[Office 2007 plugin error] Could not open OOXML (error reading from: C:\Users ...etc.

[rkg82]

Greetings all,

I'm using the latest Zoom Search Pro 8.0 build 1002, and systematically getting "Microsoft Office 2007 plugin error" when indexing local files .xlsx and .docx files.

These files were all created with Office 365/Word and Excel 2016 (32 bit) and saved with the default formats. I can open these files without any problem on my Windows 10 Entreprise (64 bits) system.

My Zoom Search configuration properly references xlsx and .docx files.formats for indexing.

I'm puzzled by the "Could not open OOXML" part of the logged error.

Has anyone encountered this issue or have any suggestions?

[Ray]

We haven't seen this problem, and we've just tested with some DOCX files created fresh from Office 365 Word.

So not sure what the cause is. Maybe there's some new feature or something unusual with your DOCX and XLSX files.

If you can e-mail these files to us, we can take a closer look.

[rkg82]

Unfortunately I can't send you the actual files, but I can certainly create new ones with a few target test words to index, and send you those by e-mail if I can reproduce the problem on my end. It could very well be something specific about the actual files, as they are synchronized local copies of files located on the corporate SharePoint server, hosted on the MS Cloud.

[Ray]

Yes, if you can reproduce the problems in some sample files, then please send them to us.

We have not (knowingly) encountered files synchronised by Sharepoint server, so that is interesting to note.

[rkg82]

The problematic files also include, in addition to the ones synchronized with SharePoint server, local synchronized copies of files on a SubVersion (SVN) source code server. I haven't been able to reproduce the issue with a new clean Word or Excel file created on my local machine, so I'll need to upload a clean, empty file to these servers, then download them as synchronized copies, and see what happens. Unless you have other suggestions.

[Ray]

As we can't reproduce the error here, we won't know what actions are needed to recreate your situation. So let's go with your proposed attempt and see if that produces the problematic files.

[rkg82]

Update to this thread:

So I cannot reproduce the problem with Office 2016/365 files I created from scratch, uploaded to corporate SharePoint and SVN servers, then downloaded a local copy for local indexing, which worked fine.

The issue persists however with certain (not all) Office .docx and .xlsx files created by other people, and perhaps modified by still other people, so perhaps that has something to do with NFS permissions or system security software.

To be continued ...

[David]

Can you take one of the problematic files, strip any confidential content, check that the file is still problematic and EMail it to us.

[rkg82]

Yes of course, as soon as I have a bit more time ... to be continued.

[rkg82]

OK, I have been able to reproduce the error on an isolated test file, a copy of an original from which all content and file properties have been stripped. W

hich e-mail address should I send my

Zoom Search test config and the target test file?

To resume my test configuration:

Here's the error I get:

14:07:25 - Start indexing (offline mode) at Thu Aug 29 14:07:25 2019
14:07:26 - [ERROR] [Office 2007 plugin error] Could not recognise OOXML file format (C:\Users\richardg\Personal - DO NOT BACKUP\Temp\ZoomSearch8.0B1005_Test_ErrorReadingFi le\ZoomSerach_Test_File.xlsx)
14:07:27 - Indexing completed at Thu Aug 29 14:07:27 2019

The test file was physically present on my local machine over which I have full admin permissions, running Zoom Search v8.0 Build 1005 under Windows 10 Enterprise 64 bits, with latest updates installed, 20 GB RAM Intel i5 6300U CPU 2.4 /2.5 GHz.

[David]

Contact details can be found here

[rkg82]

Have sent the test files via e-mail to your contact address.

[rkg82]

Interesting results in doing some more testing with the problematic files.

Here's my test case using three indexing runs on a test folder to which a new problematic file is added at each run:

1. Clear all files from test target folder to be indexed.
2. Copy one file "A", from the local "real" indexing target folder where it normally resides and is indexed along with 16 other Excel files in all, and paste it into an empty local "test" indexing target folder.
3. Run indexer on test target folder.
4. Log reports no failure indexing file A.
5. Copy another file, "B", into the test target folder, which now contains files A and B.
6. Run indexer on target folder.
7. Log reports the plugin failure for A.
8. Copy another file, "C", into the target folder, which now contains file A and B and C.
9. Run indexer on target folder.
10. Log now reports plugin failure for A and B!

I also ran the test a got the analog results when adding the same files A B C in random order. Same result, the most recent file addition to the folder passed, but previously added files now failed.

Although I cannot exclude a problem with the contents of each file, I note that any one such file can pass the test in a first indexing run, and fail on the second, and successive runs, with the latest new file copied into the folder passing each time.

I have sent by e-mail the cumulative complete log file for my test run; but I cannot send the actual Excel files without stripping them of all content. Note that this issue appears for certain PowerPoint pptx and Word docx files.

[Ray]

We have confirmed there's a problem with Excel spreadsheets which are blank (despite meta content) and spreadsheets which only contain numerical values. This will be fixed in the next release.

We are also looking into the problem with one failure triggering subsequent failures on other files. This is related to multi-threaded offline indexing -- let us know if this still happens for you when you have switched to Single thread indexing.

[rkg82]

I have recompiled my three standard configs, using only one thread, but I'm still getting the"

[ERROR] [Office 2007 plugin error] Could not recognise OOXML file format" error." for some of them.
I'm experimenting with a small subset of problematic files to check various things.

P.S. the problem also applies to other Office files, in my case, Word and PowerPoint.

To be continued...