Changeset 32090

Timestamp:

2017-12-08T20:16:36+13:00 (6 years ago)

Author:

ak19

Message:

Related to the previous commit, revision 32089. Some improvements.

Files:

• gs2-extensions/pdf-box/trunk/java/perllib/plugins/PDFBoxConverter.pm (modified) (1 diff)
• main/trunk/greenstone2/perllib/FileUtils.pm (modified) (3 diffs)

gs2-extensions/pdf-box/trunk/java/perllib/plugins/PDFBoxConverter.pm

@@ -266 +266 @@
 
     my $fulltext = &FileUtils::readUTF8File($target_file_path);
-    #$fulltext = &HTML::Entities::encode($fulltext); # doesn't seem to help
-    $fulltext =~ s@&@&@sg; # Kathy's fix to ensure doc contents don't break XML
-    &FileUtils::writeUTF8File($target_file_path, \$fulltext);
+    if(defined $fulltext) {
+        #$fulltext = &HTML::Entities::encode($fulltext); # doesn't seem to help
+        $fulltext =~ s@&@&@sg; # Kathy's fix to ensure doc contents don't break XML
+        &FileUtils::writeUTF8File($target_file_path, \$fulltext);
+    } else {
+        print STDERR "PDFBoxConverter::convert(): Unable to read from converted file\n";
+        $had_error = 1;
+    }
     }
 

main/trunk/greenstone2/perllib/FileUtils.pm

@@ -973 +973 @@
 ## @function readUTF8File()
 #
-# read contents from a file containing UTF8.
-#
-# Parameter filename, the filepath to read from
+# read contents from a file containing UTF8 using sysread, a fast implementation of file 'slurp'
+#
+# Parameter filename, the filepath to read from.
+# Returns undef if there was any trouble opening the file or reading from it.
 #
 sub readUTF8File
@@ -983 +984 @@
     print STDERR "@@@ Warning FileUtils::readFile() not yet implemented for parallel processing. Using regular version...\n";
     
-    open(FIN,"<$filename") or die "FileUtils::readFile: Unable to open $filename for reading...ERROR: $!\n";
+    #open(FIN,"<$filename") or die "FileUtils::readFile: Unable to open $filename for reading...ERROR: $!\n";
+
+    if(!open(FIN,"<$filename")) {
+    print STDERR "FileUtils::readFile: Unable to open $filename for reading...ERROR: $!\n";
+    return undef;
+    }
 
     # decode the bytes in the file with UTF8 enc,
@@ -989 +995 @@
     binmode(FIN,":utf8");
     
-    my $contents;
+    my $contents = undef;
     # Read in the entire contents of the file in one hit
     sysread(FIN, $contents, -s FIN);