ICEpdf
  1. ICEpdf
  2. PDF-459

Cannot open certain pdf files with ICEpdf 4.3.2 and 4.3.3 release

          Details

          • Type: Bug Bug
          • Status: Closed
          • Priority: Major Major
          • Resolution: Fixed
          • Affects Version/s: 4.3.2, 4.3.3
          • Fix Version/s: 4.3.4
          • Component/s: Core/Parsing
          • Labels:
            None
          • Environment:
            Tomcat 6.0

            Description

            I am trying to use icepdf-core (4.3.2 and 4.3.3 release) to render pdfs.
            These pdf files are generated using Birt.and I am unable to open these files using ICEpdf.

            Also, I could not open such pdfs with ICEpdf demo application at http://www.icesoft.org/demos/icepdf-demos.jsf using the JNLP version - http://res.icesoft.org/demo/jws/icepdf.jnlp.
            Sample PDF file is attached. Sample causing issue.pdf

              Issue Links

              depends on

              Bug - A problem which impairs or prevents the functions of the product. PDF-422 Cannot open certain pdf files with ICEpdf 4.3.2 release

              • Major - Major loss of function.
              • Closed

                Activity

                Ascending order - Click to sort in descending order
                Hide
                Permalink
                Yogesh Manware added a comment -

                PDF-422 does not resolve this issue.

                Show
                Yogesh Manware added a comment - PDF-422 does not resolve this issue.
                Hide
                Permalink
                Patrick Corless added a comment -

                Interesting corner case when parsing the content stream. The PDF has a text block setup as follows:

                BT
                /F1 10 Tf
                1 0 0 1 18.3 41.95 cm
                1 0 0 1 0 0 Tm
                (.......)Tj
                ET

                Normally a 'cm' is not allowed in a text block but already have support for this case in the content parser. This case is slightly different because of the Tm that immediately follows the cm effectively muting the the cm matrix. I altered the content parser so that we update the graphic state as per the rules of cm regardless of being in a text block or not. The fix is actually in the fact that we updated the textBlockBase matrix which is normally only set when the BT is encountered.

                Any ways long story short the fix appears to address the issue but is fairly high risk change. Closing but may have to reopen if QA blows up.

                Show
                Patrick Corless added a comment - Interesting corner case when parsing the content stream. The PDF has a text block setup as follows: BT /F1 10 Tf 1 0 0 1 18.3 41.95 cm 1 0 0 1 0 0 Tm (.......)Tj ET Normally a 'cm' is not allowed in a text block but already have support for this case in the content parser. This case is slightly different because of the Tm that immediately follows the cm effectively muting the the cm matrix. I altered the content parser so that we update the graphic state as per the rules of cm regardless of being in a text block or not. The fix is actually in the fact that we updated the textBlockBase matrix which is normally only set when the BT is encountered. Any ways long story short the fix appears to address the issue but is fairly high risk change. Closing but may have to reopen if QA blows up.

                  People

                  • Assignee:
                    Patrick Corless
                    Reporter:
                    Yogesh Manware
                  • Votes:
                    1 Vote for this issue
                    Watchers:
                    2 Start watching this issue

                    Dates

                    • Created:
                      Updated:
                      Resolved: