4 Postprocess HTML files:
5 add footer, tweak links, add language selection menu.
15 # This is to try to make the docball not too big with almost duplicate files
17 non_copied_pages = ['Documentation/out-www/notation-big-page',
18 'Documentation/out-www/internals-big-page',
19 'Documentation/out-www/learning-big-page',
20 'Documentation/out-www/usage-big-page',
21 'Documentation/out-www/music-glossary-big-page',
22 'Documentation/out-www/contributor',
23 'Documentation/out-www/changes-big-page',
24 'Documentation/out-www/essay-big-page',
25 'Documentation/out-www/extending-big-page',
26 'Documentation/out-www/snippets',
28 'Documentation/topdocs',
29 'Documentation/bibliography',
30 'Documentation/out-www/DEDICATION',
41 <!-- This footer was autogenerated by %s -->
42 <p class="footer_version">
43 %%(footer_name_version)s
45 <p class="footer_report">
46 %%(footer_report_links)s
53 <!-- This footer was autogenerated by %s -->
57 footer_name_version = _doc ('This page is for %(package_name)s-%(package_version)s (%(branch_str)s).')
58 # ugh, must not have "_doc" in strings because it is naively replaced with "_" in hacked gettext process
59 footer_report_links = _doc ('We welcome your aid; please <a href="%(help_us_url)s">help us</a> by reporting errors to our <a href="%(mail_address_url)s">bug list</a>.')
62 mail_address = 'http://post.gmane.org/post.php?group=gmane.comp.gnu.lilypond.bugs'
63 help_us_url = 'http://lilypond.org/help-us.html'
65 header_tag = '<!-- header_tag -->'
66 header_tag_re = re.compile (header_tag)
68 footer_tag = '<!-- footer_tag -->'
69 footer_tag_re = re.compile (footer_tag)
71 lang_available = _doc ("Other languages: %s.")
72 browser_lang = _doc ('About <a href="%s">automatic language selection</a>.')
73 browser_language_url = "http://www.lilypond.org/website/misc/browser-language"
75 LANGUAGES_TEMPLATE = '''
77 <!-- These links were autogenerated by %s -->
78 %%(language_available)s
85 html_re = re.compile ('(.*?)(?:[.]([^/.]*))?[.]html$')
88 def build_pages_dict (filelist):
89 """Build dictionary of available translations of each page"""
95 if len (g) <= 1 or g[1] == None:
99 if not g[0] in pages_dict:
100 pages_dict[g[0]] = [e]
102 pages_dict[g[0]].append (e)
104 def source_links_replace (m, source_val):
105 return 'href="' + os.path.join (source_val, m.group (1)) + '"'
107 # More hardcoding, yay!
108 split_docs_re = re.compile('(Documentation/out-www/(automated-engraving|essay|notation|changes|extending|music-glossary|usage|web|learning|snippets|contributor))/')
109 lily_examples_re = re.compile ('(href|src)="(ly-examples/.*?)"')
110 lily_snippets_re = re.compile ('(href|src)="([0-9a-f]{2}/lily-.*?)"')
111 pictures_re = re.compile ('src="(pictures/.*?)"')
113 docindex_link_re = re.compile (r'href="index.html"')
114 manuals_page_link_re = re.compile (r'href="((?:\.\./)+)Documentation/web/manuals')
116 ## Windows does not support symlinks.
117 # This function avoids creating symlinks for split HTML manuals
118 # Get rid of symlinks in GNUmakefile.in (local-WWW-post)
119 # this also fixes missing PNGs only present in translated docs
120 def hack_urls (s, prefix, target, is_development_branch):
121 depth = (prefix.count ('/') - 1) * '../'
123 s = css_re.sub ('<link \\1href="%(rel)sDocumentation/css/\\2"\\3>' % {'rel': depth}, s)
125 if split_docs_re.match (prefix):
126 s = lily_examples_re.sub ('\\1="../\\2"', s)
127 s = lily_snippets_re.sub ('\\1="../\\2"', s)
128 s = pictures_re.sub ('src="../\\1"', s)
130 # we also need to replace in the lsr, which is already processed above!
131 if 'input/' in prefix or 'Documentation/topdocs' in prefix or \
132 'Documentation/contributor' in prefix:
133 # fix the link from the regtest, lsr and topdoc pages to the doc index
134 # (rewrite prefix to obtain the relative path of the doc index page)
135 rel_link = re.sub (r'out-www/.*$', '', prefix)
136 rel_link = re.sub (r'[^/]*/', '../', rel_link)
137 if 'input/regression' in prefix or 'Documentation/contributor' in prefix:
138 indexfile = "Documentation/devel"
141 s = docindex_link_re.sub ('href="' + rel_link + indexfile + '.html\"', s)
142 # make the "return to doc index" work with the online website.
143 if target == 'online':
144 if (('Documentation/contributor' in prefix) or
145 is_development_branch):
146 manuals_page = 'development'
148 manuals_page = 'manuals'
149 s = manuals_page_link_re.sub (r'href="../../\1website/%s'
151 source_path = os.path.join (os.path.dirname (prefix), 'source')
152 if not os.path.islink (source_path):
154 source_val = os.readlink (source_path)
155 return re.sub ('href="source/(.*?)"', lambda m: source_links_replace (m, source_val), s)
157 body_tag_re = re.compile ('(?i)<body([^>]*)>')
158 html_tag_re = re.compile ('(?i)<html>')
159 doctype_re = re.compile ('(?i)<!DOCTYPE')
160 doctype = '<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">\n'
161 css_re = re.compile ('(?i)<link ([^>]*)href="[^">]*?(lilypond.*\.css)"([^>]*)>')
162 end_head_tag_re = re.compile ('(?i)</head>')
163 css_link = """ <link rel="stylesheet" type="text/css" title="Default design" href="%(rel)sDocumentation/css/lilypond-manuals.css">
165 <link href="%(rel)sDocumentation/css/lilypond-ie-fixes.css" rel="stylesheet" type="text/css">
170 def add_header (s, prefix):
171 """Add header (<body>, doctype and CSS)"""
172 if header_tag_re.search (s) == None:
174 (s, n) = body_tag_re.subn (body + header, s, 1)
176 (s, n) = html_tag_re.subn ('<html>' + header, s, 1)
180 if doctype_re.search (s) == None:
181 s = doctype + header_tag + '\n' + s
183 if css_re.search (s) == None:
184 depth = (prefix.count ('/') - 1) * '../'
185 s = end_head_tag_re.sub ((css_link % {'rel': depth}) + '</head>', s)
188 title_tag_re = re.compile ('.*?<title>(.*?)</title>', re.DOTALL)
189 AT_web_title_re = re.compile ('@WEB-TITLE@')
193 # maybe find first node?
194 fallback_web_title = '-- --'
195 m = title_tag_re.match (s)
197 fallback_web_title = m.group (1)
198 s = AT_web_title_re.sub (fallback_web_title, s)
201 footer_insert_re = re.compile ('<!--\s*FOOTER\s*-->')
202 end_body_re = re.compile ('(?i)</body>')
203 end_html_re = re.compile ('(?i)</html>')
205 def add_footer (s, footer_text):
207 (s, n) = footer_insert_re.subn (footer_text + '\n' + '<!-- FOOTER -->', s, 1)
209 (s, n) = end_body_re.subn (footer_text + '\n' + '</body>', s, 1)
211 (s, n) = end_html_re.subn (footer_text + '\n' + '</html>', s, 1)
213 s += footer_text + '\n'
216 def find_translations (prefix, lang_ext):
217 """find available translations of a page"""
220 for l in langdefs.LANGUAGES:
223 if e in pages_dict[prefix]:
225 elif lang_ext == '' and l.enabled and reduce (operator.and_,
226 [not prefix.startswith (s)
227 for s in non_copied_pages]):
228 # English version of missing translated pages will be written
230 return available, missing
232 online_links_re = re.compile ('''(href|src)=['"]\
233 ((?!Compiling-from-source.html")[^/][.]*[^.:'"]*)\
234 ([.]html)(#[^"']*|)['"]''')
235 offline_links_re = re.compile ('href=[\'"]\
236 ((?!Compiling-from-source.html")(?![.]{2}/contributor)[^/][.]*[^.:\'"]*)([.]html)(#[^"\']*|)[\'"]')
237 big_page_name_re = re.compile ('''(.+?)-big-page''')
239 def process_i18n_big_page_links (match, prefix, lang_ext):
240 big_page_name = big_page_name_re.match (match.group (1))
242 destination_path = os.path.normpath (os.path.join (os.path.dirname (prefix),
243 big_page_name.group (0)))
244 if not (destination_path in pages_dict and
245 lang_ext in pages_dict[destination_path]):
246 return match.group (0)
247 return 'href="' + match.group (1) + '.' + lang_ext \
248 + match.group (2) + match.group (3) + '"'
250 def process_links (s, prefix, lang_ext, file_name, missing, target):
252 if target == 'online':
253 # Strip .html, suffix for auto language selection (content
254 # negotiation). The menu must keep the full extension, so do
255 # this before adding the menu.
256 page_flavors[file_name] = \
257 [lang_ext, online_links_re.sub ('\\1="\\2\\4"', s)]
258 elif target == 'offline':
259 # in LANG doc index: don't rewrite .html suffixes
260 # as not all .LANG.html pages exist;
261 # the doc index should be translated and contain links with the right suffixes
263 if prefix in ('Documentation/out-www/index', 'Documentation/topdocs/out-www/NEWS'):
264 page_flavors[file_name] = [lang_ext, s]
266 page_flavors[file_name] = [lang_ext, s]
268 page_flavors[langdefs.lang_file_name (prefix, e, '.html')] = \
269 [e, offline_links_re.sub ('href="\\1.' + e + '\\2\\3"', s)]
271 # For saving bandwidth and disk space, we don't duplicate big pages
272 # in English, so we must process translated big pages links differently.
273 if 'big-page' in prefix:
274 page_flavors[file_name] = \
276 offline_links_re.sub \
277 (lambda match: process_i18n_big_page_links (match, prefix, lang_ext),
280 page_flavors[file_name] = \
282 offline_links_re.sub ('href="\\1.' + lang_ext + '\\2\\3"', s)]
285 def add_menu (page_flavors, prefix, available, target, translation):
286 for k in page_flavors:
289 if page_flavors[k][0] != '':
290 t = translation[page_flavors[k][0]]
293 for lang in available:
294 lang_file = lang.file_name (os.path.basename (prefix), '.html')
295 if language_menu != '':
296 language_menu += ', '
297 language_menu += '<a href="%s">%s</a>' % (lang_file, t (lang.name))
298 browser_language = t (browser_lang) % browser_language_url
300 language_available = t (lang_available) % language_menu
301 languages = LANGUAGES_TEMPLATE % vars ()
302 page_flavors[k][1] = add_footer (page_flavors[k][1], languages)
306 def process_html_files (package_name = '',
307 package_version = '',
309 name_filter = lambda s: s):
310 """Add header, footer and tweak links to a number of HTML files
313 package_name=NAME set package_name to NAME
314 package_version=VERSION set package version to VERSION
315 targets=offline|online set page processing depending on the target
316 offline is for reading HTML pages locally
317 online is for hosting the HTML pages on a website with content
319 name_filter a HTML file name filter
321 translation = langdefs.translation
322 localtime = time.strftime ('%c %Z', time.localtime (time.time ()))
324 if "http://" in mail_address:
325 mail_address_url = mail_address
327 mail_address_url= 'mailto:' + mail_address
329 versiontup = package_version.split ('.')
330 branch_str = _doc ('stable-branch')
331 if int (versiontup[1]) % 2:
332 branch_str = _doc ('development-branch')
334 # Initialize dictionaries for string formatting
336 subst[''] = dict ([i for i in globals ().items() if type (i[1]) is str])
337 subst[''].update (dict ([i for i in locals ().items() if type (i[1]) is str]))
338 for l in translation:
339 e = langdefs.LANGDICT[l].webext
342 for name in subst['']:
343 subst[e][name] = translation[l] (subst[''][name])
344 # Do deeper string formatting as early as possible,
345 # so only one '%' formatting pass is needed later
347 subst[e]['footer_name_version'] = subst[e]['footer_name_version'] % subst[e]
348 subst[e]['footer_report_links'] = subst[e]['footer_report_links'] % subst[e]
350 for prefix, ext_list in pages_dict.items ():
351 for lang_ext in ext_list:
352 file_name = langdefs.lang_file_name (prefix, lang_ext, '.html')
353 source_time = os.path.getmtime(file_name)
355 if os.path.exists(name_filter(file_name)):
356 dest_time = os.path.getmtime(name_filter(file_name))
357 if dest_time < source_time:
359 in_f = open (file_name)
363 s = s.replace ('%', '%%')
364 s = hack_urls (s, prefix, target, bool (int (versiontup[1]) % 2))
365 s = add_header (s, prefix)
368 if footer_tag_re.search (s) == None:
369 if 'web' in file_name:
370 s = add_footer (s, footer_tag + web_footer)
372 s = add_footer (s, footer_tag + footer)
374 available, missing = find_translations (prefix, lang_ext)
375 page_flavors = process_links (s, prefix, lang_ext, file_name, missing, target)
376 # Add menu after stripping: must not have autoselection for language menu.
377 page_flavors = add_menu (page_flavors, prefix, available, target, translation)
378 for k in page_flavors:
379 page_flavors[k][1] = page_flavors[k][1] % subst[page_flavors[k][0]]
380 out_f = open (name_filter (k), 'w')
381 out_f.write (page_flavors[k][1])
383 # if the page is translated, a .en.html symlink is necessary for content negotiation
384 if target == 'online' and ext_list != [''] and not os.path.lexists (name_filter (prefix + '.en.html')):
385 os.symlink (os.path.basename (prefix) + '.html', name_filter (prefix + '.en.html'))