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/out-www/notation-big-page',
17 'Documentation/out-www/internals-big-page',
18 'Documentation/out-www/learning-big-page',
19 'Documentation/out-www/usage-big-page',
20 'Documentation/out-www/music-glossary-big-page',
21 'Documentation/out-www/contributor',
22 'Documentation/out-www/changes-big-page',
23 'Documentation/out-www/essay-big-page',
24 'Documentation/out-www/extending-big-page',
25 'Documentation/out-www/snippets',
27 'Documentation/topdocs',
28 'Documentation/bibliography',
29 'Documentation/out-www/DEDICATION',
40 <p class="footer_version">
41 %(footer_name_version)s
43 <p class="footer_report">
44 %(footer_report_links)s
54 footer_name_version = _doc ('This page is for %(package_name)s-%(package_version)s (%(branch_str)s).')
55 # ugh, must not have "_doc" in strings because it is naively replaced with "_" in hacked gettext process
56 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>.')
59 mail_address = 'http://post.gmane.org/post.php?group=gmane.comp.gnu.lilypond.bugs'
60 help_us_url = 'http://lilypond.org/help-us.html'
62 header_tag = '<!-- header_tag -->'
63 header_tag_re = re.compile (header_tag)
65 footer_tag = '<!-- footer_tag -->'
66 footer_tag_re = re.compile (footer_tag)
68 lang_available = _doc ("Other languages: %s.")
69 browser_lang = _doc ('About <a href="%s">automatic language selection</a>.')
70 browser_language_url = "http://www.lilypond.org/website/misc/browser-language"
72 LANGUAGES_TEMPLATE = '''
74 %(language_available)s
81 html_re = re.compile ('(.*?)(?:[.]([^/.]*))?[.]html$')
84 def build_pages_dict (filelist):
85 """Build dictionary of available translations of each page"""
91 if len (g) <= 1 or g[1] == None:
95 if not g[0] in pages_dict:
96 pages_dict[g[0]] = [e]
98 pages_dict[g[0]].append (e)
100 def source_links_replace (m, source_val):
101 return 'href="' + os.path.join (source_val, m.group (1)) + '"'
103 # More hardcoding, yay!
104 split_docs_re = re.compile('(Documentation/out-www/(automated-engraving|essay|notation|changes|extending|music-glossary|usage|web|learning|snippets|contributor))/')
105 lily_examples_re = re.compile ('(href|src)="(ly-examples/.*?)"')
106 lily_snippets_re = re.compile ('(href|src)="([0-9a-f]{2}/lily-.*?)"')
107 pictures_re = re.compile ('src="(pictures/.*?)"')
109 docindex_link_re = re.compile (r'href="index.html"')
110 manuals_page_link_re = re.compile (r'href="((?:\.\./)+)Documentation/web/manuals')
112 ## Windows does not support symlinks.
113 # This function avoids creating symlinks for split HTML manuals
114 # Get rid of symlinks in GNUmakefile.in (local-WWW-post)
115 # this also fixes missing PNGs only present in translated docs
116 def hack_urls (s, prefix, target, is_development_branch):
117 depth = (prefix.count ('/') - 1) * '../'
119 s = css_re.sub ('<link \\1href="%(rel)sDocumentation/css/\\2"\\3>' % {'rel': depth}, s)
121 if split_docs_re.match (prefix):
122 s = lily_examples_re.sub ('\\1="../\\2"', s)
123 s = lily_snippets_re.sub ('\\1="../\\2"', s)
124 s = pictures_re.sub ('src="../\\1"', s)
126 # we also need to replace in the lsr, which is already processed above!
127 if 'input/' in prefix or 'Documentation/topdocs' in prefix or \
128 'Documentation/contributor' in prefix:
129 # fix the link from the regtest, lsr and topdoc pages to the doc index
130 # (rewrite prefix to obtain the relative path of the doc index page)
131 rel_link = re.sub (r'out-www/.*$', '', prefix)
132 rel_link = re.sub (r'[^/]*/', '../', rel_link)
133 if 'input/regression' in prefix or 'Documentation/contributor' in prefix:
134 indexfile = "Documentation/devel"
137 s = docindex_link_re.sub ('href="' + rel_link + indexfile + '.html\"', s)
138 # make the "return to doc index" work with the online website.
139 if target == 'online':
140 if (('Documentation/contributor' in prefix) or
141 is_development_branch):
142 manuals_page = 'development'
144 manuals_page = 'manuals'
145 s = manuals_page_link_re.sub (r'href="../../\1website/%s'
147 source_path = os.path.join (os.path.dirname (prefix), 'source')
148 if not os.path.islink (source_path):
150 source_val = os.readlink (source_path)
151 return re.sub ('href="source/(.*?)"', lambda m: source_links_replace (m, source_val), s)
153 body_tag_re = re.compile ('(?i)<body([^>]*)>')
154 html_tag_re = re.compile ('(?i)<html>')
155 doctype_re = re.compile ('(?i)<!DOCTYPE')
156 doctype = '<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">\n'
157 css_re = re.compile ('(?i)<link ([^>]*)href="[^">]*?(lilypond.*\.css)"([^>]*)>')
158 end_head_tag_re = re.compile ('(?i)</head>')
159 css_link = """ <link rel="stylesheet" type="text/css" title="Default design" href="%(rel)sDocumentation/css/lilypond-manuals.css">
161 <link href="%(rel)sDocumentation/css/lilypond-ie-fixes.css" rel="stylesheet" type="text/css">
166 def add_header (s, prefix):
167 """Add header (<body>, doctype and CSS)"""
168 if header_tag_re.search (s) == None:
170 (s, n) = body_tag_re.subn (body + header, s, 1)
172 (s, n) = html_tag_re.subn ('<html>' + header, s, 1)
176 if doctype_re.search (s) == None:
177 s = doctype + header_tag + '\n' + s
179 if css_re.search (s) == None:
180 depth = (prefix.count ('/') - 1) * '../'
181 s = end_head_tag_re.sub ((css_link % {'rel': depth}) + '</head>', s)
184 title_tag_re = re.compile ('.*?<title>(.*?)</title>', re.DOTALL)
185 AT_web_title_re = re.compile ('@WEB-TITLE@')
189 # maybe find first node?
190 fallback_web_title = '-- --'
191 m = title_tag_re.match (s)
193 fallback_web_title = m.group (1)
194 s = AT_web_title_re.sub (fallback_web_title, s)
197 footer_insert_re = re.compile ('<!--\s*FOOTER\s*-->')
198 end_body_re = re.compile ('(?i)</body>')
199 end_html_re = re.compile ('(?i)</html>')
201 def add_footer (s, footer_text):
203 (s, n) = footer_insert_re.subn (footer_text + '\n' + '<!-- FOOTER -->', s, 1)
205 (s, n) = end_body_re.subn (footer_text + '\n' + '</body>', s, 1)
207 (s, n) = end_html_re.subn (footer_text + '\n' + '</html>', s, 1)
209 s += footer_text + '\n'
212 def find_translations (prefix, lang_ext):
213 """find available translations of a page"""
216 for l in langdefs.LANGUAGES:
219 if e in pages_dict[prefix]:
221 elif lang_ext == '' and l.enabled and reduce (operator.and_,
222 [not prefix.startswith (s)
223 for s in non_copied_pages]):
224 # English version of missing translated pages will be written
226 return available, missing
228 online_links_re = re.compile ('''(href|src)=['"]\
229 ((?!Compiling-from-source.html")[^/][.]*[^.:'"]*)\
230 ([.]html)(#[^"']*|)['"]''')
231 offline_links_re = re.compile ('href=[\'"]\
232 ((?!Compiling-from-source.html")(?![.]{2}/contributor)[^/][.]*[^.:\'"]*)([.]html)(#[^"\']*|)[\'"]')
233 big_page_name_re = re.compile ('''(.+?)-big-page''')
235 def process_i18n_big_page_links (match, prefix, lang_ext):
236 big_page_name = big_page_name_re.match (match.group (1))
238 destination_path = os.path.normpath (os.path.join (os.path.dirname (prefix),
239 big_page_name.group (0)))
240 if not (destination_path in pages_dict and
241 lang_ext in pages_dict[destination_path]):
242 return match.group (0)
243 return 'href="' + match.group (1) + '.' + lang_ext \
244 + match.group (2) + match.group (3) + '"'
246 def process_links (s, prefix, lang_ext, file_name, missing, target):
248 if target == 'online':
249 # Strip .html, suffix for auto language selection (content
250 # negotiation). The menu must keep the full extension, so do
251 # this before adding the menu.
252 page_flavors[file_name] = \
253 [lang_ext, online_links_re.sub ('\\1="\\2\\4"', s)]
254 elif target == 'offline':
255 # in LANG doc index: don't rewrite .html suffixes
256 # as not all .LANG.html pages exist;
257 # the doc index should be translated and contain links with the right suffixes
259 if prefix in ('Documentation/out-www/index', 'Documentation/topdocs/out-www/NEWS'):
260 page_flavors[file_name] = [lang_ext, s]
262 page_flavors[file_name] = [lang_ext, s]
264 page_flavors[langdefs.lang_file_name (prefix, e, '.html')] = \
265 [e, offline_links_re.sub ('href="\\1.' + e + '\\2\\3"', s)]
267 # For saving bandwidth and disk space, we don't duplicate big pages
268 # in English, so we must process translated big pages links differently.
269 if 'big-page' in prefix:
270 page_flavors[file_name] = \
272 offline_links_re.sub \
273 (lambda match: process_i18n_big_page_links (match, prefix, lang_ext),
276 page_flavors[file_name] = \
278 offline_links_re.sub ('href="\\1.' + lang_ext + '\\2\\3"', s)]
281 def add_menu (page_flavors, prefix, available, target, translation):
282 for k in page_flavors:
285 if page_flavors[k][0] != '':
286 t = translation[page_flavors[k][0]]
289 for lang in available:
290 lang_file = lang.file_name (os.path.basename (prefix), '.html')
291 if language_menu != '':
292 language_menu += ', '
293 language_menu += '<a href="%s">%s</a>' % (lang_file, t (lang.name))
294 browser_language = t (browser_lang) % browser_language_url
296 language_available = t (lang_available) % language_menu
297 languages = LANGUAGES_TEMPLATE % vars ()
298 page_flavors[k][1] = add_footer (page_flavors[k][1], languages)
302 def process_html_files (package_name = '',
303 package_version = '',
305 name_filter = lambda s: s):
306 """Add header, footer and tweak links to a number of HTML files
309 package_name=NAME set package_name to NAME
310 package_version=VERSION set package version to VERSION
311 targets=offline|online set page processing depending on the target
312 offline is for reading HTML pages locally
313 online is for hosting the HTML pages on a website with content
315 name_filter a HTML file name filter
317 translation = langdefs.translation
318 localtime = time.strftime ('%c %Z', time.localtime (time.time ()))
320 if "http://" in mail_address:
321 mail_address_url = mail_address
323 mail_address_url= 'mailto:' + mail_address
325 versiontup = package_version.split ('.')
326 branch_str = _doc ('stable-branch')
327 if int (versiontup[1]) % 2:
328 branch_str = _doc ('development-branch')
330 # Initialize dictionaries for string formatting
332 subst[''] = dict ([i for i in globals ().items() if type (i[1]) is str])
333 subst[''].update (dict ([i for i in locals ().items() if type (i[1]) is str]))
334 for l in translation:
335 e = langdefs.LANGDICT[l].webext
338 for name in subst['']:
339 subst[e][name] = translation[l] (subst[''][name])
340 # Do deeper string formatting as early as possible,
341 # so only one '%' formatting pass is needed later
343 subst[e]['footer_name_version'] = subst[e]['footer_name_version'] % subst[e]
344 subst[e]['footer_report_links'] = subst[e]['footer_report_links'] % subst[e]
346 for prefix, ext_list in pages_dict.items ():
347 for lang_ext in ext_list:
348 file_name = langdefs.lang_file_name (prefix, lang_ext, '.html')
349 source_time = os.path.getmtime(file_name)
351 if os.path.exists(name_filter(file_name)):
352 dest_time = os.path.getmtime(name_filter(file_name))
353 if dest_time < source_time:
355 in_f = open (file_name)
359 s = s.replace ('%', '%%')
360 s = hack_urls (s, prefix, target, bool (int (versiontup[1]) % 2))
361 s = add_header (s, prefix)
364 if footer_tag_re.search (s) == None:
365 if 'web' in file_name:
366 s = add_footer (s, footer_tag + web_footer)
368 s = add_footer (s, footer_tag + footer)
370 available, missing = find_translations (prefix, lang_ext)
371 page_flavors = process_links (s, prefix, lang_ext, file_name, missing, target)
372 # Add menu after stripping: must not have autoselection for language menu.
373 page_flavors = add_menu (page_flavors, prefix, available, target, translation)
374 for k in page_flavors:
375 page_flavors[k][1] = page_flavors[k][1] % subst[page_flavors[k][0]]
376 out_f = open (name_filter (k), 'w')
377 out_f.write (page_flavors[k][1])
379 # if the page is translated, a .en.html symlink is necessary for content negotiation
380 if target == 'online' and ext_list != [''] and not os.path.lexists (name_filter (prefix + '.en.html')):
381 os.symlink (os.path.basename (prefix) + '.html', name_filter (prefix + '.en.html'))