Changeset 2599

Timestamp:

2001-06-20T14:32:24+12:00 (23 years ago)

Author:

jrm21

Message:

we now do some post-processing to fix up words broken by html tags - eg
<b>Wo</b><b>rds</b> would go to mg as "Wo" and "rds", not "Words". It also
makes the html cleaner. We currently only do this for <b> and <i> tags.

File:

• trunk/gsdl/bin/script/pdftohtml.pl (modified) (4 diffs)

trunk/gsdl/bin/script/pdftohtml.pl

@@ -135 +135 @@
     }
 
+# post-process to remove </b><b> and </i><i>, as these break up
+# words, screwing up indexing and searching.
+    &util::mv("$output_filestem.html","$output_filestem.html.tmp");
+    open INFILE, "$output_filestem.html.tmp" ||
+    die "Couldn't open file: $!";
+    open OUTFILE, ">$output_filestem.html" ||
+    die "Couldn't open file for writing: $!";
+    my $line;
+    while ($line=<INFILE>) {
+    $line =~ s#</b><b>##g;
+    $line =~ s#</i><i>##g;
+    print OUTFILE $line;
+    }
+    close INFILE;
+    close OUTFILE;
+    &util::rm("$output_filestem.html.tmp");
+
+
     # Need to convert images from PPM format to PNG format
     my @images;
@@ -154 +172 @@
         if (system($cmd)!=0) {
         print STDERR "Error executing $cmd\n";
-        return 0; # not sure about whether to leave this one in or take it out
+        #return 0; # not sure about whether to leave this one in or take it out
+        next;
         }
     } else {
@@ -165 +184 @@
         if (system($cmd)!=0) {
             print STDERR "Cannot convert $image into PNG format (tried `pnmtopng' and `convert')...\n";
-            return 0; # not sure about whether to leave this one in or take it out
+            #return 0; # not sure about whether to leave this one in or take it out
+            next;
         }
         }
@@ -175 +195 @@
 }
 
-&main(@ARGV);
+# indicate our error status
+if (&main(@ARGV)) {exit 0;}
+exit 1;