Issue in PDF search

[monidipas465937]

Hi,

We have an issue with a pdf document published using one of our tools. In this particular document, search is not working properly. Below are the screenshots of the issue. To reproduce the problem, you can simply search for the string "AUX_CLK_IN" and see that the search results show only 1 output whereas the string is present twice in the page.

Could you please let us know why this is happening and what tools one can possibly use to find out details in the document to debug the issue.

P.S.  I could not find a way to attach the sample pdf file, need help regarding the same.

Regards

Monidipa Sarkar

[try67]

It seems like this is a scanned document that was later OCRed. The results of an OCR process are rarely perfect. That is most likely what you're seeing here. The OCR process was able to recognize some of the text, but not all, and possibly not in the right location.

[monidipas465937]

Well,this is not a scanned document. This document is created using Cadence Allegro DEHDL tool to export schematic drawing to a pdf format.

Also the search works fine when I use another reader such as FoxIt Reader, then why doesn't it work in Adobe Reader. Also ss there a way to attach the pdf here?

Regards

[try67]

No, not directly. You need to upload it to something like Dropbox, Google Drive, Adobe Cloud, etc., and then post the link to it here.

[monidipas465937]

Here is a link to download the pdf -

search_issue.pdf - Box

Do let me know if there are any problems accessing the same.

Regards

[try67]

It's an issue with how the file was generated... The only way I could find around it is to export it as an image and then create a new PDF file from that image and run OCR on it. That produces better (although not perfect) results.

[monidipas465937]

How do I find out what is exactly wrong with the document. Also if there is an issue with how this file is generated, how is it working is other readers?

[try67]

- It has to do with the internal structure of the file. It's not so easy to find out exactly what is wrong. It would require analyzing the file in depth.

- I can't answer the second question. It's possible these applications fix the issues with the file behind the scenes, or simply ignore them.

[monidipas465937]

Hi,

So I got a suggestion that - " the search issue seems to stem from the “-“ at the end of the "AUX_CLK_IN-" text run,which is apparently interpreted by the APDFL (and Acrobat?) Wordfinder as indicating a hyphen with the next text run, which is "AUX_CLK_IN+", even though it is physically placed on the page above the previous textrun.

There are a couple of possible work-arounds that come to mind. On the search-side, you may want to turn off  hyphen detection by setting noHyphenDetection in the WordFinder parameters to true."

Can anyone tell me how to turn off hyphen detection by setting noHyphenDetection in the WordFinder parameters to true using the Adobe reader settings/UI ?

Regards