X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=buildscripts%2Fhtml-gettext.py;h=65efe8f28396637b75ffd2f2fea7655632205066;hb=29b060250128c577ab8f3cc4837d35b4ace625bf;hp=97f8e5f6f66fbc806490d01a12625800cb3f34e9;hpb=e0fe5d924082970ba7280ccb148eadababf00983;p=lilypond.git

diff --git a/buildscripts/html-gettext.py b/buildscripts/html-gettext.py
index 97f8e5f6f6..65efe8f283 100644
--- a/buildscripts/html-gettext.py
+++ b/buildscripts/html-gettext.py
@@ -1,45 +1,58 @@
 #!@PYTHON@
 # html-gettext.py
 
-# Usage:  html-gettext.py [-o OUTDIR] LOCALEDIR LANG FILES
+# USAGE:  html-gettext.py [-o OUTDIR] BUILDSCRIPT-DIR LOCALEDIR LANG FILES
 #
 # -o OUTDIR specifies that output files should be written in OUTDIR
 #    rather than be overwritten
 #
-# LANG
-# LOCALEDIR should contain 'lilypond-doc' message catalogs
-
-
-### DATA
-# Currently, typo_rules[LANG] only defines the HTML-coded space occuring
-# before double punctuation characters (i.e. : ; ? ! ) for LANG
-
-typo_rules = { 'fr':' ', 'default':'' }
-
-
-### PROGRAM
 
 import sys
 import re
 import os
-import string
-import gettext
 import getopt
+import gettext
 
 optlist, args = getopt.getopt(sys.argv[1:],'o:')
+buildscript_dir, localedir, lang = args[0:3]
 
 outdir = '.'
 for x in optlist:
 	if x[0] == '-o':
 		outdir = x[1]
 
-if args[1] in typo_rules.keys():
-	dbl_punct_char_separator = typo_rules[args[1]]
-else:
-	dbl_punct_char_separator = typo_rules['default']
-
-t = gettext.translation('lilypond-doc', args[0], [args[1]])
-_ = t.gettext
+sys.path.append (buildscript_dir)
+import langdefs
+
+double_punct_char_separator = langdefs.LANGDICT[lang].double_punct_char_sep
+t = gettext.translation('lilypond-doc', localedir, [lang])
+my_gettext = t.gettext
+
+html_codes = ((' -- ', ' – '),
+	      (' --- ', ' — '))
+html2texi = {'command': (re.compile (r'<samp><span class="command">(.*?)</span></samp>'), r'@command{\1}'),
+	     'code': (re.compile (r'<code>(.*?)</code>'), r'@code{\1}')
+	     }
+texi2html = {'command': (re.compile (r'@command{(.*?)}'), r'<samp><span class="command">\1</span></samp>'),
+	     'code': (re.compile (r'@code{(.*?)}'), r'<code>\1</code>')
+	     }
+whitespaces = re.compile (r'\s+')
+
+
+def _ (s):
+	if not s:
+		return ''
+	s = whitespaces.sub (' ', s)
+	for c in html_codes:
+		s = s.replace (c[1], c[0])
+	for u in html2texi.values():
+		s = u[0].sub (u[1], s)
+	s = my_gettext (s)
+	for u in texi2html.values():
+		s = u[0].sub (u[1], s)
+	for c in html_codes:
+		s = s.replace (c[0], c[1])
+	return s
 
 def link_gettext (m):
 	return '<link rel="' + m.group(1) + '" ' + m.group(2) + ' title="' + _(m.group(3)) + '">'
@@ -48,26 +61,40 @@ def title_gettext (m):
 	return '<title>' + _(m.group(1)) + ' - ' + m.group(2) + '</title>'
 
 def a_href_gettext (m):
-	if m.group(4) == ':':
-		s = dbl_punct_char_separator + ':'
-	elif m.group(4) == None:
-		s = ''
-	return '<a ' + (m.group(1) or '') + m.group(2) + _(m.group(3)) + '</a>' + s
+	s = ''
+	if m.group(0)[-1] == ':':
+		s = double_punct_char_separator + ':'
+	t = ''
+	if m.lastindex == 7:
+		t = m.group(7)
+	return '<a ' + (m.group(1) or '') + m.group(2) + (m.group(3) or '') + _(m.group(4)) + m.group(5) + _(m.group(6)) + t + '</a>' + s
 
 def h_gettext (m):
-	return '<h' + m.group(1) + ' class="' + m.group(2) + '">' + \
-	       (m.group(3) or '') + _(m.group(4)) + '</h' + m.group(1) + '>'
+	if m.group (3):
+		s = _(m.group(3))
+	else:
+		s= ''
+	return '<h' + m.group(1) + m.group(2) + '>' + s +\
+	       m.group(4) + _(m.group(5)) + '</h' + m.group(1) + '>'
+
+def crossmanual_ref_gettext (m):
+	return '<a href="' + m.group(1) + '">' + _(m.group(2)) + '</a>'
 
-for filename in args[2:]:
+for filename in args[3:]:
 	f = open (filename, 'r')
 	page = f.read ()
 	f.close()
 	page = re.sub (r'<link rel="(up|prev|next)" (.*?) title="([^"]*?)">', link_gettext, page)
 	page = re.sub (r'<title>([^<]*?) - ([^<]*?)</title>', title_gettext, page)
-	page = re.sub (r'<a ((?:rel="\w+")? ?(?:accesskey="[^"]+?" ?)?)(href="[^"]+?">)([^<]+)</a>(:)?', a_href_gettext, page)
-	page = re.sub (r'<h(\d) class="(\w+)">([\d.]+ )?([^<]+)</h\1>', h_gettext, page)
+	# ugh
+	page = re.sub (r'(?ms)<a ((?:rel="\w+")? ?(?:accesskey="[^"]+?")? ?(?:name=".*?")? ?)(href=".+?">)(<code>)?(Appendix )?([A-Z\d.]+ |)(.+?)(?(3)</code>)</a>:?', a_href_gettext, page)
+	page = re.sub (r'<h(\d)( class="\w+"|)>(Appendix |)([A-Z\d.]+ |)?([^<]+)</h\1>', h_gettext, page)
+	page = re.sub (r'<a href="(\.\./(?:music-glossary|lilypond-program/)?(?:.+?))">(.+?)</a>', crossmanual_ref_gettext, page)
+	# this is necessary for entries not translated by a_href_gettext
+	page = re.sub (r'<a href="(.+?)">(.+?)</a>', crossmanual_ref_gettext, page)
 	for w in ('Next:', 'Previous:', 'Up:'):
 		page = re.sub (w, _(w), page)
+	page = langdefs.LANGDICT[lang].html_filter (page)
 	f = open (os.path.join (outdir, filename), 'w')
 	f.write (page)
 	f.close ()