File: update-translations.py

package info (click to toggle)
litecoin 0.17.1-1
  • links: PTS, VCS
  • area: main
  • in suites: sid
  • size: 23,076 kB
  • sloc: cpp: 120,253; python: 22,203; ansic: 14,322; sh: 4,390; makefile: 1,509; asm: 730; java: 438
file content (215 lines) | stat: -rwxr-xr-x 8,426 bytes parent folder | download | duplicates (2)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
#!/usr/bin/env python3
# Copyright (c) 2014 Wladimir J. van der Laan
# Distributed under the MIT software license, see the accompanying
# file COPYING or http://www.opensource.org/licenses/mit-license.php.
'''
Run this script from the root of the repository to update all translations from
transifex.
It will do the following automatically:

- fetch all translations using the tx tool
- post-process them into valid and committable format
  - remove invalid control characters
  - remove location tags (makes diffs less noisy)

TODO:
- auto-add new translations to the build system according to the translation process
'''
import subprocess
import re
import sys
import os
import io
import xml.etree.ElementTree as ET

# Name of transifex tool
TX = 'tx'
# Name of source language file
SOURCE_LANG = 'bitcoin_en.ts'
# Directory with locale files
LOCALE_DIR = 'src/qt/locale'
# Minimum number of messages for translation to be considered at all
MIN_NUM_MESSAGES = 10
# Regexp to check for Bitcoin addresses
ADDRESS_REGEXP = re.compile('([13]|bc1)[a-zA-Z0-9]{30,}')

def check_at_repository_root():
    if not os.path.exists('.git'):
        print('No .git directory found')
        print('Execute this script at the root of the repository', file=sys.stderr)
        sys.exit(1)

def fetch_all_translations():
    if subprocess.call([TX, 'pull', '-f', '-a']):
        print('Error while fetching translations', file=sys.stderr)
        sys.exit(1)

def find_format_specifiers(s):
    '''Find all format specifiers in a string.'''
    pos = 0
    specifiers = []
    while True:
        percent = s.find('%', pos)
        if percent < 0:
            break
        specifiers.append(s[percent+1])
        pos = percent+2
    return specifiers

def split_format_specifiers(specifiers):
    '''Split format specifiers between numeric (Qt) and others (strprintf)'''
    numeric = []
    other = []
    for s in specifiers:
        if s in {'1','2','3','4','5','6','7','8','9'}:
            numeric.append(s)
        else:
            other.append(s)

    # If both numeric format specifiers and "others" are used, assume we're dealing
    # with a Qt-formatted message. In the case of Qt formatting (see https://doc.qt.io/qt-5/qstring.html#arg)
    # only numeric formats are replaced at all. This means "(percentage: %1%)" is valid, without needing
    # any kind of escaping that would be necessary for strprintf. Without this, this function
    # would wrongly detect '%)' as a printf format specifier.
    if numeric:
        other = []

    # numeric (Qt) can be present in any order, others (strprintf) must be in specified order
    return set(numeric),other

def sanitize_string(s):
    '''Sanitize string for printing'''
    return s.replace('\n',' ')

def check_format_specifiers(source, translation, errors, numerus):
    source_f = split_format_specifiers(find_format_specifiers(source))
    # assert that no source messages contain both Qt and strprintf format specifiers
    # if this fails, go change the source as this is hacky and confusing!
    assert(not(source_f[0] and source_f[1]))
    try:
        translation_f = split_format_specifiers(find_format_specifiers(translation))
    except IndexError:
        errors.append("Parse error in translation for '%s': '%s'" % (sanitize_string(source), sanitize_string(translation)))
        return False
    else:
        if source_f != translation_f:
            if numerus and source_f == (set(), ['n']) and translation_f == (set(), []) and translation.find('%') == -1:
                # Allow numerus translations to omit %n specifier (usually when it only has one possible value)
                return True
            errors.append("Mismatch between '%s' and '%s'" % (sanitize_string(source), sanitize_string(translation)))
            return False
    return True

def all_ts_files(suffix=''):
    for filename in os.listdir(LOCALE_DIR):
        # process only language files, and do not process source language
        if not filename.endswith('.ts'+suffix) or filename == SOURCE_LANG+suffix:
            continue
        if suffix: # remove provided suffix
            filename = filename[0:-len(suffix)]
        filepath = os.path.join(LOCALE_DIR, filename)
        yield(filename, filepath)

FIX_RE = re.compile(b'[\x00-\x09\x0b\x0c\x0e-\x1f]')
def remove_invalid_characters(s):
    '''Remove invalid characters from translation string'''
    return FIX_RE.sub(b'', s)

# Override cdata escape function to make our output match Qt's (optional, just for cleaner diffs for
# comparison, disable by default)
_orig_escape_cdata = None
def escape_cdata(text):
    text = _orig_escape_cdata(text)
    text = text.replace("'", '&apos;')
    text = text.replace('"', '&quot;')
    return text

def contains_bitcoin_addr(text, errors):
    if text != None and ADDRESS_REGEXP.search(text) != None:
        errors.append('Translation "%s" contains a bitcoin address. This will be removed.' % (text))
        return True
    return False

def postprocess_translations(reduce_diff_hacks=False):
    print('Checking and postprocessing...')

    if reduce_diff_hacks:
        global _orig_escape_cdata
        _orig_escape_cdata = ET._escape_cdata
        ET._escape_cdata = escape_cdata

    for (filename,filepath) in all_ts_files():
        os.rename(filepath, filepath+'.orig')

    have_errors = False
    for (filename,filepath) in all_ts_files('.orig'):
        # pre-fixups to cope with transifex output
        parser = ET.XMLParser(encoding='utf-8') # need to override encoding because 'utf8' is not understood only 'utf-8'
        with open(filepath + '.orig', 'rb') as f:
            data = f.read()
        # remove control characters; this must be done over the entire file otherwise the XML parser will fail
        data = remove_invalid_characters(data)
        tree = ET.parse(io.BytesIO(data), parser=parser)

        # iterate over all messages in file
        root = tree.getroot()
        for context in root.findall('context'):
            for message in context.findall('message'):
                numerus = message.get('numerus') == 'yes'
                source = message.find('source').text
                translation_node = message.find('translation')
                # pick all numerusforms
                if numerus:
                    translations = [i.text for i in translation_node.findall('numerusform')]
                else:
                    translations = [translation_node.text]

                for translation in translations:
                    if translation is None:
                        continue
                    errors = []
                    valid = check_format_specifiers(source, translation, errors, numerus) and not contains_bitcoin_addr(translation, errors)

                    for error in errors:
                        print('%s: %s' % (filename, error))

                    if not valid: # set type to unfinished and clear string if invalid
                        translation_node.clear()
                        translation_node.set('type', 'unfinished')
                        have_errors = True

                # Remove location tags
                for location in message.findall('location'):
                    message.remove(location)

                # Remove entire message if it is an unfinished translation
                if translation_node.get('type') == 'unfinished':
                    context.remove(message)

        # check if document is (virtually) empty, and remove it if so
        num_messages = 0
        for context in root.findall('context'):
            for message in context.findall('message'):
                num_messages += 1
        if num_messages < MIN_NUM_MESSAGES:
            print('Removing %s, as it contains only %i messages' % (filepath, num_messages))
            continue

        # write fixed-up tree
        # if diff reduction requested, replace some XML to 'sanitize' to qt formatting
        if reduce_diff_hacks:
            out = io.BytesIO()
            tree.write(out, encoding='utf-8')
            out = out.getvalue()
            out = out.replace(b' />', b'/>')
            with open(filepath, 'wb') as f:
                f.write(out)
        else:
            tree.write(filepath, encoding='utf-8')
    return have_errors

if __name__ == '__main__':
    check_at_repository_root()
    fetch_all_translations()
    postprocess_translations()