« back to all changes in this revision
Viewing changes to edit-here/translations.py
- browse files at revision 54
- compare with another revision
- download diff
- download tarball
- view history from revision 54
- Committer: Daniel Holbach
- Date: 2015-02-26 12:46:07 UTC
- mfrom: (48.1.48 help.1425010)
- Revision ID: daniel.holbach@canonical.com-20150226124607-d8r1pespk41a1wh3
merged lp:~dholbach/ubuntu-devices-help/1425010
added
removed
edit-here/translations.py
6
6
import re
7
7
import shutil
8
8
import subprocess
9
import tempfile
10
9
11
10
from pelicanconf import PATH
12
11
13
METADATA_TAGS = [
14
'title',
15
'lang',
16
'date',
17
]
18
19
12
# This defines how complete we expect translations to be before we
20
13
# generate HTML from them. Needs to be string.
21
14
TRANSLATION_COMPLETION_PERCENTAGE = '0'
22
15
16
BCP47_OVERRIDES = (
17
('zh_CN', 'zh-hans'),
18
('zh_TW', 'zh-hant'),
19
)
20
23
21
24
22
class PO4A(object):
25
def __init__(self, translations_dir, temp_dir):
23
def __init__(self, pot_file):
26
24
self.default_args = [
27
25
'-f', 'text',
28
26
'-o', 'markdown',
29
27
'-M', 'utf-8',
30
28
]
31
self.translations_dir = translations_dir
32
self.temp_dir = temp_dir
29
self.pot_file = pot_file
33
30
34
def run(self, po4a_command, additional_args, with_output=False,
35
working_dir=None):
36
if not working_dir:
37
working_dir = self.temp_dir
38
pwd = os.getcwd()
31
def run(self, po4a_command, additional_args, with_output=False):
39
32
args = copy.copy(self.default_args)
40
33
args += additional_args
41
os.chdir(working_dir)
42
34
if with_output:
43
35
ret = subprocess.Popen([po4a_command]+args, stdout=subprocess.PIPE)
44
36
else:
45
37
ret = subprocess.call([po4a_command]+args)
46
os.chdir(pwd)
47
38
return ret
48
39
49
40
def gettextize(self, documents):
51
42
for document in documents:
52
43
args += ['-m', document]
53
44
args += [
54
'-p', os.path.join(self.translations_dir, 'help.pot'),
45
'-p', self.pot_file,
55
46
'-L', 'utf-8',
56
47
]
57
48
return self.run('po4a-gettextize', args)
74
65
'-p', lang.file_name,
75
66
'-L', 'utf-8',
76
67
]
77
return self.run('po4a-translate', args, with_output=True,
78
working_dir=os.path.join(PATH, '..'))
68
return self.run('po4a-translate', args, with_output=True)
69
70
71
class Documents(object):
72
def __init__(self):
73
self.docs = []
74
for dirpath, dirnames, filenames in os.walk(PATH):
75
for filename in filenames:
76
self.docs += [os.path.join(dirpath, filename)]
77
78
def translated_doc(self, file_name, lang):
79
match = [doc for doc in self.docs
80
if os.path.basename(doc) == os.path.basename(file_name)]
81
if not match:
82
return None
83
return '%s.%s.md' % (match[0].split('.md')[0],
84
lang.bcp47_code)
85
86
def _call_po4a_translate(self, doc, lang, po4a):
87
res = po4a.translate(doc, lang)
88
output = codecs.decode(res.communicate()[0])
89
broken_title_line = [line for line in output.split('\n')
90
if line.lower().startswith('title:')][0]
91
rest = [line for line in output.split('\n')
92
if not line.lower().startswith('title')]
93
output = '\n'.join(rest)
94
return (broken_title_line, output)
95
96
def write_translated_markdown(self, lang, po4a):
97
for doc in self.docs:
98
(broken_title_line, output) = \
99
self._call_po4a_translate(doc, lang, po4a)
100
new_path = self.translated_doc(doc, lang)
101
text = "%s\nDate:\n\n" % (broken_title_line)
102
text += output
103
if os.path.exists(new_path):
104
os.remove(new_path)
105
if not os.path.exists(os.path.dirname(new_path)):
106
os.makedirs(os.path.dirname(new_path))
107
with open(new_path, 'w', encoding='utf-8') as f:
108
f.write(text)
79
109
80
110
81
111
class Language(object):
82
def __init__(self, po_file):
112
def __init__(self, po_file, documents):
83
113
self.file_name = po_file
84
self.gettext_lang_code = os.path.basename(po_file).split('.po')[0]
114
self.gettext_code = os.path.basename(po_file).split('.po')[0]
115
self.bcp47_code = self._find_bcp47_code()
116
self.documents = documents
117
118
def _find_bcp47_code(self):
119
if self.gettext_code not in [c[0] for c in BCP47_OVERRIDES]:
120
return self.gettext_code.lower().replace('_', '-')
121
return [c[1] for c in BCP47_OVERRIDES
122
if c[0] == self.gettext_code][0]
85
123
86
124
def rewrite_links(self):
87
125
po_file = polib.pofile(self.file_name)
88
for entry_group in [po_file.untranslated_entries(),
89
po_file.translated_entries(),
90
po_file.fuzzy_entries()]:
126
link_regex = r'\[.+?\]\(\{filename\}(.+?)\)'
127
for entry_group in [po_file.translated_entries(),
128
po_file.fuzzy_entries(),
129
po_file.untranslated_entries()]:
91
130
for entry in entry_group:
92
131
if '{filename}' in entry.msgid:
93
if not entry.msgstr:
132
link_msgid = re.findall(link_regex, entry.msgid)[0]
133
link_msgstr = list(re.findall(link_regex, entry.msgstr))
134
translated_doc = os.path.basename(
135
self.documents.translated_doc(
136
link_msgid, self))
137
if not link_msgstr:
94
138
entry.msgstr = entry.msgid
95
link = re.findall(r'\[.+?\]\(\{filename\}(.+?)\)',
96
entry.msgid)[0]
97
entry.msgstr = entry.msgstr.replace(
98
link,
99
'lang-%s-%s' % (self.gettext_lang_code, link))
139
link_msgstr = [link_msgid]
140
entry.msgstr = entry.msgstr.replace(link_msgstr[0],
141
translated_doc)
100
142
po_file.save(self.file_name)
101
143
102
144
105
147
self._cleanup()
106
148
self.translations_dir = os.path.abspath(os.path.join(PATH, '../po'))
107
149
self.available_languages = []
150
self.documents = Documents()
108
151
for po_filename in glob.glob(self.translations_dir+'/*.po'):
109
self.available_languages += [Language(po_filename)]
110
self.documents = self._find_documents()
111
self.temp_dir = tempfile.mkdtemp()
112
self.po4a = PO4A(self.translations_dir, self.temp_dir)
152
self.available_languages += [Language(po_filename, self.documents)]
153
self.fake_lang_code = 'en_US'
154
self.fake_po_file = os.path.join(self.translations_dir,
155
'%s.po' % self.fake_lang_code)
156
self.pot_file = os.path.join(self.translations_dir,
157
"help.pot")
158
self.po4a = PO4A(self.pot_file)
113
159
114
160
def __del__(self):
115
shutil.rmtree(self.temp_dir)
116
117
def _find_documents(self):
118
documents = []
119
for dirpath, dirnames, filenames in os.walk(PATH):
120
for filename in filenames:
121
documents += [os.path.join(dirpath, filename)]
122
return documents
161
if os.path.exists(self.fake_po_file):
162
os.remove(self.fake_po_file)
123
163
124
164
def _cleanup(self):
125
165
r = subprocess.Popen(['bzr', 'ignored'], stdout=subprocess.PIPE)
133
173
except NotADirectoryError:
134
174
os.remove(f)
135
175
136
def _remove_metadata(self, filename):
137
new_dir = os.path.join(self.temp_dir, os.path.dirname(filename))
138
os.makedirs(new_dir, exist_ok=True)
139
new_filename = os.path.join(new_dir, os.path.basename(filename))
140
shutil.copy(filename, new_filename)
141
lines = open(new_filename).readlines()
142
title_line = [x for x in lines if
143
(x.startswith('Title:') or x.startswith('title:'))]
144
index = 0
145
for line in lines:
146
if not [x for x in METADATA_TAGS
147
if line.lower().startswith(x+':')] and \
148
line.strip() != '':
149
index = lines.index(line)
150
break
151
if not title_line:
152
print('No line starting with "Title: " found in "%s".' %
153
new_filename)
154
return False
155
os.remove(new_filename)
156
with open(new_filename, 'w', encoding='utf-8') as new_file:
157
new_file.write(title_line[0]+'\n')
158
new_file.write(''.join(lines[index:]))
159
return True
160
161
def clean_documents(self):
162
for document in self.documents:
163
if not self._remove_metadata(document):
164
return False
165
return True
166
167
176
def generate_pot_file(self):
168
return self.po4a.gettextize(self.documents)
177
return self.po4a.gettextize(self.documents.docs)
169
178
170
179
def update_po_files(self):
171
return self.po4a.updatepo(self.available_languages, self.documents)
172
173
def _call_po4a_translate(self, doc, lang):
174
res = self.po4a.translate(doc, lang)
175
output = codecs.decode(res.communicate()[0])
176
broken_title_line = [line for line in output.split('\n')
177
if line.lower().startswith('title:')][0]
178
rest = [line for line in output.split('\n')
179
if not line.lower().startswith('title')]
180
output = '\n'.join(rest)
181
return (broken_title_line, output)
182
183
def _new_header(self, lang, broken_title_line):
184
title_line = broken_title_line
185
for metadata_tag in [x for x in METADATA_TAGS if not x == 'title']:
186
title_line = title_line.split(metadata_tag)[0]
187
title_line = title_line.split(metadata_tag.capitalize())[0]
188
title_line = title_line.split(metadata_tag.upper())[0]
189
return "%s\nLang: %s\nDate:\n\n" % \
190
(title_line, lang.gettext_lang_code)
180
return self.po4a.updatepo(self.available_languages,
181
self.documents.docs)
191
182
192
183
def rewrite_links(self):
184
self._generate_fake_pofile()
193
185
for lang in self.available_languages:
194
186
lang.rewrite_links()
195
187
188
# we generate a fake translation for en-US which is going to be
189
# the default
190
def _generate_fake_pofile(self):
191
if os.path.exists(self.fake_po_file):
192
os.remove(self.fake_po_file)
193
shutil.copy(self.pot_file, self.fake_po_file)
194
self.available_languages += [Language(self.fake_po_file,
195
self.documents)]
196
196
197
def generate_translations(self):
197
198
for lang in self.available_languages:
198
for doc in self.documents:
199
(broken_title_line, output) = \
200
self._call_po4a_translate(doc, lang)
201
new_path = os.path.join(PATH, 'pages',
202
'lang-%s-%s' %
203
(lang.gettext_lang_code,
204
os.path.basename(doc)))
205
text = self._new_header(lang, broken_title_line)
206
text += output
207
if os.path.exists(new_path):
208
os.remove(new_path)
209
if not os.path.exists(os.path.dirname(new_path)):
210
os.makedirs(os.path.dirname(new_path))
211
with open(new_path, 'w', encoding='utf-8') as f:
212
f.write(text)
199
self.documents.write_translated_markdown(lang, self.po4a)
Loggerhead is a web-based interface for Breezy
Version: 2.0.1