Bug #66

GET_TEXT crashes server

Added by Lionel Martin over 12 years ago. Updated over 12 years ago.

Status:ClosedStart date:06/20/2012
Priority:NormalDue date:
Assignee:Pierre Marc% Done:

100%

Category:-
Target version:2.12
Operating System:Any Tested:Yes
Version:2.11

Description

Hello,

We have an issue with the latest version of PDF service (2.11).
Here is the code to reproduce :
my $filepath = "/<path_to>/main2.pdf";
NV::Command("NV_CMD=|OBJECT:CREATE| NAME=|file| TYPE=|FILE| PERSIST=|-1| REPLACE=|YES| FILENAME=|$filepath| NV_CONTAINER=|nvdef|");

NV::Command("NV_CMD=|PDF:PDF:GET_TEXT| IN=|file| OUT_TEXT=|out_text| PAGE=|1| AREA=|28,28,427,567| NV_CONTAINER=|nvdef|");

It has been reproduced on Linux32 and Windows 64 (both version 4.7.003).

We could reproduce it with PDF version 2.07, but with version 2.06, the server does not crash. (The content is still weird)

main2.pdf - PDF file to reproduce the bug (72.1 KB) Lionel Martin, 06/20/2012 03:33 PM

History

#1 Updated by Lionel Martin over 12 years ago

  • Description updated (diff)

Try wiki formatting for description

#2 Updated by Lionel Martin over 12 years ago

Further investigation showed that the problem is linked to the content of the table, starting with the text:
BL N° 1496510 du 12/06/12 Nre...
Vs Référ Commande 11/06/12...
Notre Réf C.de U 315578...
... and following lines

When we try to copy this text from Adobe Reader, we do not paste text but weird characters.

Hope this will help track down the issue.

#3 Updated by Pierre Marc over 12 years ago

  • Assignee set to Pierre Marc
  • Status changed from New to In Progress

#4 Updated by Pierre Marc over 12 years ago

  • Tested changed from No to Yes
  • % Done changed from 0 to 90

The problem has been corrected. This was a bug in X sorting algorithm of the text blocks. Text blocks are first sorted by Y coordinate and then by X coordinate. The issue was in the X sorting. In some situations, the blocks were not sorted correctly and were duplicated. This was producing a crash when deleting duplicated blocks (deleting an already deleted block). The sort algorithm on X has been completly rebuilt.
Regression testing has been made successfully.

PLEASE CONTACT ME TO GET THE BETA VERSION OF THE CORRECTION AND TO TEST ON YOUR SIDE.

#5 Updated by Pierre Marc over 12 years ago

  • Status changed from In Progress to Resolved
  • Target version set to 2.12
  • % Done changed from 90 to 100

Testing of the beta version ok so we release. The correction is in v 2.12.

#6 Updated by Pierre Marc over 12 years ago

  • Status changed from Resolved to Closed

Also available in: Atom PDF