Changeset 32283 for main/trunk

Timestamp:

2018-07-17T22:19:34+12:00 (6 years ago)

Author:

ak19

Message:

More stable behaviour by PDFv2Plugin: 1. when pdfbox_conversion is on, but an output option supported by xpdftools is selected, it now uses xpdftools anyway instead of attempting to use pdfbox_conversion. 2. when pdfbox_conversion is not on and an output format that it alone supports (and not supported by xpdftools) is selected, a warning message is displayed that xpdftools will be used to output to a fallback output format and that the user to switch on pdfbox_conversion otherwise. This message was present and displayed in a recent commit, but the behaviour was not set up yet then. In future, we may solve this differently if it's decided that PDFBoxConverter is not an AutoLoadConverter and will therefore be always available with PDFv2Plugin (but what about GS2, where PDFBox is an optional extension?)

Location:

main/trunk/greenstone2/perllib

Files:

• plugins/PDFv2Plugin.pm (modified) (5 diffs)
• strings.properties (modified) (1 diff)

main/trunk/greenstone2/perllib/plugins/PDFv2Plugin.pm

@@ -109 +109 @@
        'deft' => "1",
        'type' => "string" }, # xpdftools' zoom takes fractions
-     { 'name' => "use_sections",
-       'desc' => "{PDFPlugin.use_sections}",
-       'type' => "flag" },
-     { 'name' => "description_tags",
-       'desc' => "{HTMLPlugin.description_tags}",
-       'type' => "flag" },
+#     { 'name' => "use_sections",
+#       'desc' => "{PDFPlugin.use_sections}",
+#       'type' => "flag" },
+#     { 'name' => "description_tags",
+#       'desc' => "{HTMLPlugin.description_tags}",
+#       'type' => "flag" },
       { 'name' => "use_realistic_book",
         'desc' => "{PDFPlugin.use_realistic_book}",
@@ -159 +159 @@
     $self->{'convert_options'} .= " -pdf_zoom $zoom";
 
-    # check convert_to
-    
-    # Not all available conversion output options are possible with xpdftools, as some are
-    # only handled by pdfbox. If a format is unavailable with xpdftools, default to pretty_html
-    if ($self->{'convert_to'} =~ /^html$/) {
-    &gsprintf::gsprintf(STDERR, "{PDFv2Plugin.conversion_needs_pdfbox}\n", ($self->{'convert_to'}, "pretty_html"));
-#   $self->{'convert_to'} = "pretty_html";
-    }
-    elsif ($self->{'convert_to'} =~ /^pagedimg/) {
-    &gsprintf::gsprintf(STDERR, "{PDFv2Plugin.conversion_needs_pdfbox}\n", ($self->{'convert_to'}, "paged_pretty_html"));
-#   $self->{'convert_to'} = "paged_pretty_html";
-    }
-    elsif ($self->{'convert_to'} =~ /^paged_text$/) {
-    # TODO
-    print STDERR "@@@ Conversion to " . $self->{'convert_to'} , " with Xpdf Tools is not yet implemented.\n";
-    #print STDERR "@@@ Converting to text instead.\n";
-    #$self->{'convert_to'} = "text";
-    }
-    
     # PDFv2Plugin now supports PDF to txt conversion on Windows too:
     # using XPDF Tools (incl pdftotext) on Windows/Linux/Mac
-    elsif ($self->{'convert_to'} eq "text" && $ENV{'GSDLOS'} =~ /^windows$/i) {
+    if ($self->{'convert_to'} eq "text" && $ENV{'GSDLOS'} =~ /^windows$/i) {
     &gsprintf::gsprintf(STDERR, "{PDFv2Plugin.win_pdftotext_info}\n");
     }
@@ -193 +174 @@
     }
     }
+
+    # if pdfbox_conversion is not on, check convert_to to make sure that xpdftools can
+    # support the selected output format, or fallback on a sensible default
+    # Not all available conversion output options are possible with xpdftools, as some are
+    # only handled by pdfbox. If a format is unavailable with xpdftools, default to pretty_html
+    if (!$self->{"pdfbox_conversion"}) {
+    my $convert_to = $self->{'convert_to'};
+    my $fallback_convert_to = $convert_to;
+    if($convert_to =~ /^html$/) {
+        $fallback_convert_to = "pretty_html";
+    }
+    elsif ($self->{'convert_to'} =~ /^pagedimg/) {
+        $fallback_convert_to = "paged_pretty_html";
+    }
+    elsif ($self->{'convert_to'} =~ /^paged_text$/) {
+        #   print STDERR "@@@ Conversion to " . $self->{'convert_to'} , " with Xpdf Tools is not yet implemented.\n";
+        $fallback_convert_to = "text";
+    }
+
+    if($convert_to =~ /^(html|pagedimg|paged_text)/) {
+        &gsprintf::gsprintf(STDERR, "{PDFv2Plugin.conversion_needs_pdfbox}\n", ($self->{'convert_to'}, $fallback_convert_to));
+        $self->{'convert_to'} = $fallback_convert_to;
+    }
+    }
+    
     # set convert_to_plugin and convert_to_ext
     $self->set_standard_convert_settings();
@@ -310 +316 @@
 
     my $self = shift (@_);
+    
+    if($self->{'convert_to'} =~ m/pretty_html$/) { # if outputting paged_pretty_html or pretty_html:
+    # only xpdftools can output pretty_html regardless of whether pdfbox_conversion is switched on
+    print STDERR "@@@@ PDFBox_conversion is switched on, but pretty_html variants are generated by xpdftools.\n";
+    return $self->ConvertBinaryFile::tmp_area_convert_file(@_);
+    }
+    # else, output format uses pdfbox:
     return $self->AutoLoadConverters::tmp_area_convert_file(@_);
 
@@ -319 +332 @@
     my ($tmp_dirname, $tmp_inputPDFname, $utf8_tailname, $lc_suffix, $tailname, $suffix) = @_;    
 
-    if($self->{'convert_to'} !~ /pretty_html/) {
+    if($self->{'convert_to'} !~ m/pretty_html$/) {
     return $self->ConvertBinaryFile::run_conversion_command(@_);
     }

main/trunk/greenstone2/perllib/strings.properties

@@ -1199 +1199 @@
 PDFv2Plugin.win_pdftotext_info:PDFv2Plugin uses Xpdf Tools to support pdf to text conversion, including on Windows.
 
-PDFv2Plugin.conversion_needs_pdfbox:*** Conversion to %s not supported with Xpdf Tools, defaulting to %s. Turn on pdfbox_conversion if you wish to enable output to selected format.
+PDFv2Plugin.conversion_needs_pdfbox:*** Conversion to %s not supported with Xpdf Tools, defaulting to %s.\nTurn on pdfbox_conversion if you wish to enable output to the selected format.
 
 PostScriptPlugin.desc:This is a \"poor man's\" ps to text converter. If you are serious, consider using the PRESCRIPT package, which is available for download at http://www.nzdl.org/html/software.html