gimp-help-2 r2590 - in branches/xml2po-support: . tools

[ulfehlert svn gnome org]

Author: ulfehlert
Date: Fri Oct 31 20:57:56 2008
New Revision: 2590
URL: http://svn.gnome.org/viewvc/gimp-help-2?rev=2590&view=rev

Log:
2008-10-31  Ulf-D. Ehlert  <ulfehlert svn gnome org>

	* tools/xml2po_for_gimp-help.patch1
	* tools/xml2po: fixed output of "tools/xml2po --version"

	* tools/migrate.sh: more minor changes

	* tools/split_xml_multi_lang.py: don't skip 'xi:include' elements
	any more; a few more comments


Modified:
   branches/xml2po-support/ChangeLog
   branches/xml2po-support/tools/migrate.sh
   branches/xml2po-support/tools/split_xml_multi_lang.py
   branches/xml2po-support/tools/xml2po
   branches/xml2po-support/tools/xml2po_for_gimp-help.patch1

Modified: branches/xml2po-support/tools/migrate.sh
==============================================================================
--- branches/xml2po-support/tools/migrate.sh	(original)
+++ branches/xml2po-support/tools/migrate.sh	Fri Oct 31 20:57:56 2008
@@ -2,25 +2,26 @@
 # this script migrates the content from DocBook XML files to PO/GETTEXT
 # supported XML files ??!?!
 
+# XXX: what about fi?
 #LINGUAS="de en es fr it ko nl no pl ru sv"
-LINGUAS="de fr"
+: ${LINGUAS:="de fr"}
 
-srcdir="src"
-oldsrcdir="oldsrc"
-xmldir="xml"
-potdir="pot"
-podir="po"
+: ${srcdir:=src}
+: ${oldsrcdir:=oldsrc}
+: ${xmldir:=xml}
+: ${potdir:=pot}
+: ${podir:=po}
 
 SPLIT="python tools/split_xml_multi_lang.py"
 XML2PO="python tools/xml2po"	# patched version!!
 
-exclude_patterns='.svn key-reference* glossary dictionary'
+: ${exclude_patterns:='.svn key-reference* glossary dictionary'}
 exclude=$(echo "$exclude_patterns" | \
           sed -e 's/[^ ]\+/"&"/g; s/ / -o -name /g; s/^/\\( -name /; s/$/ \\) -prune /')
 
 # clean-up
 if [ -d "$oldsrcdir" ]; then
-    echo >&2 Removing $srcdir, $potdir, $podir, $xmldir, ...
+    echo >&2 Removing $srcdir, $potdir, $podir, $xmldir ...
     test -L $xmldir/en && rm $xmldir/en
     test -d $srcdir && rm -rf $srcdir
     test -d $potdir && rm -rf $potdir
@@ -29,21 +30,28 @@
     mv $oldsrcdir $srcdir
 fi
 
-echo "Splitting the source XML"
+echo >&2 "Splitting the source XML"
+echo >&2 "Warning: the following files and directories will be skipped:"
+echo "$exclude_patterns" | sed -e 's/ /, /g; s/^/    /' >&2
 time \
 eval find $srcdir $exclude -o -name '*.xml' -print |
 while read srcfile
 do
     base=${srcfile%/*}
-    base=${base#$srcdir/}
+    base=${base#$srcdir}
     $SPLIT --lang="$LINGUAS" --file="$srcfile" \
-           --dest="$xmldir"/'*'/"$base"/
+           --dest="$xmldir"/'*'/"${base#/}"/
 done
 echo
 
-mv -vi "$srcdir" "$oldsrcdir" && mv -vi "$xmldir"/en "$srcdir"
+echo >&2 Moving $srcdir and $xmldir ...
+mv -vi "$srcdir" "$oldsrcdir" && \
+mv -vi "$xmldir"/en "$srcdir" && \
+echo >&2 Creating "$xmldir"/en link to "$srcdir" ...
+ln -vs $PWD/"$srcdir" "$xmldir"/en
 echo
 
+test "$1" = "split" && exit 0
 
 echo "Creating POT files"
 time \
@@ -60,6 +68,7 @@
 done
 echo
 
+test "$1" = "pot" && exit 0
 
 echo "Creating PO files"
 time \
@@ -81,3 +90,11 @@
     done
 done
 echo
+
+echo Simple check: searching for empty files...
+trap "rm -f 'empty files'" HUP INT QUIT PIPE TERM
+find ${podir} ${potdir} -type f -size 0 | sort | tee "empty files"
+echo $(wc -l "empty files") found. && rm -f "empty files"
+echo
+
+

Modified: branches/xml2po-support/tools/split_xml_multi_lang.py
==============================================================================
--- branches/xml2po-support/tools/split_xml_multi_lang.py	(original)
+++ branches/xml2po-support/tools/split_xml_multi_lang.py	Fri Oct 31 20:57:56 2008
@@ -120,8 +120,7 @@
 
         This method is used to skip comments etc.
         """
-        return self._node.nodeType in (xml.dom.minidom.Node.COMMENT_NODE,) \
-            or self._node.nodeName in ('xi:include',)
+        return self._node.nodeType in (xml.dom.minidom.Node.COMMENT_NODE,)
 # <<<<<<<<<<<<<<<< not used <<<<<<<<<<<<<<<<
 
 
@@ -269,11 +268,16 @@
                     for lang in self.languages:
                         dest[lang].appendChild(child.cloneNode(NONRECURSIVE))
 
-            # (3) skip every non-English element
+            # (3) add xi:include elements
+            elif child.nodeName == 'xi:include':
+                for lang in self.languages:
+                    dest[lang].appendChild(child.cloneNode(NONRECURSIVE))
+
+            # (4) skip every non-English element
             elif self.skip(child):
                 pass
 
-            # (4) at last, handle non-trivial cases...
+            # (5) at last, handle non-trivial cases...
             else:
                 assert child.nodeType == child.ELEMENT_NODE \
                    and 'en' in self.get_langs(child)
@@ -284,13 +288,13 @@
                 # element.
                 copies = self.vectorize(child, parentnodes)
 
-                # (4a) append recursively localized clones of nodes we don't
+                # (5a) append recursively localized clones of nodes we don't
                 # need/want to process any further (para, phrase, etc.)
                 if self.final(child):
                     self.logger.debug("split(%s) --> adding %s (final)" % \
                             (elem.nodeName, child.nodeName))
                     clones = self.append_clones(copies, dest, RECURSIVE)
-                # (4b) append non-recursively localized clones of nodes and
+                # (5b) append non-recursively localized clones of nodes and
                 # then traverse child nodes recursively (sect[1-4], note, etc.)
                 else:
                     self.logger.debug("split(%s) --> adding %s" % \
@@ -333,24 +337,39 @@
         # parent node.
         if not parents:
 
-            # copy element for every element language
+            # Algorithm:
+            #   (1) for every language defined by the element's "lang"
+            #       attribute select the element itself
+            #   (2) create set of all sibling elements of the same type/name
+            #       without a "seqnum" attribute (which indicates that this
+            #       sibling has already been used)
+            #   (3a) if element is final: select the first matching element
+            #        for every remaining language 
+            #   (3b) if element is not final: same as (3a), but stop on the
+            #        first sibling with a language that was already processed
+            #   (4) select the element itself for every missing language
+
             nodes = dict([(lang, elem) for lang in self.get_langs(elem)])
             assert nodes.has_key('en')
             assert len(nodes) != len(self.languages)
 
-            # TODO: describe algorithm
-
-            siblings = self.get_siblings(elem)
             found = 0
-
-            for sibl in siblings:
+            for sibl in self.get_siblings(elem):
                 sibl_langs = self.get_langs(sibl)
                 new_langs = [k for k in sibl_langs if k not in nodes]
+                # for non-final nodes (which will be processed recursively)
+                # it's important that we stop on any language already "seen",
+                # otherwise different sections will be mixed-up, .e.g.
+                #     <section lang="en"> ... <section lang="en;xx;yy">
                 if not self.final(elem):
                     if len(sibl_langs) > len(new_langs):
                         break
                 elif not new_langs:
+                    # here we just skip any non-matching sibling, e.g.
+                    #     <para lang="en"> ... <para lang="en"> ...
+                    #     <para lang="xx"> ... <para lang="xx"> ...
                     continue
+                # mark as "seen"
                 sibl.setAttribute("seqnum", str(self.seqnum))
                 for lang in new_langs:
                     nodes[lang] = sibl
@@ -496,8 +515,7 @@
 
         This method is used to skip comments etc.
         """
-        return node.nodeType in (node.COMMENT_NODE,) \
-            or node.nodeName in ('xi:include',)
+        return node.nodeType in (node.COMMENT_NODE,)
 
     def text(self, node):
         """Whether or not a node is a text node."""

Modified: branches/xml2po-support/tools/xml2po
==============================================================================
--- branches/xml2po-support/tools/xml2po	(original)
+++ branches/xml2po-support/tools/xml2po	Fri Oct 31 20:57:56 2008
@@ -777,7 +777,7 @@
     elif opt in ('-o', '--output'):
         output = arg
     elif opt in ('-v', '--version'):
-        print VERSION
+        print VERSION + " (patched by GIMP Documentation Team 2008-10-23)"
         sys.exit(0)
     elif opt in ('-h', '--help'):
     	usage(True)

Modified: branches/xml2po-support/tools/xml2po_for_gimp-help.patch1
==============================================================================
--- branches/xml2po-support/tools/xml2po_for_gimp-help.patch1	(original)
+++ branches/xml2po-support/tools/xml2po_for_gimp-help.patch1	Fri Oct 31 20:57:56 2008
@@ -1,5 +1,5 @@
---- xml2po	2008-10-11 14:10:16.000000000 +0200
-+++ xml2po	2008-10-23 20:00:00.000000000 +0200
+--- tools/xml2po	2008-10-23 20:00:00.000000000 +0200
++++ tools/xml2po	2008-10-31 21:00:00.000000000 +0100
 @@ -430,39 +430,30 @@
      return final
  
@@ -68,3 +68,12 @@
              sys.exit(0)
          elif opt in ('-h', '--help'):
              sys.stderr.write("Error: If you want help, please use `%s --help' without '-u' option.\n" % (allargs[0]))
+@@ -786,7 +777,7 @@
+     elif opt in ('-o', '--output'):
+         output = arg
+     elif opt in ('-v', '--version'):
+-        print VERSION
++        print VERSION + " (patched by GIMP Documentation Team 2008-10-23)"
+         sys.exit(0)
+     elif opt in ('-h', '--help'):
+     	usage(True)