Forum Discussion

hoadoan's avatar
hoadoan
Contributor
14 years ago

how to get the content of a pdf file on window

Hi all,



I have a scenario as following

- Enter a year

- Click on View button

- A pdf file will be displayed in another window



I read some methods and I see that MSAAObject is one of solutions.

1) The MSAA Open Applications plug-in is installed and enabled in TestComplete (refer to MSAA config 1.jpg)

2) Enable MSAA support for AVL_AVView (refer to MSAA config 2.jpg)



I read the blog "http://blog.smartbear.com/post/06-06-01/Use-TestComplete-to-Fill-InfoPath-2003-Forms-Automatically/" and see that after config two above steps, TestComplete’s Object Browser will displayed MSAA object with 


Could anybody help me to explain it ? And it would be great if giving me a script to get the content of pdf file on window.



Thanks so much.

Hoa








  • HKosova's avatar
    HKosova
    SmartBear Alumni (Retired)
    14 years ago

    Hi Hoa,


    For the MSAA engine to expose internal structure of a PDF file, this file must be tagged to enable accessibility. Open Document Properties and check out the Description > Advanced > Tagged PDF value - what does it say?




    If your PDF is tagged, make sure that Internet Explorer isn't filtered out from Open Applications by the Process Filter in your project's properties.




    If your PDF is untagged, you can try one of the following approaches:


    * Add tags to the PDF file to make it MSAA-compatible. If the developers of your tested application could modify it so that it generates tagged PDF, that would be easier for you. If they cannot, you can add the tags yourself by using an authoring application such as Adobe Acrobat.


    * If you have Adobe Acrobat (not Acrobat Reader) installed, you can access the PDF file contents using Acrobat's OLE Automation API (the AcroExch.App scripting object). For the Acrobat Automation API reference, check out this document.


    * Alternatively, you can save the PDF file to a text file and retrieve data from this text file. You can find sample scripts for reading text files here.

  • hoadoan's avatar
    hoadoan
    Contributor
    14 years ago

    Thanks Helen so much. Your information is useful for me



    Actually, my PDF isn't tagged. I will try tagging PDF as your suggestion later



    I found another solution to read PDF which using dll file without changing pdf file to text file. 

    - Add IKVM.GNU.Classpath.dll, PDFBox-0.7.3.dll as references

    - Copy FontBox-0.1.0-dev.dll, IKVM.Runtime.dll into TC’s bin directory






    And write script as following

    var doc=dotNET["org_pdfbox_pdmodel"]["PDDocument"]["load"]("C:\\BroadcastCalendarReport_pr.pdf");    

    var pdfStripper=dotNET["org_pdfbox_util"]["PDFTextStripper"]["zctor"]();    

    var str=pdfStripper["getText"](doc);


    I will get the content of pdf file through str. I hope that it's useful for user to need PDF directly in Test complete 6



    But I still have a problem in TC 8 trial. The message "Invalid procedure call or argument" is always shown at function pdfStripper["getText"](doc)



    Could anybody help me on this?



    Thanks,

    Hoa.










  • HKosova's avatar
    HKosova
    SmartBear Alumni (Retired)
    14 years ago

    Hi Hoa,


    The PDFBox library is a great find, thanks for sharing!


    The message "Invalid procedure call or argument" is always shown at function pdfStripper["getText"](doc)


    I guess, this error happens because your PDF is untagged, as your code works fine for me for tagged PDFs like this one. Or, your PDF might be encrypted - the getText method description says that it doesn't work with encrypted files.
  • hoadoan's avatar
    hoadoan
    Contributor
    14 years ago
    Hi Helen



    It works fine for untagged Pdf in Test complete 6. I don't know why it doesn't work in Test complete 8. It's a strange



    Regards,

    Hoa
  • HKosova's avatar
    HKosova
    SmartBear Alumni (Retired)
    14 years ago
    Hi Hoa,



    There're two implementations of the getText method with different parameter types: one uses org.pdfbox.cos.COSDocument, the other - org.pdfbox.pdmodel.PDDocument. In TestComplete scripts, they are available as getText and getText_2. If your script creates a PDDocument instance, make sure that you're using the getText implementation for PDDocument - on my computer, it's getText_2 (see the attached image).



    - Copy FontBox-0.1.0-dev.dll, IKVM.Runtime.dll into TC’s bin directory


    Try placing these files in the folder with PDFBox-0.7.3.dll. Does this help?
  • hoadoan's avatar
    hoadoan
    Contributor
    14 years ago
    Hi Helen,



    Thanks. Actually, I used getText_2 to read document. That's the reason it works well on TC 6. But it doesn't work TC 8.
  • HKosova's avatar
    HKosova
    SmartBear Alumni (Retired)
    14 years ago

    Hi Hoa,


    Try the following setup:


      * Place all the dlls -- PDFBox-0.7.3.dll, IKVM.Runtime.dll, IKVM.GNU.Classpath.dll and FontBox-0.1.0-dev.dll -- to the same folder.

        DO NOT copy any of these to the <TestComplete 8>\Bin folder (if you've previously copied them there, remove them.)


      * Add only PDFBox-0.7.3.dll to your project's CLR Bridge in TestComplete 8.


    Does it work now?

  • skrishtc's avatar
    skrishtc
    New Contributor
    13 years ago
    Hi,



    we have scenario to check the data from PDF,



    can any one please let me know where can get/download the .dll files (IKVM.GNU.Classpath.dll, PDFBox-0.7.3.dll, FontBox-0.1.0-dev.dll, IKVM.Runtime.dll)



    Bit urgent please



    Thanks in Advance
  • ravisangam's avatar
    ravisangam
    Occasional Contributor
    13 years ago
    Can anyone post the complete details about configuring and about where can we download those dll files