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 can then be used by our texi2html init script to determine
22 # the correct file name and anchor for external refs
29 options_list, files = getopt.getopt (sys.argv[1:],'o:s:hI:m:',
34 help_text = r"""Usage: %(program_name)s [OPTIONS]... TEXIFILE...
35 Extract files names for texinfo (sub)sections from the texinfo files.
38 -h, --help print this help
39 -I, --include=DIRECTORY append DIRECTORY to include search path
40 -m, --master-map-file=FILE use FILE as master map file
41 -o, --output=DIRECTORY write .xref-map files to DIRECTORY
42 -s, --split=MODE split manual according to MODE. Possible values
43 are section and custom (default)
47 sys.stdout.write ( text)
55 for opt in options_list:
58 if o == '-h' or o == '--help':
59 help (help_text % vars ())
60 if o == '-I' or o == '--include':
62 include_path.append (a)
64 print 'NOT A DIR from: ', os.getcwd (), a
65 elif o == '-o' or o == '--output':
67 elif o == '-s' or o == '--split':
69 elif o == '-m' or o == '--master-map-file':
70 if os.path.isfile (a):
73 raise Exception ('unknown option: ' + o)
76 if not os.path.isdir (outdir):
77 if os.path.exists (outdir):
81 include_re = re.compile (r'@include ((?!../lily-).*?\.i?texi)$', re.M)
82 whitespaces = re.compile (r'\s+')
83 section_translation_re = re.compile ('^@(node|(?:unnumbered|appendix)\
84 (?:(?:sub){0,2}sec)?|top|chapter|(?:sub){0,2}section|\
85 (?:major|chap|(?:sub){0,2})heading|lydoctitle|translationof) \
87 external_node_re = re.compile (r'\s+@c\s+external.*')
89 def expand_includes (m, filename):
90 filepath = os.path.join (os.path.dirname (filename), m.group(1))
91 if os.path.exists (filepath):
92 return extract_sections (filepath)[1]
94 for directory in include_path:
95 filepath = os.path.join (directory, m.group(1))
96 if os.path.exists (filepath):
97 return extract_sections (filepath)[1]
98 print 'No such file: ' + filepath
101 lang_re = re.compile (r'^@documentlanguage (.+)', re.M)
103 def extract_sections (filename):
105 f = open (filename, 'r')
108 # Search document language
109 m = lang_re.search (page)
110 if m and m.group (1) != 'en':
111 lang_suffix = '.' + m.group (1)
114 # Replace all includes by their list of sections and extract all sections
115 page = include_re.sub (lambda m: expand_includes (m, filename), page)
116 sections = section_translation_re.findall (page)
118 result += "@" + sec[0] + " " + sec[1] + "\n"
119 return (lang_suffix, result)
121 # Convert a given node name to its proper file name (normalization as
122 # explained in the texinfo manual:
123 # http://www.gnu.org/software/texinfo/manual/texinfo/html_node/HTML-Xref-Node-Name-Expansion.html
124 def texinfo_file_name(title):
125 # exception: The top node is always mapped to index.html
128 # File name normalization by texinfo (described in the texinfo manual):
129 # 1/2: letters and numbers are left unchanged
130 # 3/4: multiple, leading and trailing whitespace is removed
131 title = title.strip ();
132 title = whitespaces.sub (' ', title)
133 # 5: all remaining spaces are converted to '-'
134 # 6: all other 7- or 8-bit chars are replaced by _xxxx (xxxx=ascii character code)
136 for index in range(len(title)):
138 if char == ' ': # space -> '-'
140 elif ( ('0' <= char and char <= '9' ) or
141 ('A' <= char and char <= 'Z' ) or
142 ('a' <= char and char <= 'z' ) ): # number or letter
147 result += "_%04x" % ccode
149 result += "__%06x" % ccode
150 # 7: if name begins with number, prepend 't_g' (so it starts with a letter)
151 if (result != '') and (ord(result[0]) in range (ord('0'), ord('9'))):
152 result = 't_g' + result
155 texinfo_re = re.compile (r'@.*{(.*)}')
156 def remove_texinfo (title):
157 return texinfo_re.sub (r'\1', title)
159 def create_texinfo_anchor (title):
160 return texinfo_file_name (remove_texinfo (title))
162 unnumbered_re = re.compile (r'unnumbered.+|lydoctitle')
163 file_name_section_level = {
172 'unnumberedsubsec':1,
175 'unnumberedsubsubsec':0,
176 'appendixsubsubsec':0
178 if split in file_name_section_level:
179 splitting_level = file_name_section_level[split]
182 def process_sections (filename, lang_suffix, page):
183 sections = section_translation_re.findall (page)
184 basename = os.path.splitext (os.path.basename (filename))[0]
185 p = os.path.join (outdir, basename) + lang_suffix + '.xref-map'
189 this_filename = 'index'
191 this_unnumbered = False
195 # Write out the cached values to the file and start a new
197 if this_title and this_title != 'Top':
198 f.write (this_title + "\t" + this_filename + "\t" + this_anchor + "\n")
200 this_title = remove_texinfo (sec[1])
201 this_anchor = create_texinfo_anchor (sec[1])
202 # delete entry from master map file
203 if this_title in initial_map:
204 del initial_map[this_title]
205 elif sec[0] == "translationof":
206 (original_node, external_node) = external_node_re.subn ('', sec[1])
207 original_node = remove_texinfo (original_node)
208 # The following binds the translator to use the
209 # translated node name in cross-references in case
211 if external_node and original_node in initial_map:
212 del initial_map[original_node]
213 anchor = create_texinfo_anchor (sec[1])
214 # If @translationof is used, it gives the original
215 # node name, which we use for the anchor and the file
216 # name (if it is a numbered node)
218 if not this_unnumbered:
219 this_filename = anchor
220 elif original_node in initial_map:
221 this_filename = initial_map[original_node][2]
223 # Some pages might not use a node for every section, so
224 # treat this case here, too: If we already had a section
225 # and encounter another one before the next @node, we
226 # write out the old one and start with the new values
227 if had_section and this_title:
228 f.write (this_title + "\t" + this_filename + "\t" + this_anchor + "\n")
229 this_title = remove_texinfo (sec[1])
230 this_anchor = create_texinfo_anchor (sec[1])
233 if split == 'custom':
234 # unnumbered nodes use the previously used file name,
235 # only numbered nodes get their own filename! However,
236 # top-level @unnumbered still get their own file.
237 this_unnumbered = unnumbered_re.match (sec[0])
238 if not this_unnumbered:
239 this_filename = this_anchor
240 elif split == 'node':
241 this_filename = this_anchor
243 if sec[0] in file_name_section_level and \
244 file_name_section_level[sec[0]] >= splitting_level:
245 this_filename = this_anchor
247 if this_title and this_title != 'Top':
248 f.write (this_title + "\t" + this_filename + "\t" + this_anchor + "\n")
250 for node in initial_map:
251 f.write ("\t".join (initial_map[node]) + "\n")
254 xref_map_line_re = re.compile (r'(.*?)\t(.*?)\t(.*?)$')
256 for line in open (master_map_file):
257 m = xref_map_line_re.match (line)
259 initial_map[m.group (1)] = (m.group (1), m.group (2), m.group (3))
261 for filename in files:
262 print "extract_texi_filenames.py: Processing %s" % filename
263 (lang_suffix, sections) = extract_sections (filename)
264 process_sections (filename, lang_suffix, sections)