diff -r 000000000000 -r ae7b3fa753dc sf2bz.sh
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/sf2bz.sh	Thu Nov 16 08:30:26 2006 +0000
@@ -0,0 +1,109 @@
+#!/bin/sh
+if [ $# -ne 2 ]; then
+    echo "Usage: sf2bz <sf-project-export>.xml <bz-import>.xml" >&2
+    exit 1
+fi
+input=$1
+output=$2
+topdir=/home/ali/wk/slashem/web.scripts
+cd $topdir
+cache_dir=$topdir/cache
+tagsoup=tagsoup/tagsoup-1.0.1.jar
+#tmpdir=`mktemp -d /tmp/XXXXXXXX` || exit 1
+tmpdir=/tmp/sf2bz; rm -rf $tmpdir; mkdir $tmpdir
+mkdir -p $cache_dir/sf/artifacts
+# Sourceforge's export is pretty good, but misses a few things.
+# Deal with one of these (descriptions for attached files) here.
+echo '<?xml version="1.0"?>' > $tmpdir/descriptions.xml
+echo '<artifacts xmlns:xi="http://www.w3.org/2001/XInclude">' \
+  >> $tmpdir/descriptions.xml
+./get_attached_files $input | cut -s "-d " -f 1,2 | sort -n | uniq | \
+  while read atid aid; do
+    mkdir -p $cache_dir/sf/artifacts/$atid
+    if [ ! -r $cache_dir/sf/artifacts/$atid/$aid.html ]; then
+	wget -q -O $cache_dir/sf/artifacts/$atid/$aid.html \
+	  "http://sourceforge.net/tracker/index.php?func=detail&aid=$aid&group_id=9746&atid=$atid"
+    fi
+    mkdir -p $cache_dir/artifacts/$atid
+    if [ ! -r $cache_dir/artifacts/$atid/$aid.xhtml ]; then
+	java -jar $tagsoup $cache_dir/sf/artifacts/$atid/$aid.html \
+	  > $cache_dir/artifacts/$atid/$aid.xhtml
+    fi
+    echo "<xi:include href=\"$cache_dir/artifacts/$atid/$aid.xhtml\"/>" >> $tmpdir/descriptions.xml
+done
+echo '</artifacts>' >> $tmpdir/descriptions.xml
+xsltproc --xinclude attachment_descs.xsl $tmpdir/descriptions.xml \
+  > $tmpdir/attachment-descriptions.xml
+# Some details of attachments are included in the project export.
+xsltproc attachment_details.xsl $input > $tmpdir/attachment-details.xml
+# Then convert the attachments to XML (as required by bugzilla)
+# pulling in the various bits of information we have gathered.
+mkdir -p $cache_dir/sf/attachments $cache_dir/attachments
+mkdir -p $tmpdir/jail
+./get_attached_files artifacts.xml | while read atid aid file_id; do
+    if [ ! -r $cache_dir/sf/attachments/$file_id ]; then
+	wget -N -O $cache_dir/sf/attachments/$file_id "http://sourceforge.net/tracker/download.php?group_id=9746&atid=$atid&file_id=$file_id&aid=$aid"
+    fi
+    if [ ! -r $cache_dir/attachments/$file_id.xml ]; then
+	type=`file -bi $cache_dir/sf/attachments/$file_id | sed 's/;.*//'`
+	case "$type" in
+	    text/*)
+		(cd $tmpdir/jail; patch --dry-run -s -f \
+		  < $cache_dir/sf/attachments/$file_id > /dev/null 2>&1)
+		if [ $? -lt 2 ]; then
+		    pflag="-p"
+		else
+		    pflag=""
+		fi
+		;;
+	    *)
+	        pflag=""
+		;;
+	esac
+	./xml_attached_file $pflag -i $file_id -t $type \
+	  $cache_dir/sf/attachments/$file_id > $tmpdir/$file_id.1.xml
+	xsltproc attachment_gather.xsl $tmpdir/$file_id.1.xml \
+	  > $tmpdir/$file_id.2.xml
+	xsltproc --xinclude attachment_post.xsl $tmpdir/$file_id.2.xml \
+	  > $cache_dir/attachments/$file_id.xml
+    fi
+done
+# We can't get hold of descriptions of deleted attachments, so we
+# keep a copy of the file in case we need it. Ideally, we'd make sure
+# that we only add descriptions, but that's for another day.
+cp $tmpdir/attachment-descriptions.xml $topdir
+xsltproc get_users.xsl $input > $tmpdir/raw_users.xml
+echo '<artifacts xmlns:xi="http://www.w3.org/2001/XInclude">' \
+  >> $tmpdir/users.1.xml
+for file in $cache_dir/artifacts/*/*.xhtml; do
+    echo "<xi:include href=\"$file\"/>" >> $tmpdir/users.1.xml
+done
+echo "<xi:include href=\"raw_users.xml\"/>" >> $tmpdir/users.1.xml
+echo '</artifacts>' >> $tmpdir/users.1.xml
+xsltproc --xinclude user_names.xsl $tmpdir/users.1.xml > $tmpdir/users.2.xml
+mkdir -p $cache_dir/sf/users $cache_dir/users
+echo '<sf_users xmlns:xi="http://www.w3.org/2001/XInclude">' \
+  >> $tmpdir/users.3.xml
+xsltproc unknown_users.xsl $tmpdir/users.2.xml | while read user; do
+    case $user in
+	*@users.sourceforge.net)
+	    user_id=`echo $user | sed 's/@users.sourceforge.net$//'`
+	    if [ ! -r $cache_dir/sf/users/$user_id.html ]; then
+		wget -N -O $cache_dir/sf/users/$user_id.html \
+		  "http://sourceforge.net/users/$user_id/"
+	    fi
+	    if [ ! -r $cache_dir/users/$user_id.xhtml ]; then
+		java -jar $tagsoup $cache_dir/sf/users/$user_id.html \
+		  > $cache_dir/users/$user_id.xhtml
+	    fi
+	    echo "<xi:include href=\"$cache_dir/users/$user_id.xhtml\"/>" \
+	      >> $tmpdir/users.3.xml
+    esac
+done
+echo "<xi:include href=\"users.2.xml\"/>" >> $tmpdir/users.3.xml
+echo '</sf_users>' >> $tmpdir/users.3.xml
+xsltproc --xinclude user_names.xsl $tmpdir/users.3.xml > users.xml
+ln -s $cache_dir/attachments $tmpdir
+xsltproc bugzilla.xsl $input > artifacts.tmp.xml
+xsltproc --xinclude bugzilla_post.xsl artifacts.tmp.xml > $output
+#rm -rf $tmpdir artifacts.tmp.xml