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:',
42 help_text = r"""Usage: %(program_name)s [OPTIONS]... TEXIFILE...
43 Extract files names for texinfo (sub)sections from the texinfo files.
46 -h, --help print this help
47 -I, --include=DIRECTORY append DIRECTORY to include search path
48 -m, --master-map-file=FILE use FILE as master map file
49 -o, --output=DIRECTORY write .xref-map files to DIRECTORY
50 -s, --split=MODE split manual according to MODE. Possible values
51 are section and custom (default)
55 sys.stdout.write ( text)
63 for opt in options_list:
66 if o == '-h' or o == '--help':
67 help (help_text % vars ())
68 if o == '-I' or o == '--include':
70 include_path.append (a)
72 print 'NOT A DIR from: ', os.getcwd (), a
73 elif o == '-o' or o == '--output':
75 elif o == '-s' or o == '--split':
77 elif o == '-m' or o == '--master-map-file':
78 if os.path.isfile (a):
81 raise Exception ('unknown option: ' + o)
84 if not os.path.isdir (outdir):
85 if os.path.exists (outdir):
89 include_re = re.compile (r'@include ((?!../lily-).*?\.i?texi)$', re.M)
90 whitespaces = re.compile (r'\s+')
91 section_translation_re = re.compile ('^@(node|(?:unnumbered|appendix)\
92 (?:(?:sub){0,2}sec)?|top|chapter|(?:sub){0,2}section|\
93 (?:major|chap|(?:sub){0,2})heading|lydoctitle|translationof) \
95 external_node_re = re.compile (r'\s+@c\s+external.*')
97 def expand_includes (m, filename):
98 include_name = m.group (1)
99 filepath = os.path.join (os.path.dirname (filename), include_name)
100 if os.path.exists (filepath):
101 return extract_sections (filepath)[1]
103 for directory in include_path:
104 filepath = os.path.join (directory, include_name)
105 if os.path.exists (filepath):
106 return extract_sections (filepath)[1]
107 print 'No such file: ' + include_name
108 print 'Search path: ' + ':'.join (include_path)
111 lang_re = re.compile (r'^@documentlanguage (.+)', re.M)
113 def extract_sections (filename):
115 f = open (filename, 'r')
118 # Search document language
119 m = lang_re.search (page)
120 if m and m.group (1) != 'en':
121 lang_suffix = '.' + m.group (1)
124 # Replace all includes by their list of sections and extract all sections
125 page = include_re.sub (lambda m: expand_includes (m, filename), page)
126 sections = section_translation_re.findall (page)
128 result += "@" + sec[0] + " " + sec[1] + "\n"
129 return (lang_suffix, result)
131 # Convert a given node name to its proper file name (normalization as
132 # explained in the texinfo manual:
133 # http://www.gnu.org/software/texinfo/manual/texinfo/html_node/HTML-Xref-Node-Name-Expansion.html
134 def texinfo_file_name(title):
135 # exception: The top node is always mapped to index.html
138 # File name normalization by texinfo (described in the texinfo manual):
139 # 1/2: letters and numbers are left unchanged
140 # 3/4: multiple, leading and trailing whitespace is removed
141 title = title.strip ();
142 title = whitespaces.sub (' ', title)
143 # 5: all remaining spaces are converted to '-'
144 # 6: all other 7- or 8-bit chars are replaced by _xxxx (xxxx=ascii character code)
146 for index in range(len(title)):
148 if char == ' ': # space -> '-'
150 elif ( ('0' <= char and char <= '9' ) or
151 ('A' <= char and char <= 'Z' ) or
152 ('a' <= char and char <= 'z' ) ): # number or letter
157 result += "_%04x" % ccode
159 result += "__%06x" % ccode
160 # 7: if name begins with number, prepend 't_g' (so it starts with a letter)
161 if (result != '') and (ord(result[0]) in range (ord('0'), ord('9'))):
162 result = 't_g' + result
165 texinfo_re = re.compile (r'@.*?{(.*?)}')
166 def remove_texinfo (title):
167 title = title.replace ('--', '-')
168 return texinfo_re.sub (r'\1', title).strip ()
170 def create_texinfo_anchor (title):
171 return texinfo_file_name (remove_texinfo (title))
173 unnumbered_re = re.compile (r'unnumbered.+|lydoctitle')
174 file_name_section_level = {
183 'unnumberedsubsec':1,
186 'unnumberedsubsubsec':0,
187 'appendixsubsubsec':0
189 if split in file_name_section_level:
190 splitting_level = file_name_section_level[split]
193 def process_sections (filename, lang_suffix, page):
194 sections = section_translation_re.findall (page)
195 basename = os.path.splitext (os.path.basename (filename))[0]
196 p = os.path.join (outdir, basename) + lang_suffix + '.xref-map'
201 this_filename = 'index'
203 this_unnumbered = False
207 # Write out the cached values to the file and start a new
209 if this_title and this_title != 'Top':
210 f.write (this_title + "\t" + this_filename + "\t" + this_anchor + "\n")
212 this_title = remove_texinfo (sec[1])
213 this_anchor = create_texinfo_anchor (sec[1])
214 # delete entry from master map file
215 if this_title in initial_map:
216 del initial_map[this_title]
217 elif sec[0] == "translationof":
218 (original_node, external_node) = external_node_re.subn ('', sec[1])
219 original_node = remove_texinfo (original_node)
220 # The following binds the translator to use the
221 # translated node name in cross-references in case
223 if external_node and original_node in initial_map:
224 del initial_map[original_node]
225 anchor = create_texinfo_anchor (sec[1])
226 # If @translationof is used, it gives the original
227 # node name, which we use for the anchor and the file
228 # name (if it is a numbered node)
230 if not this_unnumbered:
231 this_filename = anchor
232 elif original_node in initial_map:
233 this_filename = initial_map[original_node][2]
235 # Some pages might not use a node for every section, so
236 # treat this case here, too: If we already had a section
237 # and encounter another one before the next @node, we
238 # write out the old one and start with the new values
239 if had_section and this_title:
240 f.write (this_title + "\t" + this_filename + "\t" + this_anchor + "\n")
241 this_title = remove_texinfo (sec[1])
242 this_anchor = create_texinfo_anchor (sec[1])
245 if split == 'custom':
246 # unnumbered nodes use the previously used file name,
247 # only numbered nodes get their own filename! However,
248 # top-level @unnumbered still get their own file.
249 this_unnumbered = unnumbered_re.match (sec[0])
250 if not this_unnumbered:
251 this_filename = this_anchor
252 elif split == 'node':
253 this_filename = this_anchor
255 if sec[0] in file_name_section_level and \
256 file_name_section_level[sec[0]] >= splitting_level:
257 this_filename = this_anchor
259 if this_title and this_title != 'Top':
260 f.write (this_title + "\t" + this_filename + "\t" + this_anchor + "\n")
262 for node in initial_map:
263 f.write ("\t".join (initial_map[node]) + "\n")
266 xref_map_line_re = re.compile (r'(.*?)\t(.*?)\t(.*?)$')
268 for line in open (master_map_file):
269 m = xref_map_line_re.match (line)
271 initial_map[m.group (1)] = (m.group (1), m.group (2), m.group (3))
273 for filename in files:
274 print "extract_texi_filenames.py: Processing %s" % filename
275 (lang_suffix, sections) = extract_sections (filename)
276 process_sections (filename, lang_suffix, sections)