revision | |
12 # history and logs, available at http://babel.edgewall.org/log/. | |
13 | |
14 """Reading and writing of files in the ``gettext`` PO (portable object) | |
15 format. | |
16 | |
17 :see: `The Format of PO Files | |
18 <http://www.gnu.org/software/gettext/manual/gettext.html#PO-Files>`_ | |
19 """ | |
20 | |
531 | 21 from datetime import datetime |
134 | 22 import os |
1 | 23 import re |
24 | |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
25 from babel.messages.catalog import Catalog, Message |
531 | 26 from babel.util import wraptext |
1 | 27 |
178
749c0f6863bc
Minor change to what symbols are ?exported?, primarily for the generated docs.
cmlenz
parents:
175
diff
changeset
|
28 __all__ = ['read_po', 'write_po'] |
161 | 29 __docformat__ = 'restructuredtext en' |
158 | 30 |
31 def unescape(string): | |
32 r"""Reverse `escape` the given string. | |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
33 |
158 | 34 >>> print unescape('"Say:\\n \\"hello, world!\\"\\n"') |
35 Say: | |
36 "hello, world!" | |
37 <BLANKLINE> | |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
38 |
158 | 39 :param string: the string to unescape |
40 :return: the unescaped string | |
41 :rtype: `str` or `unicode` | |
42 """ | |
582 | 43 def replace_escapes(match): |
44 m = match.group(1) | |
45 if m == 'n': | |
46 return '\n' | |
47 elif m == 't': | |
48 return '\t' | |
49 elif m == 'r': | |
50 return '\r' | |
51 # m is \ or " | |
52 return m | |
53 return re.compile(r'\\([\\trn"])').sub(replace_escapes, string[1:-1]) | |
158 | 54 |
55 def denormalize(string): | |
56 r"""Reverse the normalization done by the `normalize` function. | |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
57 |
158 | 58 >>> print denormalize(r'''"" |
59 ... "Say:\n" | |
60 ... " \"hello, world!\"\n"''') | |
61 Say: | |
62 "hello, world!" | |
63 <BLANKLINE> | |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
64 |
158 | 65 >>> print denormalize(r'''"" |
66 ... "Say:\n" | |
67 ... " \"Lorem ipsum dolor sit " | |
68 ... "amet, consectetur adipisicing" | |
69 ... " elit, \"\n"''') | |
70 Say: | |
71 "Lorem ipsum dolor sit amet, consectetur adipisicing elit, " | |
72 <BLANKLINE> | |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
73 |
158 | 74 :param string: the string to denormalize |
75 :return: the denormalized string | |
76 :rtype: `unicode` or `str` | |
77 """ | |
78 if string.startswith('""'): | |
79 lines = [] | |
80 for line in string.splitlines()[1:]: | |
81 lines.append(unescape(line)) | |
82 return ''.join(lines) | |
83 else: | |
84 return unescape(string) | |
1 | 85 |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
86 def read_po(fileobj, locale=None, domain=None, ignore_obsolete=False): |
6
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
87 """Read messages from a ``gettext`` PO (portable object) file from the given |
64 | 88 file-like object and return a `Catalog`. |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
89 |
6
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
90 >>> from StringIO import StringIO |
106
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
91 >>> buf = StringIO(''' |
6
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
92 ... #: main.py:1 |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
93 ... #, fuzzy, python-format |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
94 ... msgid "foo %(name)s" |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
95 ... msgstr "" |
21
cd9aa202568e
Change pot header's first line, "Translations Template for %%(project)s." instead of "SOME DESCRIPTIVE TITLE.". '''`project`''' and '''`version`''' now default to '''PROJECT''' and '''VERSION''' respectively. Fixed a bug regarding '''Content-Transfer-Encoding''', it shouldn't be the charset, and we're defaulting to `8bit` untill someone complains.
palgarvio
parents:
17
diff
changeset
|
96 ... |
94
96037779b518
Updated `read_po` to add user comments besides just auto comments.
palgarvio
parents:
84
diff
changeset
|
97 ... # A user comment |
96037779b518
Updated `read_po` to add user comments besides just auto comments.
palgarvio
parents:
84
diff
changeset
|
98 ... #. An auto comment |
6
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
99 ... #: main.py:3 |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
100 ... msgid "bar" |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
101 ... msgid_plural "baz" |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
102 ... msgstr[0] "" |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
103 ... msgstr[1] "" |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
104 ... ''') |
64 | 105 >>> catalog = read_po(buf) |
104
395704fda00b
Merged `write_pot` and `write_po` functions by moving more functionality to the `Catalog` class. This is certainly not perfect yet, but moves us in the right direction.
cmlenz
parents:
103
diff
changeset
|
106 >>> catalog.revision_date = datetime(2007, 04, 01) |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
107 |
64 | 108 >>> for message in catalog: |
67 | 109 ... if message.id: |
110 ... print (message.id, message.string) | |
105
c62b68a0b65e
`Message`, `read_po` and `write_po` now all handle user/auto comments correctly.
palgarvio
parents:
104
diff
changeset
|
111 ... print ' ', (message.locations, message.flags) |
c62b68a0b65e
`Message`, `read_po` and `write_po` now all handle user/auto comments correctly.
palgarvio
parents:
104
diff
changeset
|
112 ... print ' ', (message.user_comments, message.auto_comments) |
149
d62c63280e81
Respect charset specified in PO headers in `read_po()`. Fixes #17.
cmlenz
parents:
134
diff
changeset
|
113 (u'foo %(name)s', '') |
d62c63280e81
Respect charset specified in PO headers in `read_po()`. Fixes #17.
cmlenz
parents:
134
diff
changeset
|
114 ([(u'main.py', 1)], set([u'fuzzy', u'python-format'])) |
105
c62b68a0b65e
`Message`, `read_po` and `write_po` now all handle user/auto comments correctly.
palgarvio
parents:
104
diff
changeset
|
115 ([], []) |
149
d62c63280e81
Respect charset specified in PO headers in `read_po()`. Fixes #17.
cmlenz
parents:
134
diff
changeset
|
116 ((u'bar', u'baz'), ('', '')) |
d62c63280e81
Respect charset specified in PO headers in `read_po()`. Fixes #17.
cmlenz
parents:
134
diff
changeset
|
117 ([(u'main.py', 3)], set([])) |
d62c63280e81
Respect charset specified in PO headers in `read_po()`. Fixes #17.
cmlenz
parents:
134
diff
changeset
|
118 ([u'A user comment'], [u'An auto comment']) |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
119 |
1 | 120 :param fileobj: the file-like object to read the PO file from |
196
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
121 :param locale: the locale identifier or `Locale` object, or `None` |
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
122 if the catalog is not bound to a locale (which basically |
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
123 means it's a template) |
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
124 :param domain: the message domain |
227 | 125 :param ignore_obsolete: whether to ignore obsolete messages in the input |
334 | 126 :return: a catalog object representing the parsed PO file |
127 :rtype: `Catalog` | |
1 | 128 """ |
196
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
129 catalog = Catalog(locale=locale, domain=domain) |
64 | 130 |
196
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
131 counter = [0] |
220
97b4b289e792
Added infrastructure for adding catalog checkers, and implement a checker that validations Python format parameters in translations, closing #19.
cmlenz
parents:
203
diff
changeset
|
132 offset = [0] |
6
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
133 messages = [] |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
134 translations = [] |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
135 locations = [] |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
136 flags = [] |
105
c62b68a0b65e
`Message`, `read_po` and `write_po` now all handle user/auto comments correctly.
palgarvio
parents:
104
diff
changeset
|
137 user_comments = [] |
c62b68a0b65e
`Message`, `read_po` and `write_po` now all handle user/auto comments correctly.
palgarvio
parents:
104
diff
changeset
|
138 auto_comments = [] |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
139 obsolete = [False] |
335 | 140 context = [] |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
141 in_msgid = [False] |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
142 in_msgstr = [False] |
342
2ee7dc04836c
Fixed a bug in pofile (in_msgctxt was not defined). Test follows.
aronacher
parents:
335
diff
changeset
|
143 in_msgctxt = [False] |
6
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
144 |
64 | 145 def _add_message(): |
6
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
146 translations.sort() |
64 | 147 if len(messages) > 1: |
106
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
148 msgid = tuple([denormalize(m) for m in messages]) |
64 | 149 else: |
106
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
150 msgid = denormalize(messages[0]) |
370
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
151 if isinstance(msgid, (list, tuple)): |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
152 string = [] |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
153 for idx in range(catalog.num_plurals): |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
154 try: |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
155 string.append(translations[idx]) |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
156 except IndexError: |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
157 string.append((idx, '')) |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
158 string = tuple([denormalize(t[1]) for t in string]) |
64 | 159 else: |
106
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
160 string = denormalize(translations[0][1]) |
335 | 161 if context: |
162 msgctxt = denormalize('\n'.join(context)) | |
163 else: | |
164 msgctxt = None | |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
165 message = Message(msgid, string, list(locations), set(flags), |
335 | 166 auto_comments, user_comments, lineno=offset[0] + 1, |
167 context=msgctxt) | |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
168 if obsolete[0]: |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
169 if not ignore_obsolete: |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
170 catalog.obsolete[msgid] = message |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
171 else: |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
172 catalog[msgid] = message |
335 | 173 del messages[:]; del translations[:]; del context[:]; del locations[:]; |
174 del flags[:]; del auto_comments[:]; del user_comments[:]; | |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
175 obsolete[0] = False |
196
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
176 counter[0] += 1 |
6
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
177 |
220
97b4b289e792
Added infrastructure for adding catalog checkers, and implement a checker that validations Python format parameters in translations, closing #19.
cmlenz
parents:
203
diff
changeset
|
178 def _process_message_line(lineno, line): |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
179 if line.startswith('msgid_plural'): |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
180 in_msgid[0] = True |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
181 msg = line[12:].lstrip() |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
182 messages.append(msg) |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
183 elif line.startswith('msgid'): |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
184 in_msgid[0] = True |
220
97b4b289e792
Added infrastructure for adding catalog checkers, and implement a checker that validations Python format parameters in translations, closing #19.
cmlenz
parents:
203
diff
changeset
|
185 offset[0] = lineno |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
186 txt = line[5:].lstrip() |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
187 if messages: |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
188 _add_message() |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
189 messages.append(txt) |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
190 elif line.startswith('msgstr'): |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
191 in_msgid[0] = False |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
192 in_msgstr[0] = True |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
193 msg = line[6:].lstrip() |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
194 if msg.startswith('['): |
441
942afedbb3d7
Make sure to only strip on the first occurence of ].
jruigrok
parents:
428
diff
changeset
|
195 idx, msg = msg[1:].split(']', 1) |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
196 translations.append([int(idx), msg.lstrip()]) |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
197 else: |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
198 translations.append([0, msg]) |
335 | 199 elif line.startswith('msgctxt'): |
428
a974d298400f
Fix for msgctxt parsing in PO files. Thanks to Asheesh Laroia for the patch. Closes #159.
cmlenz
parents:
423
diff
changeset
|
200 if messages: |
a974d298400f
Fix for msgctxt parsing in PO files. Thanks to Asheesh Laroia for the patch. Closes #159.
cmlenz
parents:
423
diff
changeset
|
201 _add_message() |
335 | 202 in_msgid[0] = in_msgstr[0] = False |
203 context.append(line[7:].lstrip()) | |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
204 elif line.startswith('"'): |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
205 if in_msgid[0]: |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
206 messages[-1] += u'\n' + line.rstrip() |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
207 elif in_msgstr[0]: |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
208 translations[-1][1] += u'\n' + line.rstrip() |
335 | 209 elif in_msgctxt[0]: |
210 context.append(line.rstrip()) | |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
211 |
220
97b4b289e792
Added infrastructure for adding catalog checkers, and implement a checker that validations Python format parameters in translations, closing #19.
cmlenz
parents:
203
diff
changeset
|
212 for lineno, line in enumerate(fileobj.readlines()): |
414
05487ae7696e
fix Python 2.3 compat: rearrange set/itemgetter/rsplit/sorted/unicode.decode
pjenvey
parents:
370
diff
changeset
|
213 line = line.strip() |
05487ae7696e
fix Python 2.3 compat: rearrange set/itemgetter/rsplit/sorted/unicode.decode
pjenvey
parents:
370
diff
changeset
|
214 if not isinstance(line, unicode): |
05487ae7696e
fix Python 2.3 compat: rearrange set/itemgetter/rsplit/sorted/unicode.decode
pjenvey
parents:
370
diff
changeset
|
215 line = line.decode(catalog.charset) |
1 | 216 if line.startswith('#'): |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
217 in_msgid[0] = in_msgstr[0] = False |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
218 if messages and translations: |
106
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
219 _add_message() |
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
220 if line[1:].startswith(':'): |
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
221 for location in line[2:].lstrip().split(): |
356
4cdca48fc832
Fixed #59 by falling back silently on invalid location comments.
aronacher
parents:
342
diff
changeset
|
222 pos = location.rfind(':') |
4cdca48fc832
Fixed #59 by falling back silently on invalid location comments.
aronacher
parents:
342
diff
changeset
|
223 if pos >= 0: |
4cdca48fc832
Fixed #59 by falling back silently on invalid location comments.
aronacher
parents:
342
diff
changeset
|
224 try: |
4cdca48fc832
Fixed #59 by falling back silently on invalid location comments.
aronacher
parents:
342
diff
changeset
|
225 lineno = int(location[pos + 1:]) |
4cdca48fc832
Fixed #59 by falling back silently on invalid location comments.
aronacher
parents:
342
diff
changeset
|
226 except ValueError: |
4cdca48fc832
Fixed #59 by falling back silently on invalid location comments.
aronacher
parents:
342
diff
changeset
|
227 continue |
4cdca48fc832
Fixed #59 by falling back silently on invalid location comments.
aronacher
parents:
342
diff
changeset
|
228 locations.append((location[:pos], lineno)) |
106
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
229 elif line[1:].startswith(','): |
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
230 for flag in line[2:].lstrip().split(','): |
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
231 flags.append(flag.strip()) |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
232 elif line[1:].startswith('~'): |
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
233 obsolete[0] = True |
220
97b4b289e792
Added infrastructure for adding catalog checkers, and implement a checker that validations Python format parameters in translations, closing #19.
cmlenz
parents:
203
diff
changeset
|
234 _process_message_line(lineno, line[2:].lstrip()) |
106
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
235 elif line[1:].startswith('.'): |
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
236 # These are called auto-comments |
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
237 comment = line[2:].strip() |
199
a0d22f2f2df0
Handle obsolete messages when parsing catalogs. Closes #32.
cmlenz
parents:
196
diff
changeset
|
238 if comment: # Just check that we're not adding empty comments |
106
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
239 auto_comments.append(comment) |
120 | 240 else: |
106
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
241 # These are called user comments |
120 | 242 user_comments.append(line[1:].strip()) |
104
395704fda00b
Merged `write_pot` and `write_po` functions by moving more functionality to the `Catalog` class. This is certainly not perfect yet, but moves us in the right direction.
cmlenz
parents:
103
diff
changeset
|
243 else: |
220
97b4b289e792
Added infrastructure for adding catalog checkers, and implement a checker that validations Python format parameters in translations, closing #19.
cmlenz
parents:
203
diff
changeset
|
244 _process_message_line(lineno, line) |
6
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
245 |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
246 if messages: |
64 | 247 _add_message() |
196
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
248 |
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
249 # No actual messages found, but there was some info in comments, from which |
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
250 # we'll construct an empty header message |
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
251 elif not counter[0] and (flags or user_comments or auto_comments): |
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
252 messages.append(u'') |
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
253 translations.append([0, u'']) |
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
254 _add_message() |
b38a6b220ea2
Fix for #35, and a minor improvement to how we parse the catalog fuzzy bit.
cmlenz
parents:
191
diff
changeset
|
255 |
64 | 256 return catalog |
1 | 257 |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
258 WORD_SEP = re.compile('(' |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
259 r'\s+|' # any whitespace |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
260 r'[^\s\w]*\w+[a-zA-Z]-(?=\w+[a-zA-Z])|' # hyphenated words |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
261 r'(?<=[\w\!\"\'\&\.\,\?])-{2,}(?=\w)' # em-dash |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
262 ')') |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
263 |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
264 def escape(string): |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
265 r"""Escape the given string so that it can be included in double-quoted |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
266 strings in ``PO`` files. |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
267 |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
268 >>> escape('''Say: |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
269 ... "hello, world!" |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
270 ... ''') |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
271 '"Say:\\n \\"hello, world!\\"\\n"' |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
272 |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
273 :param string: the string to escape |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
274 :return: the escaped string |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
275 :rtype: `str` or `unicode` |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
276 """ |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
277 return '"%s"' % string.replace('\\', '\\\\') \ |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
278 .replace('\t', '\\t') \ |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
279 .replace('\r', '\\r') \ |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
280 .replace('\n', '\\n') \ |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
281 .replace('\"', '\\"') |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
282 |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
283 def normalize(string, prefix='', width=76): |
106
2cd83f77cc98
Fix for #16: the header message (`msgid = ""`) is now treated specially by `read_po` and `Catalog`.
cmlenz
parents:
105
diff
changeset
|
284 r"""Convert a string into a format that is appropriate for .po files. |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
285 |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
286 >>> print normalize('''Say: |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
287 ... "hello, world!" |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
288 ... ''', width=None) |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
289 "" |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
290 "Say:\n" |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
291 " \"hello, world!\"\n" |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
292 |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
293 >>> print normalize('''Say: |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
294 ... "Lorem ipsum dolor sit amet, consectetur adipisicing elit, " |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
295 ... ''', width=32) |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
296 "" |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
297 "Say:\n" |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
298 " \"Lorem ipsum dolor sit " |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
299 "amet, consectetur adipisicing" |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
300 " elit, \"\n" |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
301 |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
302 :param string: the string to normalize |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
303 :param prefix: a string that should be prepended to every line |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
304 :param width: the maximum line width; use `None`, 0, or a negative number |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
305 to completely disable line wrapping |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
306 :return: the normalized string |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
307 :rtype: `unicode` |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
308 """ |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
309 if width and width > 0: |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
310 prefixlen = len(prefix) |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
311 lines = [] |
568 | 312 for line in string.splitlines(True): |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
313 if len(escape(line)) + prefixlen > width: |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
314 chunks = WORD_SEP.split(line) |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
315 chunks.reverse() |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
316 while chunks: |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
317 buf = [] |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
318 size = 2 |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
319 while chunks: |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
320 l = len(escape(chunks[-1])) - 2 + prefixlen |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
321 if size + l < width: |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
322 buf.append(chunks.pop()) |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
323 size += l |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
324 else: |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
325 if not buf: |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
326 # handle long chunks by putting them on a |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
327 # separate line |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
328 buf.append(chunks.pop()) |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
329 break |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
330 lines.append(u''.join(buf)) |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
331 else: |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
332 lines.append(line) |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
333 else: |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
334 lines = string.splitlines(True) |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
335 |
67 | 336 if len(lines) <= 1: |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
337 return escape(string) |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
338 |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
339 # Remove empty trailing line |
67 | 340 if lines and not lines[-1]: |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
341 del lines[-1] |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
342 lines[-1] += '\n' |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
343 return u'""\n' + u'\n'.join([(prefix + escape(l)) for l in lines]) |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
344 |
104
395704fda00b
Merged `write_pot` and `write_po` functions by moving more functionality to the `Catalog` class. This is certainly not perfect yet, but moves us in the right direction.
cmlenz
parents:
103
diff
changeset
|
345 def write_po(fileobj, catalog, width=76, no_location=False, omit_header=False, |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
346 sort_output=False, sort_by_file=False, ignore_obsolete=False, |
203 | 347 include_previous=False): |
56
f40fc143439c
Add actual data structures for handling message catalogs, so that more code can be reused here between the frontends.
cmlenz
parents:
55
diff
changeset
|
348 r"""Write a ``gettext`` PO (portable object) template file for a given |
f40fc143439c
Add actual data structures for handling message catalogs, so that more code can be reused here between the frontends.
cmlenz
parents:
55
diff
changeset
|
349 message catalog to the provided file-like object. |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
350 |
56
f40fc143439c
Add actual data structures for handling message catalogs, so that more code can be reused here between the frontends.
cmlenz
parents:
55
diff
changeset
|
351 >>> catalog = Catalog() |
f40fc143439c
Add actual data structures for handling message catalogs, so that more code can be reused here between the frontends.
cmlenz
parents:
55
diff
changeset
|
352 >>> catalog.add(u'foo %(name)s', locations=[('main.py', 1)], |
f40fc143439c
Add actual data structures for handling message catalogs, so that more code can be reused here between the frontends.
cmlenz
parents:
55
diff
changeset
|
353 ... flags=('fuzzy',)) |
544
ea0254950175
catalog.add() now returns the message instance (closes #245)
fschwarz
parents:
531
diff
changeset
|
354 <Message...> |
56
f40fc143439c
Add actual data structures for handling message catalogs, so that more code can be reused here between the frontends.
cmlenz
parents:
55
diff
changeset
|
355 >>> catalog.add((u'bar', u'baz'), locations=[('main.py', 3)]) |
544
ea0254950175
catalog.add() now returns the message instance (closes #245)
fschwarz
parents:
531
diff
changeset
|
356 <Message...> |
1 | 357 >>> from StringIO import StringIO |
358 >>> buf = StringIO() | |
104
395704fda00b
Merged `write_pot` and `write_po` functions by moving more functionality to the `Catalog` class. This is certainly not perfect yet, but moves us in the right direction.
cmlenz
parents:
103
diff
changeset
|
359 >>> write_po(buf, catalog, omit_header=True) |
1 | 360 >>> print buf.getvalue() |
361 #: main.py:1 | |
6
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
362 #, fuzzy, python-format |
c3b1b0b3d129
Add basic PO file parsing, and change the PO writing procedure to also take flags (such as "python-format" or "fuzzy").
cmlenz
parents:
5
diff
changeset
|
363 msgid "foo %(name)s" |
1 | 364 msgstr "" |
365 <BLANKLINE> | |
366 #: main.py:3 | |
367 msgid "bar" | |
368 msgid_plural "baz" | |
369 msgstr[0] "" | |
370 msgstr[1] "" | |
371 <BLANKLINE> | |
372 <BLANKLINE> | |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
373 |
1 | 374 :param fileobj: the file-like object to write to |
67 | 375 :param catalog: the `Catalog` instance |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
376 :param width: the maximum line width for the generated output; use `None`, |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
377 0, or a negative number to completely disable line wrapping |
1 | 378 :param no_location: do not emit a location comment for every message |
379 :param omit_header: do not include the ``msgid ""`` entry at the top of the | |
380 output | |
227 | 381 :param sort_output: whether to sort the messages in the output by msgid |
382 :param sort_by_file: whether to sort the messages in the output by their | |
383 locations | |
384 :param ignore_obsolete: whether to ignore obsolete messages and not include | |
385 them in the output; by default they are included as | |
386 comments | |
203 | 387 :param include_previous: include the old msgid as a comment when |
229 | 388 updating the catalog |
1 | 389 """ |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
390 def _normalize(key, prefix=''): |
547
1631d555d7e4
babel.messages.pofile should only apply encoding when actually writing a file (eases Python 3 transition, closes #251)
fschwarz
parents:
544
diff
changeset
|
391 return normalize(key, prefix=prefix, width=width) |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
392 |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
393 def _write(text): |
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
394 if isinstance(text, unicode): |
547
1631d555d7e4
babel.messages.pofile should only apply encoding when actually writing a file (eases Python 3 transition, closes #251)
fschwarz
parents:
544
diff
changeset
|
395 text = text.encode(catalog.charset, 'backslashreplace') |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
396 fileobj.write(text) |
1 | 397 |
181
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
398 def _write_comment(comment, prefix=''): |
423
fb926f48efba
Now, the `--width` option, although with a default value of 76, it's not set to any value initially so that the `--no-wrap` option can be passed without throwing an error. Fixes #145.
palgarvio
parents:
421
diff
changeset
|
399 # xgettext always wraps comments even if --no-wrap is passed; |
fb926f48efba
Now, the `--width` option, although with a default value of 76, it's not set to any value initially so that the `--no-wrap` option can be passed without throwing an error. Fixes #145.
palgarvio
parents:
421
diff
changeset
|
400 # provide the same behaviour |
181
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
401 if width and width > 0: |
423
fb926f48efba
Now, the `--width` option, although with a default value of 76, it's not set to any value initially so that the `--no-wrap` option can be passed without throwing an error. Fixes #145.
palgarvio
parents:
421
diff
changeset
|
402 _width = width |
fb926f48efba
Now, the `--width` option, although with a default value of 76, it's not set to any value initially so that the `--no-wrap` option can be passed without throwing an error. Fixes #145.
palgarvio
parents:
421
diff
changeset
|
403 else: |
fb926f48efba
Now, the `--width` option, although with a default value of 76, it's not set to any value initially so that the `--no-wrap` option can be passed without throwing an error. Fixes #145.
palgarvio
parents:
421
diff
changeset
|
404 _width = 76 |
fb926f48efba
Now, the `--width` option, although with a default value of 76, it's not set to any value initially so that the `--no-wrap` option can be passed without throwing an error. Fixes #145.
palgarvio
parents:
421
diff
changeset
|
405 for line in wraptext(comment, _width): |
181
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
406 _write('#%s %s\n' % (prefix, line.strip())) |
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
407 |
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
408 def _write_message(message, prefix=''): |
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
409 if isinstance(message.id, (list, tuple)): |
421 | 410 if message.context: |
411 _write('%smsgctxt %s\n' % (prefix, | |
412 _normalize(message.context, prefix))) | |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
413 _write('%smsgid %s\n' % (prefix, _normalize(message.id[0], prefix))) |
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
414 _write('%smsgid_plural %s\n' % ( |
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
415 prefix, _normalize(message.id[1], prefix) |
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
416 )) |
370
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
417 |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
418 for idx in range(catalog.num_plurals): |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
419 try: |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
420 string = message.string[idx] |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
421 except IndexError: |
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
422 string = '' |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
423 _write('%smsgstr[%d] %s\n' % ( |
370
bc18179832b7
We no longer neglect `catalog.plurals`. Added tests for it. Fixes #120.
palgarvio
parents:
356
diff
changeset
|
424 prefix, idx, _normalize(string, prefix) |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
425 )) |
181
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
426 else: |
421 | 427 if message.context: |
428 _write('%smsgctxt %s\n' % (prefix, | |
429 _normalize(message.context, prefix))) | |
190
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
430 _write('%smsgid %s\n' % (prefix, _normalize(message.id, prefix))) |
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
431 _write('%smsgstr %s\n' % ( |
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
432 prefix, _normalize(message.string or '', prefix) |
5041d90edf0c
Correctly write out obsolete messages spanning multiple lines. Fixes #33.
cmlenz
parents:
181
diff
changeset
|
433 )) |
181
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
434 |
104
395704fda00b
Merged `write_pot` and `write_po` functions by moving more functionality to the `Catalog` class. This is certainly not perfect yet, but moves us in the right direction.
cmlenz
parents:
103
diff
changeset
|
435 messages = list(catalog) |
71 | 436 if sort_output: |
248
f0b1ee94628c
add a __cmp__ to Message that correctly sorts by id, taking into account plurals
pjenvey
parents:
229
diff
changeset
|
437 messages.sort() |
71 | 438 elif sort_by_file: |
439 messages.sort(lambda x,y: cmp(x.locations, y.locations)) | |
68 | 440 |
71 | 441 for message in messages: |
67 | 442 if not message.id: # This is the header "message" |
443 if omit_header: | |
444 continue | |
104
395704fda00b
Merged `write_pot` and `write_po` functions by moving more functionality to the `Catalog` class. This is certainly not perfect yet, but moves us in the right direction.
cmlenz
parents:
103
diff
changeset
|
445 comment_header = catalog.header_comment |
103
dacfbaf0d1e0
Implement wrapping of header comments in PO(T) output. Related to #14.
cmlenz
parents:
102
diff
changeset
|
446 if width and width > 0: |
dacfbaf0d1e0
Implement wrapping of header comments in PO(T) output. Related to #14.
cmlenz
parents:
102
diff
changeset
|
447 lines = [] |
104
395704fda00b
Merged `write_pot` and `write_po` functions by moving more functionality to the `Catalog` class. This is certainly not perfect yet, but moves us in the right direction.
cmlenz
parents:
103
diff
changeset
|
448 for line in comment_header.splitlines(): |
315 | 449 lines += wraptext(line, width=width, |
450 subsequent_indent='# ') | |
581 | 451 comment_header = u'\n'.join(lines) |
452 _write(comment_header + u'\n') | |
102
14a3d766a701
Project name and version, and the charset are available via the `Catalog` object, and do not need to be passed to `write_pot()`.
cmlenz
parents:
97
diff
changeset
|
453 |
227 | 454 for comment in message.user_comments: |
181
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
455 _write_comment(comment) |
227 | 456 for comment in message.auto_comments: |
181
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
457 _write_comment(comment, prefix='.') |
1 | 458 |
459 if not no_location: | |
134 | 460 locs = u' '.join([u'%s:%d' % (filename.replace(os.sep, '/'), lineno) |
461 for filename, lineno in message.locations]) | |
181
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
462 _write_comment(locs, prefix=':') |
56
f40fc143439c
Add actual data structures for handling message catalogs, so that more code can be reused here between the frontends.
cmlenz
parents:
55
diff
changeset
|
463 if message.flags: |
f40fc143439c
Add actual data structures for handling message catalogs, so that more code can be reused here between the frontends.
cmlenz
parents:
55
diff
changeset
|
464 _write('#%s\n' % ', '.join([''] + list(message.flags))) |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
465 |
203 | 466 if message.previous_id and include_previous: |
309
43dca73da5b5
Fix for unicode problem when the previous message id is included as a comment in PO serialization. Closes #78.
cmlenz
parents:
248
diff
changeset
|
467 _write_comment('msgid %s' % _normalize(message.previous_id[0]), |
203 | 468 prefix='|') |
469 if len(message.previous_id) > 1: | |
309
43dca73da5b5
Fix for unicode problem when the previous message id is included as a comment in PO serialization. Closes #78.
cmlenz
parents:
248
diff
changeset
|
470 _write_comment('msgid_plural %s' % _normalize( |
203 | 471 message.previous_id[1] |
472 ), prefix='|') | |
200
1c778cccd330
Added `--no-fuzzy-matching` to the frontends and also `--previous` which adds the old msgid's as comments. The latest closes #31.
palgarvio
parents:
199
diff
changeset
|
473 |
181
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
474 _write_message(message) |
24
b09e90803d1b
Reimplement line wrapping for PO writing (as the `textwrap` module is too destructive with white space) and move it to the `normalize` function (which was already doing some handling of line breaks).
cmlenz
parents:
23
diff
changeset
|
475 _write('\n') |
181
8a762ce37bf7
The frontends now provide ways to update existing translations catalogs from a template. Closes #22.
cmlenz
parents:
178
diff
changeset
|
476 |
191
c171a0041ad2
Add an option to the frontend commands for catalog updating that removes completely any obsolete messages, instead of putting them comments.
cmlenz
parents:
190
diff
changeset
|
477 if not ignore_obsolete: |
c171a0041ad2
Add an option to the frontend commands for catalog updating that removes completely any obsolete messages, instead of putting them comments.
cmlenz
parents:
190
diff
changeset
|
478 for message in catalog.obsolete.values(): |
227 | 479 for comment in message.user_comments: |
191
c171a0041ad2
Add an option to the frontend commands for catalog updating that removes completely any obsolete messages, instead of putting them comments.
cmlenz
parents:
190
diff
changeset
|
480 _write_comment(comment) |
c171a0041ad2
Add an option to the frontend commands for catalog updating that removes completely any obsolete messages, instead of putting them comments.
cmlenz
parents:
190
diff
changeset
|
481 _write_message(message, prefix='#~ ') |
c171a0041ad2
Add an option to the frontend commands for catalog updating that removes completely any obsolete messages, instead of putting them comments.
cmlenz
parents:
190
diff
changeset
|
482 _write('\n') |