Changeset 32284 for main

Timestamp:

2018-07-18T18:45:52+12:00 (6 years ago)

Author:

ak19

Message:

PDFv2Plugin doesn't offer a zoom flag anymore, replaced with a dpi flag to set the resolution.

Location:

main/trunk/greenstone2

Files:

• bin/script/gsConvert.pl (modified) (4 diffs)
• perllib/plugins/PDFv2Plugin.pm (modified) (2 diffs)

main/trunk/greenstone2/bin/script/gsConvert.pl

@@ -65 +65 @@
 my $pdf_nohidden;
 my $pdf_zoom;
+my $pdf_dpi;
 my $pdf_ignore_images;
 my $pdf_allow_images_only;
@@ -90 +91 @@
     print STDERR "\t-pdf_zoom\tfactor by which to zoom PDF (only useful if\n";
     print STDERR "\t\t-pdf_complex is set\n";
+    print STDERR "\t-pdf_dpi\tSet the resolution in DPI of background images produced by xpdf's pdftohtml\n";
     exit(1);
 }
@@ -132 +134 @@
              'pdf_allow_images_only', \$pdf_allow_images_only,
              'pdf_nohidden', \$pdf_nohidden,
-             'pdf_zoom/\d+/2', \$pdf_zoom
+             'pdf_zoom/\d+/2', \$pdf_zoom,
+             'pdf_dpi/\d+/96', \$pdf_dpi
              ))
     {
@@ -899 +902 @@
     # xpdf's pdftohtml tool also takes a zoom factor, where a zoom of 1 is 100%
     $cmd .= "\"$xpdf_pdftohtml\"";
-    $cmd .= " -z $pdf_zoom" if ($pdf_zoom);
-#    $cmd .= " -c" if ($pdf_complex);
-#    $cmd .= " -i" if ($pdf_ignore_images);
-#    $cmd .= " -a" if ($pdf_allow_images_only);
-#    $cmd .= " -hidden" unless ($pdf_nohidden);    
+    # resolution, -r in DPI of background images, see https://www.xpdfreader.com/pdftohtml-man.html
+    $cmd .= " -r $pdf_dpi" if ($pdf_dpi);
     $cmd .= " \"$input_filename\" \"$tmp_dirname\"";
     #$cmd .= " \"$input_filename\" \"$output_filestem\"";

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

@@ -89 +89 @@
        'type' => "string",
        'deft' => "Title,Author,Subject,Keywords" },
-      { 'name' => "metadata_field_separator",
+     { 'name' => "metadata_field_separator",
     'desc' => "{HTMLPlugin.metadata_field_separator}",
     'type' => "string",
     'deft' => "" },
-#     { 'name' => "noimages",
-#       'desc' => "{PDFPlugin.noimages}",
-#       'type' => "flag" },
-#     { 'name' => "allowimagesonly",
-#       'desc' => "{PDFPlugin.allowimagesonly}",
-#       'type' => "flag" },
-#     { 'name' => "complex",
-#       'desc' => "{PDFPlugin.complex}",
-#       'type' => "flag" },
-#     { 'name' => "nohidden",
-#       'desc' => "{PDFPlugin.nohidden}",
-#       'type' => "flag" },
-     { 'name' => "zoom",
-       'desc' => "{PDFv2Plugin.zoom}",
-       'deft' => "1",
-       'type' => "string" }, # xpdftools' zoom takes fractions
+     { 'name' => "dpi",
+       'desc' => "{PDFv2Plugin.dpi}",
+       'deft' => "96",
+       'type' => "int" }, # 72DPI is xpdf's pdftohtml's default. pdfbox' default is 96DPI in headless mode else detected from the screen resolution, see https://pdfbox.apache.org/2.0/commandline.html#pdftoimage
 #     { 'name' => "use_sections",
 #       'desc' => "{PDFPlugin.use_sections}",
@@ -155 +143 @@
     $self->{'convert_options'} = "-pdf_tool xpdftools"; # default for PDFv2Plugin. If pdfbox_conversion is on, the pdfbpox GS extension sets pdf_tool to pdfbox
 
-    # pdf_zoom is supported by xpdftools' pdftohtml. So for pretty_html and paged_pretty_html
-    my $zoom = $self->{"zoom"};
-    $self->{'convert_options'} .= " -pdf_zoom $zoom";
+    # Setting dpi has meaning for xpdftools pdftohtml (so paged_pretty_html and pretty_html)
+    # and for when pdfbox outputs an image for each page (pagedimg, pagedimgtxt).
+    # dpi has no effect on (paged_)text and html output modes.
+    my $dpi = $self->{"dpi"};
+    $self->{'convert_options'} .= " -pdf_dpi $dpi";
 
     # PDFv2Plugin now supports PDF to txt conversion on Windows too: