Changeset 32284 for gs2-extensions/pdf-box/trunk/java/perllib/plugins/PDFBoxConverter.pm

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.

File:

• gs2-extensions/pdf-box/trunk/java/perllib/plugins/PDFBoxConverter.pm (modified) (4 diffs)

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

@@ -129 +129 @@
     $self->{'pdfbox_txt_launch_cmd'} = "$java -cp \"$pbajar\" org.apache.pdfbox.tools.ExtractText";
     $self->{'pdfbox_html_launch_cmd'} = "$java -cp \"$pbajar\" -Dline.separator=\"<br />\" org.apache.pdfbox.tools.ExtractText";
-#   $self->{'pdfbox_img_launch_cmd'} = "java -cp \"$pbajar\" org.apache.pdfbox.tools.PDFToImage"; # pdfbox 2.09 cmd for converting each PDF page to an image (gif, jpg, png)
+#   $self->{'pdfbox_img_launch_cmd'} = "java -cp \"$pbajar\" org.apache.pdfbox.tools.PDFToImage"; # pdfbox 2.09 cmd for converting each PDF page to an image (jpg, png)
 
     # We use this next cmd to launch our new custom PDFBox class (PDFBoxToImagesAndText.java) to convert each PDF page into an image (gif, jpg, png)
@@ -179 +179 @@
     # and ends up going through gsConvert.pl
     $self->{'convert_options'} .= " -pdf_tool pdfbox";
-    
+
     my $img_output_mode = 0;
     
@@ -192 +192 @@
     if ($target_file_type eq "html") {
     $self->{'converted_to'} = "HTML";
-    } elsif ($target_file_type eq "jpg" || $target_file_type eq "gif" || $target_file_type eq "png") { 
+    } elsif ($target_file_type eq "jpg" || $target_file_type eq "png") { # || $target_file_type eq "gif"
     # GIF not supported by PDFBox at present, see https://pdfbox.apache.org/1.8/commandline.html#pdftoimage
     $self->{'converted_to'} = $target_file_type;    
@@ -248 +248 @@
     if($img_output_mode || $paged_txt_output_mode) { # converting each page to image and/or text
     my $output_prefix = &FileUtils::filenameConcatenate($target_file_path, $tailname);
-    
+
+    # Our custom class does renaming of the pages (simplified to just numbers) for PagedImagePlugin
     #$convert_cmd = $paged_txt_output_mode ? $self->{'pdfbox_imgtxt_launch_cmd'} : $self->{'pdfbox_img_launch_cmd'};
     $convert_cmd = $self->{'pdfbox_imgtxt_launch_cmd'};
-    $convert_cmd .= " -textOnly" unless($img_output_mode); # if paged txt only and no images
     $convert_cmd .= " -imagesOnly" unless($paged_txt_output_mode); # set to images only unless there's text too
-    $convert_cmd .= " -imageType $target_file_type" if($img_output_mode);
+    if($img_output_mode) { # whether images-only or images-and-text mode
+        $convert_cmd .= " -imageType $target_file_type";
+        $convert_cmd .= " -dpi ". $self->{"dpi"} if defined $self->{"dpi"};
+    } else { # img_output_mode off, so paged txt only and no images
+        $convert_cmd .= " -textOnly";       
+    }
     $convert_cmd .= " -outputPrefix \"$output_prefix\"";
     $convert_cmd .= " \"$source_file_full_path\"";