2 # -*- coding: utf-8 -*-
3 # extract_texi_filenames.py
5 # USAGE: extract_texi_filenames.py [-o OUTDIR] FILES
7 # -o OUTDIR specifies that output files should rather be written in OUTDIR
10 # This script parses the .texi file given and creates a file with the
11 # nodename <=> filename/anchor map.
12 # The idea behind: Unnumbered subsections go into the same file as the
13 # previous numbered section, @translationof gives the original node name,
14 # which is then used for the filename/anchor.
16 # If this script is run on a file texifile.texi, it produces a file
17 # texifile[.LANG].xref-map with tab-separated entries of the form
18 # NODE\tFILENAME\tANCHOR
19 # LANG is the document language in case it's not 'en'
20 # Note: The filename does not have any extension appended!
21 # This file should then be used by our texi2html init script to determine
22 # the correct file name and anchor for external refs
24 # For translated documentation: cross-references to nodes that exist
25 # only in documentation in English are allowed, that's why the already
26 # generated map file of docs in English is loaded with
27 # --master-map-file option, then the node names that are defined in
28 # the map for the manual in English but not in the translated manual
29 # are added to the map for the translated manual.
37 options_list, files = getopt.getopt (sys.argv[1:],'o:s:hI:m:k:q',
41 'known-missing-files=',
44 help_text = r"""Usage: %(program_name)s [OPTIONS]... TEXIFILE...
45 Extract files names for texinfo (sub)sections from the texinfo files.
48 -h, --help print this help
49 -I, --include=DIRECTORY append DIRECTORY to include search path
50 -m, --master-map-file=FILE use FILE as master map file
51 -o, --output=DIRECTORY write .xref-map files to DIRECTORY
52 -s, --split=MODE split manual according to MODE. Possible values
53 are section and custom (default)
54 -k, --known-missing-files a filename which has a list of files known
55 to be missing for this make
56 -q, --quiet suppress most messages
60 sys.stdout.write ( text)
67 known_missing_files = []
68 known_missing_files_file = ''
69 suppress_output = False
71 for opt in options_list:
74 if o == '-h' or o == '--help':
75 help (help_text % vars ())
76 if o == '-I' or o == '--include':
78 include_path.append (a)
80 print 'NOT A DIR from: ', os.getcwd (), a
81 elif o == '-o' or o == '--output':
83 elif o == '-s' or o == '--split':
85 elif o == '-m' or o == '--master-map-file':
86 if os.path.isfile (a):
88 elif o == '--known-missing-files':
89 if os.path.isfile (a):
90 known_missing_files_file = a
92 print 'Missing files list file not found: ', a
93 elif o == '-q' or o == '--quiet':
94 suppress_output = True
96 raise Exception ('unknown option: ' + o)
98 if known_missing_files_file:
99 missing_files = open (known_missing_files_file, 'r')
100 known_missing_files = missing_files.read().splitlines()
101 missing_files.close()
103 if not os.path.isdir (outdir):
104 if os.path.exists (outdir):
108 include_re = re.compile (r'@include ((?!../lily-).*?\.i?te(xi|ly))$', re.M)
109 whitespaces = re.compile (r'\s+')
110 section_translation_re = re.compile ('^@(node|(?:unnumbered|appendix)\
111 (?:(?:sub){0,2}sec)?|top|chapter|(?:sub){0,2}section|\
112 (?:major|chap|(?:sub){0,2})heading|lydoctitle|translationof) \
113 (.+)$', re.MULTILINE)
114 external_node_re = re.compile (r'\s+@c\s+external.*')
116 def expand_includes (m, filename):
117 include_name = m.group (1)
118 filepath = os.path.join (os.path.dirname (filename), include_name)
119 if os.path.exists (filepath):
120 return extract_sections (filepath)[1]
122 for directory in include_path:
123 filepath = os.path.join (directory, include_name)
124 if os.path.exists (filepath):
125 return extract_sections (filepath)[1]
126 if not (include_name in known_missing_files):
128 print 'No such file: ' + include_name
129 print 'Search path: ' + ':'.join (include_path)
132 lang_re = re.compile (r'^@documentlanguage (.+)', re.M)
134 def extract_sections (filename):
136 f = open (filename, 'r')
139 # Search document language
140 m = lang_re.search (page)
141 if m and m.group (1) != 'en':
142 lang_suffix = '.' + m.group (1)
145 # Replace all includes by their list of sections and extract all sections
146 page = include_re.sub (lambda m: expand_includes (m, filename), page)
147 sections = section_translation_re.findall (page)
149 result += "@" + sec[0] + " " + sec[1] + "\n"
150 return (lang_suffix, result)
152 # Convert a given node name to its proper file name (normalization as
153 # explained in the texinfo manual:
154 # http://www.gnu.org/software/texinfo/manual/texinfo/html_node/HTML-Xref-Node-Name-Expansion.html
155 def texinfo_file_name(title):
156 # exception: The top node is always mapped to index.html
159 # File name normalization by texinfo (described in the texinfo manual):
160 # 1/2: letters and numbers are left unchanged
161 # 3/4: multiple, leading and trailing whitespace is removed
162 title = title.strip ();
163 title = whitespaces.sub (' ', title)
164 # 5: all remaining spaces are converted to '-'
165 # 6: all other 7- or 8-bit chars are replaced by _xxxx (xxxx=ascii character code)
167 for index in range(len(title)):
169 if char == ' ': # space -> '-'
171 elif ( ('0' <= char and char <= '9' ) or
172 ('A' <= char and char <= 'Z' ) or
173 ('a' <= char and char <= 'z' ) ): # number or letter
178 result += "_%04x" % ccode
180 result += "__%06x" % ccode
181 # 7: if name begins with number, prepend 't_g' (so it starts with a letter)
182 if (result != '') and (ord(result[0]) in range (ord('0'), ord('9'))):
183 result = 't_g' + result
186 texinfo_re = re.compile (r'@.*?{(.*?)}')
187 def remove_texinfo (title):
188 title = title.replace ('--', '-')
189 return texinfo_re.sub (r'\1', title).strip ()
191 def create_texinfo_anchor (title):
192 return texinfo_file_name (remove_texinfo (title))
194 unnumbered_re = re.compile (r'unnumbered.+|lydoctitle')
195 file_name_section_level = {
204 'unnumberedsubsec':1,
207 'unnumberedsubsubsec':0,
208 'appendixsubsubsec':0
210 if split in file_name_section_level:
211 splitting_level = file_name_section_level[split]
214 def process_sections (filename, lang_suffix, page):
215 sections = section_translation_re.findall (page)
216 basename = os.path.splitext (os.path.basename (filename))[0]
217 p = os.path.join (outdir, basename) + lang_suffix + '.xref-map'
218 if not suppress_output:
223 this_filename = 'index'
225 this_unnumbered = False
229 # Write out the cached values to the file and start a new
231 if this_title and this_title != 'Top':
232 f.write (this_title + "\t" + this_filename + "\t" + this_anchor + "\n")
234 this_title = remove_texinfo (sec[1])
235 this_anchor = create_texinfo_anchor (sec[1])
236 # delete entry from master map file
237 if this_title in initial_map:
238 del initial_map[this_title]
239 elif sec[0] == "translationof":
240 (original_node, external_node) = external_node_re.subn ('', sec[1])
241 original_node = remove_texinfo (original_node)
242 # The following binds the translator to use the
243 # translated node name in cross-references in case
245 if external_node and original_node in initial_map:
246 del initial_map[original_node]
247 anchor = create_texinfo_anchor (sec[1])
248 # If @translationof is used, it gives the original
249 # node name, which we use for the anchor and the file
250 # name (if it is a numbered node)
252 if not this_unnumbered:
253 this_filename = anchor
254 elif original_node in initial_map:
255 this_filename = initial_map[original_node][2]
257 # Some pages might not use a node for every section, so
258 # treat this case here, too: If we already had a section
259 # and encounter another one before the next @node, we
260 # write out the old one and start with the new values
261 if had_section and split != 'node' and this_title:
262 f.write (this_title + "\t" + this_filename + "\t" + this_anchor + "\n")
263 this_title = remove_texinfo (sec[1])
264 this_anchor = create_texinfo_anchor (sec[1])
267 if split == 'custom':
268 # unnumbered nodes use the previously used file name,
269 # only numbered nodes get their own filename! However,
270 # top-level @unnumbered still get their own file.
271 this_unnumbered = unnumbered_re.match (sec[0])
272 if not this_unnumbered:
273 this_filename = this_anchor
274 elif split == 'node':
275 this_filename = this_anchor
277 if sec[0] in file_name_section_level and \
278 file_name_section_level[sec[0]] >= splitting_level:
279 this_filename = this_anchor
281 if this_title and this_title != 'Top':
282 f.write (this_title + "\t" + this_filename + "\t" + this_anchor + "\n")
284 for node in initial_map:
285 f.write ("\t".join (initial_map[node]) + "\n")
288 xref_map_line_re = re.compile (r'(.*?)\t(.*?)\t(.*?)$')
290 for line in open (master_map_file):
291 m = xref_map_line_re.match (line)
293 initial_map[m.group (1)] = (m.group (1), m.group (2), m.group (3))
295 for filename in files:
296 if not suppress_output:
297 print "extract_texi_filenames.py: Processing %s" % filename
298 (lang_suffix, sections) = extract_sections (filename)
299 process_sections (filename, lang_suffix, sections)