Changeset 28078 for other-projects/nightly-tasks

Timestamp:

2013-08-16T19:31:19+12:00 (11 years ago)

Author:

ak19

Message:

diffcol modifications for getting the METS tutorial collection to work. The docmets.xml files also used (oai)lastmodified timestamps, so these needed to be ignored as well.

File:

• other-projects/nightly-tasks/diffcol/trunk/diffcol/diffcol.pl (modified) (3 diffs)

other-projects/nightly-tasks/diffcol/trunk/diffcol/diffcol.pl

@@ -626 +626 @@
     else
     {
-        my $ignore_line_re = "<Metadata name=\"(lastmodified|lastmodifieddate|oailastmodified|oailastmodifieddate|ex.File.FileModifyDate|ex.File.FilePermissions|ImageSize|FileSize)\">.*</Metadata>\\s*\\n*";
+        # allow for a namespace prefix to <Metadata> as happens in GreenstoneMETS docmets.xml files, e.g. <gsdl3:Metadata></gsdl3:Metadata>
+        my $ignore_line_re = "<(.*?:)?Metadata name=\"(lastmodified|lastmodifieddate|oailastmodified|oailastmodifieddate|ex.File.FileModifyDate|ex.File.FilePermissions|ImageSize|FileSize)\">.*</(.*?:)?Metadata>\\s*\\n*";
         
         my $strResult;
 
-        # for doc.xml files, need to ignore many date fields. Filter these out before diffing,
+        # for doc.xml and docmets.xml files, need to ignore many date fields. Filter these out before diffing,
         # in case these don't appear in the same order between collections, since
         # diffutil::GenerateOutput only handles the ignore_regex after a diff has been done
         # when they can show up as unnecessary diff 'errors'
-        if($strModel =~ m/doc\.xml$/) {
-        my ($model_contents, $test_contents);
+
+        my ($model_contents, $test_contents);
+
+        if($strModel =~ m/doc(mets)?\.xml$/) {
+
         open(FIN,"<$strModel") or die "Unable to open $strModel...ERROR: $!\n";
         sysread(FIN, $model_contents, -s FIN);
@@ -645 +649 @@
         $model_contents =~ s/$ignore_line_re//g;
         $test_contents =~ s/$ignore_line_re//g;
-
-
+        
+
+        # doc.xml needs to additionally be normalised, before comparing a windows test with a linux model or vice-versa
+        if($strModel =~ m/doc\.xml$/) {
             # equalise/normalise the two doc.xml files for OS differences, if there are any
             my $testIsWin = &isDocXMLFileWindows($test_contents);
@@ -731 +737 @@
 #       }
         
+        } # finished special processing of doc.xml files
+        
+        # now can diff the normalised versions of the doc.xml/docmets.xml files:
         $strResult = diff \$model_contents, \$test_contents, { STYLE => "OldStyle" };