1. Convert pdf to any other formats using Adobe DC SDK, like txt, xml, doc, docx, jpg, ps, rft etc.
    Notice: To run this project, you need install Adobe Acrobat Pro DC.
    We have not tested other versions of Adobe Acrobat. This link points to the download page of Acrobat Pro DC, you can try it for free for a few days.
  2. Using grobid, friendly to scientific paper text(not support images and tables).

Environment and Dependency

  • Operating system: Windows 10 Professional
  • IDE: Visual Studio 2017 Community
  • Development framework: . Net Framework 4.6.1
  • Dependency:
    • Acrobat
    • Adobe Acrobat 10.0 Type Library
    • System Windows Forms
  • The fist two are COM type libraries, after installing Adobe Acrobat DC, you can add reference in Visual Studio with project manager.

How to build and run it

Build PDFconvert

  1. Create a C# Console project, and choose the . Net framework version.
  2. Add references, click the COM in references manager and select Acrobat and Adobe Acrobat 10.0 Type Library.
  3. To run this project, you need add command parameters in project manager, the input file complete path and output dictionary(optional, if not specify, it will save the output file where the executable file in). You can also use console to run the executable file as follows:
    PDFConvert.exe -i inputfile -o outputdir -r true -t 20000
  1. If you run this repository directly, you can skip step 1 and 2.

Batch process with Python (Controller)

 python BatRun.py -e  C:\Users\~\PDFConvert.exe -i C:\Users\~ -o C:\Users\~ -f xml -t 60


  • python3.5+
  • pip install pypiwin32==223
  • BeautifulSoup4
  • docx
  • xlrd
  • lxml

If you run this python command in cmd console or Powershell of Windows10, you’d be better disable ‘QuickEdit Mode’ and ‘Insert Mode’ every time to avoid the process stuck in suspended state. Simply right click up the top of your powershell console, head down to properties and under Edit options you will see ‘QuickEdit Mode’.



If you want to convert pdf to other formats,
The cConvIDs supported by Acrobat library. The list of cConvIDs are as follows:

cConvID extension comment
com.adobe.acrobat.eps eps Not test
com.adobe.acrobat.html-3-20 html, htm Recommended
com.adobe.acrobat.html-4-01-css-1-00 html, htm Run well
com.adobe.acrobat.jpeg jpeg , jpg, jpe Not test
com.adobe.acrobat.jp2k jpf, jpx, jp2, j2k, j2c, jpc Not test
com.adobe.acrobat.doc doc Run well
com.adobe.acrobat.docx docx Run well
com.callas.preflight.pdfa pdf Not test
com.callas.preflight.pdfx pdf Not test
com.adobe.acrobat.png png Not test
com.adobe.acrobat.ps ps Not test
com.adobe.acrobat.rtf rft Not test
com.adobe.acrobat.accesstext txt Run well
com.adobe.acrobat.plain-text txt Run well
com.adobe.acrobat.tiff tiff, tif Not test
com.adobe.acrobat.xml-1-00 xml Recommended
com.adobe.acrobat.xlsx xlsx Run well

In Acrobat 10.0

Deprecated cConvID Equivalent Valid cConvID
com.adobe.acrobat.html-3-20 com.adobe.acrobat.html
com.adobe.acrobat.htm l- 4-01-css-1-00 com.adobe.acrobat.html

Refer to Acrobat SDK and documents to learn more. You can download the SDK package, and develop application on the samples.


Format Convert speed Extract table Complete Analyze
XML Fast Yes Good Easy
Word Slow Yes Good General
Excel General Yes Great Hard
TXT Fatest No General Hardest
HTML Fast Yes Best Easy


The source code is for learning and communication only.
Please contact Adobe for commercial use.


Apache License 2.0


View Github