Source

sphinx / sphinx / transforms.py

  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
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
# -*- coding: utf-8 -*-
"""
    sphinx.transforms
    ~~~~~~~~~~~~~~~~~

    Docutils transforms used by Sphinx when reading documents.

    :copyright: Copyright 2007-2013 by the Sphinx team, see AUTHORS.
    :license: BSD, see LICENSE for details.
"""

from os import path

from docutils import nodes
from docutils.utils import new_document, relative_path
from docutils.parsers.rst import Parser as RSTParser
from docutils.transforms import Transform
from docutils.transforms.parts import ContentsFilter

from sphinx import addnodes
from sphinx.locale import _, init as init_locale
from sphinx.util import split_index_msg
from sphinx.util.nodes import traverse_translatable_index, extract_messages
from sphinx.util.osutil import ustrftime, find_catalog
from sphinx.util.compat import docutils_version
from sphinx.util.pycompat import all
from sphinx.domains.std import (
    make_term_from_paragraph_node,
    make_termnodes_from_paragraph_node,
)


default_substitutions = set([
    'version',
    'release',
    'today',
])

class DefaultSubstitutions(Transform):
    """
    Replace some substitutions if they aren't defined in the document.
    """
    # run before the default Substitutions
    default_priority = 210

    def apply(self):
        config = self.document.settings.env.config
        # only handle those not otherwise defined in the document
        to_handle = default_substitutions - set(self.document.substitution_defs)
        for ref in self.document.traverse(nodes.substitution_reference):
            refname = ref['refname']
            if refname in to_handle:
                text = config[refname]
                if refname == 'today' and not text:
                    # special handling: can also specify a strftime format
                    text = ustrftime(config.today_fmt or _('%B %d, %Y'))
                ref.replace_self(nodes.Text(text, text))


class MoveModuleTargets(Transform):
    """
    Move module targets that are the first thing in a section to the section
    title.

    XXX Python specific
    """
    default_priority = 210

    def apply(self):
        for node in self.document.traverse(nodes.target):
            if not node['ids']:
                continue
            if (node.has_key('ismod') and
                node.parent.__class__ is nodes.section and
                # index 0 is the section title node
                node.parent.index(node) == 1):
                node.parent['ids'][0:0] = node['ids']
                node.parent.remove(node)


class HandleCodeBlocks(Transform):
    """
    Several code block related transformations.
    """
    default_priority = 210

    def apply(self):
        # move doctest blocks out of blockquotes
        for node in self.document.traverse(nodes.block_quote):
            if all(isinstance(child, nodes.doctest_block) for child
                     in node.children):
                node.replace_self(node.children)
        # combine successive doctest blocks
        #for node in self.document.traverse(nodes.doctest_block):
        #    if node not in node.parent.children:
        #        continue
        #    parindex = node.parent.index(node)
        #    while len(node.parent) > parindex+1 and \
        #            isinstance(node.parent[parindex+1], nodes.doctest_block):
        #        node[0] = nodes.Text(node[0] + '\n\n' +
        #                             node.parent[parindex+1][0])
        #        del node.parent[parindex+1]


class SortIds(Transform):
    """
    Sort secion IDs so that the "id[0-9]+" one comes last.
    """
    default_priority = 261

    def apply(self):
        for node in self.document.traverse(nodes.section):
            if len(node['ids']) > 1 and node['ids'][0].startswith('id'):
                node['ids'] = node['ids'][1:] + [node['ids'][0]]


class CitationReferences(Transform):
    """
    Replace citation references by pending_xref nodes before the default
    docutils transform tries to resolve them.
    """
    default_priority = 619

    def apply(self):
        for citnode in self.document.traverse(nodes.citation_reference):
            cittext = citnode.astext()
            refnode = addnodes.pending_xref(cittext, reftype='citation',
                                            reftarget=cittext, refwarn=True,
                                            ids=citnode["ids"])
            refnode.line = citnode.line or citnode.parent.line
            refnode += nodes.Text('[' + cittext + ']')
            citnode.parent.replace(citnode, refnode)


class CustomLocaleReporter(object):
    """
    Replacer for document.reporter.get_source_and_line method.

    reST text lines for translation do not have the original source line number.
    This class provides the correct line numbers when reporting.
    """
    def __init__(self, source, line):
        self.source, self.line = source, line

    def set_reporter(self, document):
        if docutils_version < (0, 9):
            document.reporter.locator = self.get_source_and_line
        else:
            document.reporter.get_source_and_line = self.get_source_and_line

    def get_source_and_line(self, lineno=None):
        return self.source, self.line


class Locale(Transform):
    """
    Replace translatable nodes with their translated doctree.
    """
    default_priority = 0

    def apply(self):
        env = self.document.settings.env
        settings, source = self.document.settings, self.document['source']
        # XXX check if this is reliable
        assert source.startswith(env.srcdir)
        docname = path.splitext(relative_path(env.srcdir, source))[0]
        textdomain = find_catalog(docname,
                                  self.document.settings.gettext_compact)

        # fetch translations
        dirs = [path.join(env.srcdir, directory)
                for directory in env.config.locale_dirs]
        catalog, has_catalog = init_locale(dirs, env.config.language,
                                           textdomain)
        if not has_catalog:
            return

        parser = RSTParser()

        #phase1: replace reference ids with translated names
        for node, msg in extract_messages(self.document):
            msgstr = catalog.gettext(msg)
            # XXX add marker to untranslated parts
            if not msgstr or msgstr == msg: # as-of-yet untranslated
                continue

            # Avoid "Literal block expected; none found." warnings.
            # If msgstr ends with '::' then it cause warning message at
            # parser.parse() processing.
            # literal-block-warning is only appear in avobe case.
            if msgstr.strip().endswith('::'):
                msgstr += '\n\n   dummy literal'
                # dummy literal node will discard by 'patch = patch[0]'

            patch = new_document(source, settings)
            CustomLocaleReporter(node.source, node.line).set_reporter(patch)
            parser.parse(msgstr, patch)
            patch = patch[0]
            # XXX doctest and other block markup
            if not isinstance(patch, nodes.paragraph):
                continue # skip for now

            processed = False  # skip flag

            # update title(section) target name-id mapping
            if isinstance(node, nodes.title):
                section_node = node.parent
                new_name = nodes.fully_normalize_name(patch.astext())
                old_name = nodes.fully_normalize_name(node.astext())

                if old_name != new_name:
                    # if name would be changed, replace node names and
                    # document nameids mapping with new name.
                    names = section_node.setdefault('names', [])
                    names.append(new_name)
                    if old_name in names:
                        names.remove(old_name)

                    _id = self.document.nameids.get(old_name, None)
                    explicit = self.document.nametypes.get(old_name, None)

                    # * if explicit: _id is label. title node need another id.
                    # * if not explicit:
                    #
                    #   * _id is None:
                    #
                    #     _id is None means _id was duplicated.
                    #     old_name entry still exists in nameids and
                    #     nametypes for another duplicated entry.
                    #
                    #   * _id is provided: bellow process
                    if not explicit and _id:
                        # _id was not duplicated.
                        # remove old_name entry from document ids database
                        # to reuse original _id.
                        self.document.nameids.pop(old_name, None)
                        self.document.nametypes.pop(old_name, None)
                        self.document.ids.pop(_id, None)

                    # re-entry with new named section node.
                    self.document.note_implicit_target(
                            section_node, section_node)

                    processed = True

            # glossary terms update refid
            if isinstance(node, nodes.term):
                gloss_entries = env.temp_data.setdefault('gloss_entries', set())
                ids = []
                termnodes = []
                for _id in node['names']:
                    if _id in gloss_entries:
                        gloss_entries.remove(_id)
                    _id, _, new_termnodes = \
                        make_termnodes_from_paragraph_node(env, patch, _id)
                    ids.append(_id)
                    termnodes.extend(new_termnodes)

                if termnodes and ids:
                    patch = make_term_from_paragraph_node(termnodes, ids)
                    node['ids'] = patch['ids']
                    node['names'] = patch['names']
                    processed = True

            # update leaves with processed nodes
            if processed:
                for child in patch.children:
                    child.parent = node
                node.children = patch.children
                node['translated'] = True


        #phase2: translation
        for node, msg in extract_messages(self.document):
            if node.get('translated', False):
                continue

            msgstr = catalog.gettext(msg)
            # XXX add marker to untranslated parts
            if not msgstr or msgstr == msg: # as-of-yet untranslated
                continue

            # Avoid "Literal block expected; none found." warnings.
            # If msgstr ends with '::' then it cause warning message at
            # parser.parse() processing.
            # literal-block-warning is only appear in avobe case.
            if msgstr.strip().endswith('::'):
                msgstr += '\n\n   dummy literal'
                # dummy literal node will discard by 'patch = patch[0]'

            patch = new_document(source, settings)
            CustomLocaleReporter(node.source, node.line).set_reporter(patch)
            parser.parse(msgstr, patch)
            patch = patch[0]
            # XXX doctest and other block markup
            if not isinstance(patch, nodes.paragraph):
                continue # skip for now

            # auto-numbered foot note reference should use original 'ids'.
            def is_autonumber_footnote_ref(node):
                return isinstance(node, nodes.footnote_reference) and \
                    node.get('auto') == 1
            def list_replace_or_append(lst, old, new):
                if old in lst:
                    lst[lst.index(old)] = new
                else:
                    lst.append(new)
            old_foot_refs = node.traverse(is_autonumber_footnote_ref)
            new_foot_refs = patch.traverse(is_autonumber_footnote_ref)
            if len(old_foot_refs) != len(new_foot_refs):
                env.warn_node('inconsistent footnote references in '
                              'translated message', node)
            old_foot_namerefs = {}
            for r in old_foot_refs:
                old_foot_namerefs.setdefault(r.get('refname'), []).append(r)
            for new in new_foot_refs:
                refname = new.get('refname')
                refs = old_foot_namerefs.get(refname, [])
                if not refs:
                    continue

                old = refs.pop(0)
                new['ids'] = old['ids']
                for id in new['ids']:
                    self.document.ids[id] = new
                list_replace_or_append(
                        self.document.autofootnote_refs, old, new)
                if refname:
                    list_replace_or_append(
                        self.document.footnote_refs.setdefault(refname, []),
                        old, new)
                    list_replace_or_append(
                        self.document.refnames.setdefault(refname, []),
                        old, new)

            # reference should use new (translated) 'refname'.
            # * reference target ".. _Python: ..." is not translatable.
            # * use translated refname for section refname.
            # * inline reference "`Python <...>`_" has no 'refname'.
            def is_refnamed_ref(node):
                return isinstance(node, nodes.reference) and  \
                    'refname' in node
            old_refs = node.traverse(is_refnamed_ref)
            new_refs = patch.traverse(is_refnamed_ref)
            if len(old_refs) != len(new_refs):
                env.warn_node('inconsistent references in '
                              'translated message', node)
            old_ref_names = [r['refname'] for r in old_refs]
            new_ref_names = [r['refname'] for r in new_refs]
            orphans = list(set(old_ref_names) - set(new_ref_names))
            for new in new_refs:
                if not self.document.has_name(new['refname']):
                    # Maybe refname is translated but target is not translated.
                    # Note: multiple translated refnames break link ordering.
                    if orphans:
                        new['refname'] = orphans.pop(0)
                    else:
                        # orphan refnames is already empty!
                        # reference number is same in new_refs and old_refs.
                        pass

                self.document.note_refname(new)

            # refnamed footnote and citation should use original 'ids'.
            def is_refnamed_footnote_ref(node):
                footnote_ref_classes = (nodes.footnote_reference,
                                        nodes.citation_reference)
                return isinstance(node, footnote_ref_classes) and \
                    'refname' in node
            old_refs = node.traverse(is_refnamed_footnote_ref)
            new_refs = patch.traverse(is_refnamed_footnote_ref)
            refname_ids_map = {}
            if len(old_refs) != len(new_refs):
                env.warn_node('inconsistent references in '
                              'translated message', node)
            for old in old_refs:
                refname_ids_map[old["refname"]] = old["ids"]
            for new in new_refs:
                refname = new["refname"]
                if refname in refname_ids_map:
                    new["ids"] = refname_ids_map[refname]

            # Original pending_xref['reftarget'] contain not-translated
            # target name, new pending_xref must use original one.
            # This code restricts to change ref-targets in the translation.
            old_refs = node.traverse(addnodes.pending_xref)
            new_refs = patch.traverse(addnodes.pending_xref)
            xref_reftarget_map = {}
            if len(old_refs) != len(new_refs):
                env.warn_node('inconsistent term references in '
                              'translated message', node)
            def get_ref_key(node):
                case = node["refdomain"], node["reftype"]
                if case == ('std', 'term'):
                    return None
                else:
                    return (
                        node["refdomain"],
                        node["reftype"],
                        node['reftarget'],)

            for old in old_refs:
                key = get_ref_key(old)
                if key:
                    xref_reftarget_map[key] = old["reftarget"]
            for new in new_refs:
                key = get_ref_key(new)
                if key in xref_reftarget_map:
                    new['reftarget'] = xref_reftarget_map[key]

            # update leaves
            for child in patch.children:
                child.parent = node
            node.children = patch.children
            node['translated'] = True

        # Extract and translate messages for index entries.
        for node, entries in traverse_translatable_index(self.document):
            new_entries = []
            for type, msg, tid, main in entries:
                msg_parts = split_index_msg(type, msg)
                msgstr_parts = []
                for part in msg_parts:
                    msgstr = catalog.gettext(part)
                    if not msgstr:
                        msgstr = part
                    msgstr_parts.append(msgstr)

                new_entries.append((type, ';'.join(msgstr_parts), tid, main))

            node['raw_entries'] = entries
            node['entries'] = new_entries


class RemoveTranslatableInline(Transform):
    """
    Remove inline nodes used for translation as placeholders.
    """
    default_priority = 999

    def apply(self):
        from sphinx.builders.gettext import MessageCatalogBuilder
        env = self.document.settings.env
        builder = env.app.builder
        if isinstance(builder, MessageCatalogBuilder):
            return
        for inline in self.document.traverse(nodes.inline):
            if 'translatable' in inline:
                inline.parent.remove(inline)
                inline.parent += inline.children


class SphinxContentsFilter(ContentsFilter):
    """
    Used with BuildEnvironment.add_toc_from() to discard cross-file links
    within table-of-contents link nodes.
    """
    def visit_pending_xref(self, node):
        text = node.astext()
        self.parent.append(nodes.literal(text, text))
        raise nodes.SkipNode

    def visit_image(self, node):
        raise nodes.SkipNode