4 Postprocess HTML files:
5 add footer, tweak links, add language selection menu.
14 # This is to try to make the docball not too big with almost duplicate files
16 non_copied_pages = ['Documentation/user/out-www/lilypond-big-page',
17 'Documentation/user/out-www/lilypond-internals-big-page',
18 'Documentation/user/out-www/lilypond-learning-big-page',
19 'Documentation/user/out-www/lilypond-program-big-page',
20 'Documentation/user/out-www/music-glossary-big-page',
22 'Documentation/topdocs',
23 'Documentation/bibliography',
24 'Documentation/out-www/THANKS',
25 'Documentation/out-www/DEDICATION',
26 'Documentation/out-www/devel',
37 <p class="footer_version">
38 %(footer_name_version)s
40 <p class="footer_report">
41 %(footer_report_links)s
45 footer_name_version = _doc ('This page is for %(package_name)s-%(package_version)s (%(branch_str)s).')
46 # ugh, must not have "_doc" in strings because it is naively replaced with "_" in hacked gettext process
47 footer_report_links = _doc ('Your <a href="%(suggest_Docs_url)s">suggestions for the documentation</a> are welcome, please report errors to our <a href="%(mail_address_url)s">bug list</a>.')
50 mail_address = 'http://post.gmane.org/post.php?group=gmane.comp.gnu.lilypond.bugs'
51 suggest_Docs_url = 'http://lilypond.org/web/devel/participating/documentation-adding'
53 header_tag = '<!-- header_tag -->'
54 header_tag_re = re.compile (header_tag)
56 footer_tag = '<!-- footer_tag -->'
57 footer_tag_re = re.compile (footer_tag)
59 lang_available = _doc ("Other languages: %s.")
60 browser_lang = _doc ('About <A HREF="%s">automatic language selection</A>.')
61 browser_language_url = "/web/about/browser-language"
63 LANGUAGES_TEMPLATE = '''
65 %(language_available)s
72 html_re = re.compile ('(.*?)(?:[.]([^/.]*))?[.]html$')
75 def build_pages_dict (filelist):
76 """Build dictionary of available translations of each page"""
82 if len (g) <= 1 or g[1] == None:
86 if not g[0] in pages_dict:
87 pages_dict[g[0]] = [e]
89 pages_dict[g[0]].append (e)
91 def source_links_replace (m, source_val):
92 return 'href="' + os.path.join (source_val, m.group (1)) + '"'
94 splitted_docs_re = re.compile ('(input/lsr/out-www/lilypond-snippets|\
95 Documentation/user/out-www/(lilypond|music-glossary|lilypond-program|\
96 lilypond-learning))/')
98 snippets_ref_re = re.compile (r'href="(\.\./)?lilypond-snippets')
99 user_ref_re = re.compile ('href="(?:\.\./)?lilypond\
100 (-internals|-learning|-program|(?!-snippets))')
102 docindex_link_re = re.compile (r'href="index.html"')
105 ## Windows does not support symlinks.
106 # This function avoids creating symlinks for splitted HTML manuals
107 # Get rid of symlinks in GNUmakefile.in (local-WWW-post)
108 # this also fixes missing PNGs only present in translated docs
109 def hack_urls (s, prefix):
110 if splitted_docs_re.match (prefix):
111 s = re.sub ('(href|src)="(../lily-.*?|.*?[.]png)"', '\\1="../\\2"', s)
113 # fix xrefs between documents in different directories ad hoc
114 if 'user/out-www/lilypond' in prefix:
115 s = snippets_ref_re.sub ('href="source/input/lsr/lilypond-snippets', s)
116 elif 'input/lsr' in prefix:
117 s = user_ref_re.sub ('href="source/Documentation/user/lilypond\\1', s)
119 # we also need to replace in the lsr, which is already processed above!
120 if 'input/' in prefix or 'Documentation/topdocs' in prefix:
121 # fix the link from the regtest, lsr and topdoc pages to the doc index
122 # (rewrite prefix to obtain the relative path of the doc index page)
123 rel_link = re.sub (r'out-www/.*$', '', prefix)
124 rel_link = re.sub (r'[^/]*/', '../', rel_link)
125 if 'input/regression' in prefix:
126 indexfile = "Documentation/devel/index"
129 s = docindex_link_re.sub ('href="' + rel_link + indexfile + '.html\"', s)
131 source_path = os.path.join (os.path.dirname (prefix), 'source')
132 if not os.path.islink (source_path):
134 source_val = os.readlink (source_path)
135 return re.sub ('href="source/(.*?)"', lambda m: source_links_replace (m, source_val), s)
137 body_tag_re = re.compile ('(?i)<body([^>]*)>')
138 html_tag_re = re.compile ('(?i)<html>')
139 doctype_re = re.compile ('(?i)<!DOCTYPE')
140 doctype = '<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">\n'
141 css_re = re.compile ('(?i)<link rel="stylesheet" type="text/css" ([^>]*)href="[^">]*?lilypond.*\.css"([^>]*)>')
142 end_head_tag_re = re.compile ('(?i)</head>')
143 css_link = """ <link rel="stylesheet" type="text/css" title="Patrick McCarty's design" href="%(rel)sDocumentation/lilypond-mccarty.css">
144 <link rel="alternate stylesheet" type="text/css" href="%(rel)sDocumentation/lilypond.css" title="Andrew Hawryluk's design">
145 <link rel="alternate stylesheet" type="text/css" href="%(rel)sDocumentation/lilypond-blue.css" title="Kurt Kroon's blue design">
147 <link href="%(rel)sDocumentation/lilypond-ie-fixes.css" rel="stylesheet" type="text/css">
152 def add_header (s, prefix):
153 """Add header (<body>, doctype and CSS)"""
154 if header_tag_re.search (s) == None:
156 (s, n) = body_tag_re.subn (body + header, s, 1)
158 (s, n) = html_tag_re.subn ('<html>' + header, s, 1)
162 s = header_tag + '\n' + s
164 if doctype_re.search (s) == None:
167 if css_re.search (s) == None:
168 depth = (prefix.count ('/') - 1) * '../'
169 s = end_head_tag_re.sub ((css_link % {'rel': depth}) + '</head>', s)
172 title_tag_re = re.compile ('.*?<title>(.*?)</title>', re.DOTALL)
173 AT_web_title_re = re.compile ('@WEB-TITLE@')
177 # maybe find first node?
178 fallback_web_title = '-- --'
179 m = title_tag_re.match (s)
181 fallback_web_title = m.group (1)
182 s = AT_web_title_re.sub (fallback_web_title, s)
185 footer_insert_re = re.compile ('<!--\s*FOOTER\s*-->')
186 end_body_re = re.compile ('(?i)</body>')
187 end_html_re = re.compile ('(?i)</html>')
189 def add_footer (s, footer_text):
191 (s, n) = footer_insert_re.subn (footer_text + '\n' + '<!-- FOOTER -->', s, 1)
193 (s, n) = end_body_re.subn (footer_text + '\n' + '</body>', s, 1)
195 (s, n) = end_html_re.subn (footer_text + '\n' + '</html>', s, 1)
197 s += footer_text + '\n'
200 def find_translations (prefix, lang_ext):
201 """find available translations of a page"""
204 for l in langdefs.LANGUAGES:
207 if e in pages_dict[prefix]:
209 elif lang_ext == '' and l.enabled and reduce (operator.and_,
210 [not prefix.startswith (s)
211 for s in non_copied_pages]):
212 # English version of missing translated pages will be written
214 return available, missing
216 online_links_re = re.compile ('''(href|src)=['"]\
217 ((?!Compiling-from-source.html")[^/][.]*[^.:'"]*)\
218 ([.]html)(#[^"']*|)['"]''')
219 offline_links_re = re.compile ('href=[\'"]\
220 ((?!Compiling-from-source.html")[^/][.]*[^.:\'"]*)([.]html)(#[^"\']*|)[\'"]')
221 big_page_name_re = re.compile ('''(.+?)-big-page''')
223 def process_i18n_big_page_links (match, prefix, lang_ext):
224 big_page_name = big_page_name_re.match (match.group (1))
226 destination_path = os.path.normpath (os.path.join (os.path.dirname (prefix),
227 big_page_name.group (0)))
228 if not lang_ext in pages_dict[destination_path]:
229 return match.group (0)
230 return 'href="' + match.group (1) + '.' + lang_ext \
231 + match.group (2) + match.group (3) + '"'
233 def process_links (s, prefix, lang_ext, file_name, missing, target):
235 if target == 'online':
236 # Strip .html, suffix for auto language selection (content
237 # negotiation). The menu must keep the full extension, so do
238 # this before adding the menu.
239 page_flavors[file_name] = \
240 [lang_ext, online_links_re.sub ('\\1="\\2\\4"', s)]
241 elif target == 'offline':
242 # in LANG doc index: don't rewrite .html suffixes
243 # as not all .LANG.html pages exist;
244 # the doc index should be translated and contain links with the right suffixes
246 if prefix in ('Documentation/out-www/index', 'Documentation/topdocs/out-www/NEWS'):
247 page_flavors[file_name] = [lang_ext, s]
249 page_flavors[file_name] = [lang_ext, s]
251 page_flavors[langdefs.lang_file_name (prefix, e, '.html')] = \
252 [e, offline_links_re.sub ('href="\\1.' + e + '\\2\\3"', s)]
254 # For saving bandwidth and disk space, we don't duplicate big pages
255 # in English, so we must process translated big pages links differently.
256 if 'big-page' in prefix:
257 page_flavors[file_name] = \
259 offline_links_re.sub \
260 (lambda match: process_i18n_big_page_links (match, prefix, lang_ext),
263 page_flavors[file_name] = \
265 offline_links_re.sub ('href="\\1.' + lang_ext + '\\2\\3"', s)]
268 def add_menu (page_flavors, prefix, available, target, translation):
269 for k in page_flavors:
272 if page_flavors[k][0] != '':
273 t = translation[page_flavors[k][0]]
276 for lang in available:
277 lang_file = lang.file_name (os.path.basename (prefix), '.html')
278 if language_menu != '':
279 language_menu += ', '
280 language_menu += '<a href="%s">%s</a>' % (lang_file, t (lang.name))
281 if target == 'offline':
282 browser_language = ''
283 elif target == 'online':
284 browser_language = t (browser_lang) % browser_language_url
286 language_available = t (lang_available) % language_menu
287 languages = LANGUAGES_TEMPLATE % vars ()
288 page_flavors[k][1] = add_footer (page_flavors[k][1], languages)
292 def process_html_files (package_name = '',
293 package_version = '',
295 name_filter = lambda s: s):
296 """Add header, footer and tweak links to a number of HTML files
299 package_name=NAME set package_name to NAME
300 package_version=VERSION set package version to VERSION
301 targets=offline|online set page processing depending on the target
302 offline is for reading HTML pages locally
303 online is for hosting the HTML pages on a website with content
305 name_filter a HTML file name filter
307 translation = langdefs.translation
308 localtime = time.strftime ('%c %Z', time.localtime (time.time ()))
310 if "http://" in mail_address:
311 mail_address_url = mail_address
313 mail_address_url= 'mailto:' + mail_address
315 versiontup = package_version.split ('.')
316 branch_str = _doc ('stable-branch')
317 if int (versiontup[1]) % 2:
318 branch_str = _doc ('development-branch')
320 # Initialize dictionaries for string formatting
322 subst[''] = dict ([i for i in globals ().items() if type (i[1]) is str])
323 subst[''].update (dict ([i for i in locals ().items() if type (i[1]) is str]))
324 for l in translation:
325 e = langdefs.LANGDICT[l].webext
328 for name in subst['']:
329 subst[e][name] = translation[l] (subst[''][name])
330 # Do deeper string formatting as early as possible,
331 # so only one '%' formatting pass is needed later
333 subst[e]['footer_name_version'] = subst[e]['footer_name_version'] % subst[e]
334 subst[e]['footer_report_links'] = subst[e]['footer_report_links'] % subst[e]
336 for prefix, ext_list in pages_dict.items ():
337 for lang_ext in ext_list:
338 file_name = langdefs.lang_file_name (prefix, lang_ext, '.html')
339 in_f = open (file_name)
343 s = s.replace ('%', '%%')
344 s = hack_urls (s, prefix)
345 s = add_header (s, prefix)
348 if footer_tag_re.search (s) == None:
349 s = add_footer (s, footer_tag + footer)
351 available, missing = find_translations (prefix, lang_ext)
352 page_flavors = process_links (s, prefix, lang_ext, file_name, missing, target)
353 # Add menu after stripping: must not have autoselection for language menu.
354 page_flavors = add_menu (page_flavors, prefix, available, target, translation)
355 for k in page_flavors:
356 page_flavors[k][1] = page_flavors[k][1] % subst[page_flavors[k][0]]
357 out_f = open (name_filter (k), 'w')
358 out_f.write (page_flavors[k][1])
360 # if the page is translated, a .en.html symlink is necessary for content negotiation
361 if target == 'online' and ext_list != ['']:
362 os.symlink (os.path.basename (prefix) + '.html', name_filter (prefix + '.en.html'))