Mercurial > pyntnclick

comparison scripts/msgfmt.py @ 779:087d3ff0aa08 pyntnclick

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Move po scripts into scripts. Bundle msgfmt.py from cpython too
author Stefano Rivera <stefano@rivera.za.net>
date Sat, 26 Jan 2013 18:38:27 +0200
parents
children 99e1e67f3916
comparison
equal deleted inserted replaced
778:98cdb7458225 779:087d3ff0aa08
1 #! /usr/bin/env python
2 # -*- coding: iso-8859-1 -*-
3 # Written by Martin v. Löwis <loewis@informatik.hu-berlin.de>
4
5 """Generate binary message catalog from textual translation description.
6
7 This program converts a textual Uniforum-style message catalog (.po file) into
8 a binary GNU catalog (.mo file). This is essentially the same function as the
9 GNU msgfmt program, however, it is a simpler implementation.
10
11 Usage: msgfmt.py [OPTIONS] filename.po
12
13 Options:
14 -o file
15 --output-file=file
16 Specify the output file to write to. If omitted, output will go to a
17 file named filename.mo (based off the input file name).
18
19 -h
20 --help
21 Print this message and exit.
22
23 -V
24 --version
25 Display version information and exit.
26 """
27
28 import os
29 import sys
30 import ast
31 import getopt
32 import struct
33 import array
34
35 __version__ = "1.1"
36
37 MESSAGES = {}
38
39
40
41 def usage(code, msg=''):
42 print >> sys.stderr, __doc__
43 if msg:
44 print >> sys.stderr, msg
45 sys.exit(code)
46
47
48
49 def add(id, str, fuzzy):
50 "Add a non-fuzzy translation to the dictionary."
51 global MESSAGES
52 if not fuzzy and str:
53 MESSAGES[id] = str
54
55
56
57 def generate():
58 "Return the generated output."
59 global MESSAGES
60 keys = MESSAGES.keys()
61 # the keys are sorted in the .mo file
62 keys.sort()
63 offsets = []
64 ids = strs = ''
65 for id in keys:
66 # For each string, we need size and file offset. Each string is NUL
67 # terminated; the NUL does not count into the size.
68 offsets.append((len(ids), len(id), len(strs), len(MESSAGES[id])))
69 ids += id + '\0'
70 strs += MESSAGES[id] + '\0'
71 output = ''
72 # The header is 7 32-bit unsigned integers. We don't use hash tables, so
73 # the keys start right after the index tables.
74 # translated string.
75 keystart = 7*4+16*len(keys)
76 # and the values start after the keys
77 valuestart = keystart + len(ids)
78 koffsets = []
79 voffsets = []
80 # The string table first has the list of keys, then the list of values.
81 # Each entry has first the size of the string, then the file offset.
82 for o1, l1, o2, l2 in offsets:
83 koffsets += [l1, o1+keystart]
84 voffsets += [l2, o2+valuestart]
85 offsets = koffsets + voffsets
86 output = struct.pack("Iiiiiii",
87 0x950412deL, # Magic
88 0, # Version
89 len(keys), # # of entries
90 7*4, # start of key index
91 7*4+len(keys)*8, # start of value index
92 0, 0) # size and offset of hash table
93 output += array.array("i", offsets).tostring()
94 output += ids
95 output += strs
96 return output
97
98
99
100 def make(filename, outfile):
101 ID = 1
102 STR = 2
103
104 # Compute .mo name from .po name and arguments
105 if filename.endswith('.po'):
106 infile = filename
107 else:
108 infile = filename + '.po'
109 if outfile is None:
110 outfile = os.path.splitext(infile)[0] + '.mo'
111
112 try:
113 lines = open(infile).readlines()
114 except IOError, msg:
115 print >> sys.stderr, msg
116 sys.exit(1)
117
118 section = None
119 fuzzy = 0
120
121 # Parse the catalog
122 lno = 0
123 for l in lines:
124 lno += 1
125 # If we get a comment line after a msgstr, this is a new entry
126 if l[0] == '#' and section == STR:
127 add(msgid, msgstr, fuzzy)
128 section = None
129 fuzzy = 0
130 # Record a fuzzy mark
131 if l[:2] == '#,' and 'fuzzy' in l:
132 fuzzy = 1
133 # Skip comments
134 if l[0] == '#':
135 continue
136 # Now we are in a msgid section, output previous section
137 if l.startswith('msgid') and not l.startswith('msgid_plural'):
138 if section == STR:
139 add(msgid, msgstr, fuzzy)
140 section = ID
141 l = l[5:]
142 msgid = msgstr = ''
143 is_plural = False
144 # This is a message with plural forms
145 elif l.startswith('msgid_plural'):
146 if section != ID:
147 print >> sys.stderr, 'msgid_plural not preceeded by msgid on %s:%d' %\
148 (infile, lno)
149 sys.exit(1)
150 l = l[12:]
151 msgid += '\0' # separator of singular and plural
152 is_plural = True
153 # Now we are in a msgstr section
154 elif l.startswith('msgstr'):
155 section = STR
156 if l.startswith('msgstr['):
157 if not is_plural:
158 print >> sys.stderr, 'plural without msgid_plural on %s:%d' %\
159 (infile, lno)
160 sys.exit(1)
161 l = l.split(']', 1)[1]
162 if msgstr:
163 msgstr += '\0' # Separator of the various plural forms
164 else:
165 if is_plural:
166 print >> sys.stderr, 'indexed msgstr required for plural on %s:%d' %\
167 (infile, lno)
168 sys.exit(1)
169 l = l[6:]
170 # Skip empty lines
171 l = l.strip()
172 if not l:
173 continue
174 l = ast.literal_eval(l)
175 if section == ID:
176 msgid += l
177 elif section == STR:
178 msgstr += l
179 else:
180 print >> sys.stderr, 'Syntax error on %s:%d' % (infile, lno), \
181 'before:'
182 print >> sys.stderr, l
183 sys.exit(1)
184 # Add last entry
185 if section == STR:
186 add(msgid, msgstr, fuzzy)
187
188 # Compute output
189 output = generate()
190
191 try:
192 open(outfile,"wb").write(output)
193 except IOError,msg:
194 print >> sys.stderr, msg
195
196
197
198 def main():
199 try:
200 opts, args = getopt.getopt(sys.argv[1:], 'hVo:',
201 ['help', 'version', 'output-file='])
202 except getopt.error, msg:
203 usage(1, msg)
204
205 outfile = None
206 # parse options
207 for opt, arg in opts:
208 if opt in ('-h', '--help'):
209 usage(0)
210 elif opt in ('-V', '--version'):
211 print >> sys.stderr, "msgfmt.py", __version__
212 sys.exit(0)
213 elif opt in ('-o', '--output-file'):
214 outfile = arg
215 # do it
216 if not args:
217 print >> sys.stderr, 'No input file given'
218 print >> sys.stderr, "Try `msgfmt --help' for more information."
219 return
220
221 for filename in args:
222 make(filename, outfile)
223
224
225 if __name__ == '__main__':
226 main()