Show More
@@ -1,137 +1,141 | |||||
1 | #!/usr/bin/env python |
|
1 | #!/usr/bin/env python | |
2 | # |
|
2 | # | |
3 | # hggettext - carefully extract docstrings for Mercurial |
|
3 | # hggettext - carefully extract docstrings for Mercurial | |
4 | # |
|
4 | # | |
5 | # Copyright 2009 Matt Mackall <mpm@selenic.com> and others |
|
5 | # Copyright 2009 Matt Mackall <mpm@selenic.com> and others | |
6 | # |
|
6 | # | |
7 | # This software may be used and distributed according to the terms of the |
|
7 | # This software may be used and distributed according to the terms of the | |
8 | # GNU General Public License version 2 or any later version. |
|
8 | # GNU General Public License version 2 or any later version. | |
9 |
|
9 | |||
10 | # The normalize function is taken from pygettext which is distributed |
|
10 | # The normalize function is taken from pygettext which is distributed | |
11 | # with Python under the Python License, which is GPL compatible. |
|
11 | # with Python under the Python License, which is GPL compatible. | |
12 |
|
12 | |||
13 | """Extract docstrings from Mercurial commands. |
|
13 | """Extract docstrings from Mercurial commands. | |
14 |
|
14 | |||
15 | Compared to pygettext, this script knows about the cmdtable and table |
|
15 | Compared to pygettext, this script knows about the cmdtable and table | |
16 | dictionaries used by Mercurial, and will only extract docstrings from |
|
16 | dictionaries used by Mercurial, and will only extract docstrings from | |
17 | functions mentioned therein. |
|
17 | functions mentioned therein. | |
18 |
|
18 | |||
19 | Use xgettext like normal to extract strings marked as translatable and |
|
19 | Use xgettext like normal to extract strings marked as translatable and | |
20 | join the message cataloges to get the final catalog. |
|
20 | join the message cataloges to get the final catalog. | |
21 | """ |
|
21 | """ | |
22 |
|
22 | |||
23 | import os, sys, inspect |
|
23 | from __future__ import absolute_import | |
|
24 | ||||
|
25 | import inspect | |||
|
26 | import os | |||
|
27 | import sys | |||
24 |
|
28 | |||
25 |
|
29 | |||
26 | def escape(s): |
|
30 | def escape(s): | |
27 | # The order is important, the backslash must be escaped first |
|
31 | # The order is important, the backslash must be escaped first | |
28 | # since the other replacements introduce new backslashes |
|
32 | # since the other replacements introduce new backslashes | |
29 | # themselves. |
|
33 | # themselves. | |
30 | s = s.replace('\\', '\\\\') |
|
34 | s = s.replace('\\', '\\\\') | |
31 | s = s.replace('\n', '\\n') |
|
35 | s = s.replace('\n', '\\n') | |
32 | s = s.replace('\r', '\\r') |
|
36 | s = s.replace('\r', '\\r') | |
33 | s = s.replace('\t', '\\t') |
|
37 | s = s.replace('\t', '\\t') | |
34 | s = s.replace('"', '\\"') |
|
38 | s = s.replace('"', '\\"') | |
35 | return s |
|
39 | return s | |
36 |
|
40 | |||
37 |
|
41 | |||
38 | def normalize(s): |
|
42 | def normalize(s): | |
39 | # This converts the various Python string types into a format that |
|
43 | # This converts the various Python string types into a format that | |
40 | # is appropriate for .po files, namely much closer to C style. |
|
44 | # is appropriate for .po files, namely much closer to C style. | |
41 | lines = s.split('\n') |
|
45 | lines = s.split('\n') | |
42 | if len(lines) == 1: |
|
46 | if len(lines) == 1: | |
43 | s = '"' + escape(s) + '"' |
|
47 | s = '"' + escape(s) + '"' | |
44 | else: |
|
48 | else: | |
45 | if not lines[-1]: |
|
49 | if not lines[-1]: | |
46 | del lines[-1] |
|
50 | del lines[-1] | |
47 | lines[-1] = lines[-1] + '\n' |
|
51 | lines[-1] = lines[-1] + '\n' | |
48 | lines = map(escape, lines) |
|
52 | lines = map(escape, lines) | |
49 | lineterm = '\\n"\n"' |
|
53 | lineterm = '\\n"\n"' | |
50 | s = '""\n"' + lineterm.join(lines) + '"' |
|
54 | s = '""\n"' + lineterm.join(lines) + '"' | |
51 | return s |
|
55 | return s | |
52 |
|
56 | |||
53 |
|
57 | |||
54 | def poentry(path, lineno, s): |
|
58 | def poentry(path, lineno, s): | |
55 | return ('#: %s:%d\n' % (path, lineno) + |
|
59 | return ('#: %s:%d\n' % (path, lineno) + | |
56 | 'msgid %s\n' % normalize(s) + |
|
60 | 'msgid %s\n' % normalize(s) + | |
57 | 'msgstr ""\n') |
|
61 | 'msgstr ""\n') | |
58 |
|
62 | |||
59 |
|
63 | |||
60 | def offset(src, doc, name, default): |
|
64 | def offset(src, doc, name, default): | |
61 | """Compute offset or issue a warning on stdout.""" |
|
65 | """Compute offset or issue a warning on stdout.""" | |
62 | # Backslashes in doc appear doubled in src. |
|
66 | # Backslashes in doc appear doubled in src. | |
63 | end = src.find(doc.replace('\\', '\\\\')) |
|
67 | end = src.find(doc.replace('\\', '\\\\')) | |
64 | if end == -1: |
|
68 | if end == -1: | |
65 | # This can happen if the docstring contains unnecessary escape |
|
69 | # This can happen if the docstring contains unnecessary escape | |
66 | # sequences such as \" in a triple-quoted string. The problem |
|
70 | # sequences such as \" in a triple-quoted string. The problem | |
67 | # is that \" is turned into " and so doc wont appear in src. |
|
71 | # is that \" is turned into " and so doc wont appear in src. | |
68 | sys.stderr.write("warning: unknown offset in %s, assuming %d lines\n" |
|
72 | sys.stderr.write("warning: unknown offset in %s, assuming %d lines\n" | |
69 | % (name, default)) |
|
73 | % (name, default)) | |
70 | return default |
|
74 | return default | |
71 | else: |
|
75 | else: | |
72 | return src.count('\n', 0, end) |
|
76 | return src.count('\n', 0, end) | |
73 |
|
77 | |||
74 |
|
78 | |||
75 | def importpath(path): |
|
79 | def importpath(path): | |
76 | """Import a path like foo/bar/baz.py and return the baz module.""" |
|
80 | """Import a path like foo/bar/baz.py and return the baz module.""" | |
77 | if path.endswith('.py'): |
|
81 | if path.endswith('.py'): | |
78 | path = path[:-3] |
|
82 | path = path[:-3] | |
79 | if path.endswith('/__init__'): |
|
83 | if path.endswith('/__init__'): | |
80 | path = path[:-9] |
|
84 | path = path[:-9] | |
81 | path = path.replace('/', '.') |
|
85 | path = path.replace('/', '.') | |
82 | mod = __import__(path) |
|
86 | mod = __import__(path) | |
83 | for comp in path.split('.')[1:]: |
|
87 | for comp in path.split('.')[1:]: | |
84 | mod = getattr(mod, comp) |
|
88 | mod = getattr(mod, comp) | |
85 | return mod |
|
89 | return mod | |
86 |
|
90 | |||
87 |
|
91 | |||
88 | def docstrings(path): |
|
92 | def docstrings(path): | |
89 | """Extract docstrings from path. |
|
93 | """Extract docstrings from path. | |
90 |
|
94 | |||
91 | This respects the Mercurial cmdtable/table convention and will |
|
95 | This respects the Mercurial cmdtable/table convention and will | |
92 | only extract docstrings from functions mentioned in these tables. |
|
96 | only extract docstrings from functions mentioned in these tables. | |
93 | """ |
|
97 | """ | |
94 | mod = importpath(path) |
|
98 | mod = importpath(path) | |
95 | if mod.__doc__: |
|
99 | if mod.__doc__: | |
96 | src = open(path).read() |
|
100 | src = open(path).read() | |
97 | lineno = 1 + offset(src, mod.__doc__, path, 7) |
|
101 | lineno = 1 + offset(src, mod.__doc__, path, 7) | |
98 | print poentry(path, lineno, mod.__doc__) |
|
102 | print poentry(path, lineno, mod.__doc__) | |
99 |
|
103 | |||
100 | functions = list(getattr(mod, 'i18nfunctions', [])) |
|
104 | functions = list(getattr(mod, 'i18nfunctions', [])) | |
101 | functions = [(f, True) for f in functions] |
|
105 | functions = [(f, True) for f in functions] | |
102 |
|
106 | |||
103 | cmdtable = getattr(mod, 'cmdtable', {}) |
|
107 | cmdtable = getattr(mod, 'cmdtable', {}) | |
104 | if not cmdtable: |
|
108 | if not cmdtable: | |
105 | # Maybe we are processing mercurial.commands? |
|
109 | # Maybe we are processing mercurial.commands? | |
106 | cmdtable = getattr(mod, 'table', {}) |
|
110 | cmdtable = getattr(mod, 'table', {}) | |
107 | functions.extend((c[0], False) for c in cmdtable.itervalues()) |
|
111 | functions.extend((c[0], False) for c in cmdtable.itervalues()) | |
108 |
|
112 | |||
109 | for func, rstrip in functions: |
|
113 | for func, rstrip in functions: | |
110 | if func.__doc__: |
|
114 | if func.__doc__: | |
111 | src = inspect.getsource(func) |
|
115 | src = inspect.getsource(func) | |
112 | name = "%s.%s" % (path, func.__name__) |
|
116 | name = "%s.%s" % (path, func.__name__) | |
113 | lineno = func.func_code.co_firstlineno |
|
117 | lineno = func.func_code.co_firstlineno | |
114 | doc = func.__doc__ |
|
118 | doc = func.__doc__ | |
115 | if rstrip: |
|
119 | if rstrip: | |
116 | doc = doc.rstrip() |
|
120 | doc = doc.rstrip() | |
117 | lineno += offset(src, doc, name, 1) |
|
121 | lineno += offset(src, doc, name, 1) | |
118 | print poentry(path, lineno, doc) |
|
122 | print poentry(path, lineno, doc) | |
119 |
|
123 | |||
120 |
|
124 | |||
121 | def rawtext(path): |
|
125 | def rawtext(path): | |
122 | src = open(path).read() |
|
126 | src = open(path).read() | |
123 | print poentry(path, 1, src) |
|
127 | print poentry(path, 1, src) | |
124 |
|
128 | |||
125 |
|
129 | |||
126 | if __name__ == "__main__": |
|
130 | if __name__ == "__main__": | |
127 | # It is very important that we import the Mercurial modules from |
|
131 | # It is very important that we import the Mercurial modules from | |
128 | # the source tree where hggettext is executed. Otherwise we might |
|
132 | # the source tree where hggettext is executed. Otherwise we might | |
129 | # accidentally import and extract strings from a Mercurial |
|
133 | # accidentally import and extract strings from a Mercurial | |
130 | # installation mentioned in PYTHONPATH. |
|
134 | # installation mentioned in PYTHONPATH. | |
131 | sys.path.insert(0, os.getcwd()) |
|
135 | sys.path.insert(0, os.getcwd()) | |
132 | from mercurial import demandimport; demandimport.enable() |
|
136 | from mercurial import demandimport; demandimport.enable() | |
133 | for path in sys.argv[1:]: |
|
137 | for path in sys.argv[1:]: | |
134 | if path.endswith('.txt'): |
|
138 | if path.endswith('.txt'): | |
135 | rawtext(path) |
|
139 | rawtext(path) | |
136 | else: |
|
140 | else: | |
137 | docstrings(path) |
|
141 | docstrings(path) |
General Comments 0
You need to be logged in to leave comments.
Login now