Bug with comments in pdfs

Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic

   Hi,

if there are comments in a pdf then reading data can go wrong:

TPDFStructure.SmartDecode checks for whitespace and comments but it ignores that after a comment there may be additional whitespace:

trailer

<< /ID 

  % ReportLab generated PDF document -- digest (http://www.reportlab.com) 

  [(4\264\331\222\333\020\213K\351^\240\316\000\203\255\301) (4\264\331\222\333\020\213K\351^\240\316\000\203\255\301)] 

 /Info 7 0 R

 /Root 6 0 R

 /Size 18 >>

in this sample the trailer is not read correct as after the first comment and start of ID data there is additional whitespace

Fix:

  procedure SkipWhitespace;

  begin

    // Read all whitespace

    while (X < L) and IsWhitespace(Source[X]) do Inc(X);

    // is next char the start sign for a comment, then ignore entire line

    if (X < L) and (Source[X] = '%') then

    begin

      while (X < L) and (Source[X] <> #10) and (Source[X] <> #13) do Inc(X);

      if (Source[X] = #13) and (Source[X + 1] = #10) then

        Inc(X, 2)

      else

        Inc(X);

      // UKO 2009-04-07 after a comment recheck for white space!

      SkipWhitespace;

    end;

  end;

same should be done on TPDFType1Decoder.Tokenize: also adjust the SkipWhiteSpace procedure