Attaching PDF link to a reference will cause Firefox "not Responding"

Gurdas_Sandhu

When I attach a PDF to a reference using the 'Attach Link to File' feature, the result is not instantaneous. Sometimes this operation can take up to 10 seconds, during which time my Firefox status bar says "Not Responding". Is this normal behavior (due to Zotero trying to index my PDF)? How can I avoid this or at least push back the delay to a time when my computer is idle?

Using Zotero (4.0.15) with Firefox (24.0) on Windows 7 64-bit and Word 2010 32-bit. I have updated Zotero and Firefox to their latest versions today but ti seems the issue remains.

dstillman

This isn't new, but it's something we're currently trying to address. It's definitely slow with larger files.

In the meantime, though, how big is the PDF in question? And what are your indexing settings (max characters and max pages) in the Search pane of the Zotero preferences?

Gurdas_Sandhu

Thanks, Dan. Relieved to know this is not a sign of a larger issue with my Zotero/Firefox/Windows installation. Yesterday I was working with PDFs that go from 500 KB to 20 MB. My max characters and max pages to index are set to default, 500000 and 100, respectively.

Index statistcs are:
Indexed: 539
Partial: 38
Unindexed: 99
Words: 72500

Further questions:
1. Am I right in assuming the delay is because Zotero is trying to index the PDF I just attached as a link?

2. If yes, is it safe to assume when I do simple or advanced search in Zotero, it also considers the text inside the PDFs?

3. Are the partial and unindexed files those which Zotero cannot read, such as scanned image PDFs?

4. Can Zotero index PDFs that say "Secured" in the file title bar?

5. Is there a way I can pause indexing during sessions when I am adding references at short intervals and then re-start indexing when I am done adding?

On another note, I love Zotero. Thank you for the tremendous time and effort you'll have invested in this project.

adamsmith

yes on 1-3. (re 2: in the simple search you need to use the "Everything" setting)
I believe so on 4, though it depends on the details (usually this is just a protection against editing). 5. No, not currently.

Gurdas_Sandhu

Thanks, Adam.

dstillman

Note, though, that you could set the max-pages setting to 0 to temporarily disable indexing and then increase it again and click Rebuild Index. It will then ask you if you want to rebuild the whole index or just index unindexed items.

Gurdas_Sandhu

Thanks, Dan. That is a step forward because most of my Zotero library building happens in spurts and your trick is good enough for me. I gave it a shot and confirm it works like a charm.

I was not aware of the "just unindexed items" feature since I've never rebuilt my index. I was assuming the index is automatically maintained up to date. So, I tried rebuilding and just unindexed items options hoping it will make no change to my index since no new items were added in last couple hours. But, the results were very different. Given below are the index stats before and after my action.

Indexed: 539 -> 567
Partial: 38 -> 41
Unindexed: 99 -> 68
Words: 72500 -> 115014

Questions:

6. Does this mean that items that Zotero was unable to index at first attempt, such as a PDF with scanned text, will remain unindexed even when I have OCRed the file at a later date?

7. And so, I should regularly rebuild the index using the index unindexed items option?

Gurdas_Sandhu

Dan, Q 6 and 7.

dstillman

6. Yes.

7. If you're regularly OCRing existing image PDFs, sure.