diff options
author | Peter von Kaehne <refdoc@gmx.net> | 2009-12-30 00:00:14 +0000 |
---|---|---|
committer | Peter von Kaehne <refdoc@gmx.net> | 2009-12-30 00:00:14 +0000 |
commit | 03005158d9a910699256e9cec660af2a5a2ebf2c (patch) | |
tree | ff77dc06032f3ff8cf6a0fa1220668526473b2f3 | |
parent | c8deb706d6fbdf34f663e7e71d628e6b5d1f31ee (diff) | |
download | sword-tools-03005158d9a910699256e9cec660af2a5a2ebf2c.tar.gz |
dictionary transformation added. Needs still handling of verse references
git-svn-id: https://www.crosswire.org/svn/sword-tools/trunk@261 07627401-56e2-0310-80f4-f8cd0041bdcd
-rwxr-xr-x | modules/geneve/wb.pl | 70 |
1 files changed, 70 insertions, 0 deletions
diff --git a/modules/geneve/wb.pl b/modules/geneve/wb.pl new file mode 100755 index 0000000..da8bfc2 --- /dev/null +++ b/modules/geneve/wb.pl @@ -0,0 +1,70 @@ +#!/usr/bin/perl + +my @files=`ls -1 *.abw`; + +foreach (@files){ + + my @lines; + chop; + open USFM, ">>$_.tei"; + @lines=`cat $_`; + + + foreach (@lines) { + s/props\=\".*?\"//g; + + # entry (ordinary) + s/<p\ style=\"T3\".*?><c.*?>(.*?)<\/c><c.*?><\/c><\/p>/<\/entryFree>\n<entryFree\ sortKey=\"$1\"\ split=\"$1\"><form\ type=\"headword\"><orth\ rend\=\"bold\">$1<\/orth><\/form>/g; + + # entry (referring somewhere else) + s/<p\ style=\"T3\".*?><c.*?>(.*?)<\/c><c.*?><\/c><c.*?>s\.(.*?)<\/c><\/p>/<\/entryFree>\n<entryFree\ sortKey=\"$1\"\ split=\"$1\"><form\ type=\"headword\"><orth\ rend\=\"bold\">$1<\/orth><\/form>\n\t<ref target=\"$2\">s\.\ $2<\/ref>/g; + + # content + s/<p.*?>(.*?)<\/p>/\t<def>\n\t\t$1\n\t<\/def>/; + s/<c\ style=\"Kursiv ZF\"\ >(.*?)<\/c>/<hi rend=\"italic\">$1<\/hi>/g; + + + # clean up + s/<c.*?>//g; + s/<\/c\s*>//g; + + } + print (USFM "<?xml\ version=\"1.0\"\ encoding=\"utf-8\"?>\n<TEI\ xmlns=\"http://www.crosswire.org/2008/TEIOSIS/namespace\"\nxmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\" + xsi:schemaLocation=\"http://www.crosswire.org/2008/TEIOSIS/namespace + http://www.crosswire.org/OSIS/teiP5osis.1.4.xsd\"> + + <teiHeader> + <fileDesc> + <titleStmt> + <title>Worterklaerung Schlachter</title> + <author></author> + </titleStmt> + + <editionStmt> + <edition></edition> + </editionStmt> + + <publicationStmt> + <publisher></publisher> + <date></date> + </publicationStmt> + + <sourceDesc><p>http://sourceforge.net/project/showfiles.php?group_id=89078&package_id=93370&release_id=278981</p></sourceDesc> + </fileDesc> + + <revisionDesc> + <change when=\"2008-11-25\">initial conversion to TEI</change> + </revisionDesc> + </teiHeader> + + <text> + <body>"); + + + + + print (USFM @lines,"\n"); + + print (USFM "</body>\n</text>\n</TEI>"); + close USFM; +} |