1# Copyright (C) 2011 Canonical Ltd 2# 3# This program is free software; you can redistribute it and/or modify 4# it under the terms of the GNU General Public License as published by 5# the Free Software Foundation; either version 2 of the License, or 6# (at your option) any later version. 7# 8# This program is distributed in the hope that it will be useful, 9# but WITHOUT ANY WARRANTY; without even the implied warranty of 10# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 11# GNU General Public License for more details. 12# 13# You should have received a copy of the GNU General Public License 14# along with this program; if not, write to the Free Software 15# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA 16 17# The normalize function is taken from pygettext which is distributed 18# with Python under the Python License, which is GPL compatible. 19 20"""Extract docstrings from Bazaar commands. 21 22This module only handles breezy objects that use strings not directly wrapped 23by a gettext() call. To generate a complete translation template file, this 24output needs to be combined with that of xgettext or a similar command for 25extracting those strings, as is done in the bzr Makefile. Sorting the output 26is also left to that stage of the process. 27""" 28 29import inspect 30import os 31import sys 32 33import breezy 34from . import ( 35 commands as _mod_commands, 36 errors, 37 help_topics, 38 option, 39 plugin as _mod_plugin, 40 ) 41from .trace import ( 42 mutter, 43 note, 44 ) 45from .i18n import gettext 46 47 48def _escape(s): 49 s = (s.replace('\\', '\\\\') 50 .replace('\n', '\\n') 51 .replace('\r', '\\r') 52 .replace('\t', '\\t') 53 .replace('"', '\\"') 54 ) 55 return s 56 57 58def _normalize(s): 59 # This converts the various Python string types into a format that 60 # is appropriate for .po files, namely much closer to C style. 61 lines = s.split('\n') 62 if len(lines) == 1: 63 s = '"' + _escape(s) + '"' 64 else: 65 if not lines[-1]: 66 del lines[-1] 67 lines[-1] = lines[-1] + '\n' 68 lineterm = '\\n"\n"' 69 s = '""\n"' + lineterm.join(map(_escape, lines)) + '"' 70 return s 71 72 73def _parse_source(source_text, filename='<unknown>'): 74 """Get object to lineno mappings from given source_text""" 75 import ast 76 cls_to_lineno = {} 77 str_to_lineno = {} 78 for node in ast.walk(ast.parse(source_text, filename)): 79 # TODO: worry about duplicates? 80 if isinstance(node, ast.ClassDef): 81 # TODO: worry about nesting? 82 cls_to_lineno[node.name] = node.lineno 83 elif isinstance(node, ast.Str): 84 # Python AST gives location of string literal as the line the 85 # string terminates on. It's more useful to have the line the 86 # string begins on. Unfortunately, counting back newlines is 87 # only an approximation as the AST is ignorant of escaping. 88 str_to_lineno[node.s] = node.lineno - (0 if sys.version_info >= (3, 8) else node.s.count('\n')) 89 return cls_to_lineno, str_to_lineno 90 91 92class _ModuleContext(object): 93 """Record of the location within a source tree""" 94 95 def __init__(self, path, lineno=1, _source_info=None): 96 self.path = path 97 self.lineno = lineno 98 if _source_info is not None: 99 self._cls_to_lineno, self._str_to_lineno = _source_info 100 101 @classmethod 102 def from_module(cls, module): 103 """Get new context from module object and parse source for linenos""" 104 sourcepath = inspect.getsourcefile(module) 105 # TODO: fix this to do the right thing rather than rely on cwd 106 relpath = os.path.relpath(sourcepath) 107 return cls(relpath, 108 _source_info=_parse_source("".join(inspect.findsource(module)[0]), module.__file__)) 109 110 def from_class(self, cls): 111 """Get new context with same details but lineno of class in source""" 112 try: 113 lineno = self._cls_to_lineno[cls.__name__] 114 except (AttributeError, KeyError): 115 mutter("Definition of %r not found in %r", cls, self.path) 116 return self 117 return self.__class__(self.path, lineno, 118 (self._cls_to_lineno, self._str_to_lineno)) 119 120 def from_string(self, string): 121 """Get new context with same details but lineno of string in source""" 122 try: 123 lineno = self._str_to_lineno[string] 124 except (AttributeError, KeyError): 125 mutter("String %r not found in %r", string[:20], self.path) 126 return self 127 return self.__class__(self.path, lineno, 128 (self._cls_to_lineno, self._str_to_lineno)) 129 130 131class _PotExporter(object): 132 """Write message details to output stream in .pot file format""" 133 134 def __init__(self, outf, include_duplicates=False): 135 self.outf = outf 136 if include_duplicates: 137 self._msgids = None 138 else: 139 self._msgids = set() 140 self._module_contexts = {} 141 142 def poentry(self, path, lineno, s, comment=None): 143 if self._msgids is not None: 144 if s in self._msgids: 145 return 146 self._msgids.add(s) 147 if comment is None: 148 comment = '' 149 else: 150 comment = "# %s\n" % comment 151 mutter("Exporting msg %r at line %d in %r", s[:20], lineno, path) 152 line = ( 153 "#: {path}:{lineno}\n" 154 "{comment}" 155 "msgid {msg}\n" 156 "msgstr \"\"\n" 157 "\n".format( 158 path=path, lineno=lineno, comment=comment, msg=_normalize(s))) 159 self.outf.write(line) 160 161 def poentry_in_context(self, context, string, comment=None): 162 context = context.from_string(string) 163 self.poentry(context.path, context.lineno, string, comment) 164 165 def poentry_per_paragraph(self, path, lineno, msgid, include=None): 166 # TODO: How to split long help? 167 paragraphs = msgid.split('\n\n') 168 if include is not None: 169 paragraphs = filter(include, paragraphs) 170 for p in paragraphs: 171 self.poentry(path, lineno, p) 172 lineno += p.count('\n') + 2 173 174 def get_context(self, obj): 175 module = inspect.getmodule(obj) 176 try: 177 context = self._module_contexts[module.__name__] 178 except KeyError: 179 context = _ModuleContext.from_module(module) 180 self._module_contexts[module.__name__] = context 181 if inspect.isclass(obj): 182 context = context.from_class(obj) 183 return context 184 185 186def _write_option(exporter, context, opt, note): 187 if getattr(opt, 'hidden', False): 188 return 189 optname = opt.name 190 if getattr(opt, 'title', None): 191 exporter.poentry_in_context(context, opt.title, 192 "title of {name!r} {what}".format(name=optname, what=note)) 193 for name, _, _, helptxt in opt.iter_switches(): 194 if name != optname: 195 if opt.is_hidden(name): 196 continue 197 name = "=".join([optname, name]) 198 if helptxt: 199 exporter.poentry_in_context(context, helptxt, 200 "help of {name!r} {what}".format(name=name, what=note)) 201 202 203def _standard_options(exporter): 204 OPTIONS = option.Option.OPTIONS 205 context = exporter.get_context(option) 206 for name in sorted(OPTIONS): 207 opt = OPTIONS[name] 208 _write_option(exporter, context.from_string(name), opt, "option") 209 210 211def _command_options(exporter, context, cmd): 212 note = "option of {0!r} command".format(cmd.name()) 213 for opt in cmd.takes_options: 214 # String values in Command option lists are for global options 215 if not isinstance(opt, str): 216 _write_option(exporter, context, opt, note) 217 218 219def _write_command_help(exporter, cmd): 220 context = exporter.get_context(cmd.__class__) 221 rawdoc = cmd.__doc__ 222 dcontext = context.from_string(rawdoc) 223 doc = inspect.cleandoc(rawdoc) 224 225 def exclude_usage(p): 226 # ':Usage:' has special meaning in help topics. 227 # This is usage example of command and should not be translated. 228 if p.splitlines()[0] != ':Usage:': 229 return True 230 231 exporter.poentry_per_paragraph(dcontext.path, dcontext.lineno, doc, 232 exclude_usage) 233 _command_options(exporter, context, cmd) 234 235 236def _command_helps(exporter, plugin_name=None): 237 """Extract docstrings from path. 238 239 This respects the Bazaar cmdtable/table convention and will 240 only extract docstrings from functions mentioned in these tables. 241 """ 242 243 # builtin commands 244 for cmd_name in _mod_commands.builtin_command_names(): 245 command = _mod_commands.get_cmd_object(cmd_name, False) 246 if command.hidden: 247 continue 248 if plugin_name is not None: 249 # only export builtins if we are not exporting plugin commands 250 continue 251 note(gettext("Exporting messages from builtin command: %s"), cmd_name) 252 _write_command_help(exporter, command) 253 254 plugins = _mod_plugin.plugins() 255 if plugin_name is not None and plugin_name not in plugins: 256 raise errors.BzrError(gettext('Plugin %s is not loaded' % plugin_name)) 257 core_plugins = set( 258 name for name in plugins 259 if plugins[name].path().startswith(breezy.__path__[0])) 260 # plugins 261 for cmd_name in _mod_commands.plugin_command_names(): 262 command = _mod_commands.get_cmd_object(cmd_name, False) 263 if command.hidden: 264 continue 265 if plugin_name is not None and command.plugin_name() != plugin_name: 266 # if we are exporting plugin commands, skip plugins we have not 267 # specified. 268 continue 269 if plugin_name is None and command.plugin_name() not in core_plugins: 270 # skip non-core plugins 271 # TODO: Support extracting from third party plugins. 272 continue 273 note(gettext("Exporting messages from plugin command: {0} in {1}").format( 274 cmd_name, command.plugin_name())) 275 _write_command_help(exporter, command) 276 277 278def _error_messages(exporter): 279 """Extract fmt string from breezy.errors.""" 280 context = exporter.get_context(errors) 281 base_klass = errors.BzrError 282 for name in dir(errors): 283 klass = getattr(errors, name) 284 if not inspect.isclass(klass): 285 continue 286 if not issubclass(klass, base_klass): 287 continue 288 if klass is base_klass: 289 continue 290 if klass.internal_error: 291 continue 292 fmt = getattr(klass, "_fmt", None) 293 if fmt: 294 note(gettext("Exporting message from error: %s"), name) 295 exporter.poentry_in_context(context, fmt) 296 297 298def _help_topics(exporter): 299 topic_registry = help_topics.topic_registry 300 for key in topic_registry.keys(): 301 doc = topic_registry.get(key) 302 if isinstance(doc, str): 303 exporter.poentry_per_paragraph( 304 'dummy/help_topics/' + key + '/detail.txt', 305 1, doc) 306 elif callable(doc): # help topics from files 307 exporter.poentry_per_paragraph( 308 'en/help_topics/' + key + '.txt', 309 1, doc(key)) 310 summary = topic_registry.get_summary(key) 311 if summary is not None: 312 exporter.poentry('dummy/help_topics/' + key + '/summary.txt', 313 1, summary) 314 315 316def export_pot(outf, plugin=None, include_duplicates=False): 317 exporter = _PotExporter(outf, include_duplicates) 318 if plugin is None: 319 _standard_options(exporter) 320 _command_helps(exporter) 321 _error_messages(exporter) 322 _help_topics(exporter) 323 else: 324 _command_helps(exporter, plugin) 325