URL management in library, and export

grguthrie

I am mainly using this for maintaining a collection of web references.
I entered them both from [Windows] desktop saved references (site.url), and from browser tabs.
I want to export this Zotero collection to a set of active and readable URLs.

With the online web version of the library, clicking the “export” icon does no do anything.

With the desktop application,
• Exporting to cvs is the only version that I find actually gives the original URLs for entries,
But it only exports the first 7 of the library (n=173).
• An RDF export, lists file names, but not the URLs in the saves reference.
• Export as bookmarks does show the URLs, but not the folder structure,
and similarly only exports the first few entries.
• Exporting as a CSL-JSON does export everything, but again shows only file information, not the URL which the desktop file represented for most items. (Some do get a full description including the URL.
The difference seems to be if they were tagged as “documents” and added to the library instead of “webpage”.
Some of them were just drag-and-drop of desktop URL files, some others were copied from browser tabs..
.url files are tagged as documents, and do not show the actual URL link.
Items D&D from browser are labelled “webpage” and do show the URL.

This raises a basic question, if I have desktop URL references (of course as files [Windows]),, saving them in a Library is hen different that entering a browser window tag, and instead saves the actual file?
If so, is here a simple way to save desktop saves web references into a Library, other than opening hem in a browser and D&D the tab?

Although I found an example (in reddit) of using a web URL to the API for exports, I couldn’t get that to work.
Example: https://api.zotero.org/groups/2352922/items?limit=10&start=0
→ will result in Zotero API JSON.

adamsmith

I'm not quite sure I understand the structure of your library. Could you share a screenshot of your middle panel that includes some items that aren't exported?

grguthrie

My post mixed several issues as I tried to make things, work, I should separate them out.
I'll mention the entry types here:
1) I am saving largely references to papers and websites.
Some of them are inserted into the library by D&D a url from the desktop, others by D&D a tab from a browser.
I had assumed (wrongly) that both would work the same, but they don't.
File.url entries show as a document (attachment), and opening them goes to the URL in the file.
https://s3.amazonaws.com/zotero.org/images/forums/u16799181/clxd4plawdaug3v6bq8i.png
Link (tab) entries are encoded as WebPage, but opening them does not go to the URL link, although a right-click on the icon and "show online" does open the URL..
And it also shows the web page contents in a preview ("attachment"), and will open that.
https://s3.amazonaws.com/zotero.org/images/forums/u16799181/y58hcmevd61an7yggauu.png
Then also (of course) the document entries are uploaded to the server, and take up space.
I assume that this is a known behavior, although it surprised me.
Is there a better way to add desktop web links as items to a library?
Some content items also save URL information, and will open the URL - ("Video Recording").
2) because of this, in an export or various dumps, documents do not show a URL, but file information.
3) so from all of this, I was looking for a way to export a library to get a listing of all the webpage (URL) references.
Most of the exports I tried, only gave a few of the total (n=172) list.
Some gave a full list, but with documents as files, not the URL they represent.
https://s3.amazonaws.com/zotero.org/images/forums/u16799181/xlb7m63h10mhy83t5f9q.png
Why does for example, "export as bookmarks" or as csv only five a few of the items?
** Is there a better way to do this?
TIA

grguthrie

I tried to get around this by using a browser API dump, and indeed I can get all items, although many (most) do not contain the URL, but instead a file (attachment) description.
I have asked about this separately, but not yet found a solution:
https://forums.zotero.org/discussion/comment/489329#Comment_489329

aborel

You should forget the API, it doesn't help with the main cause of your problem.
Zotero will only know the URL if it is recorded in a parent item (as in your second screenshot). The standard way to record web pages in Zotero is to capture them using the browser connector. Dragging HTML files or file.url shortcuts into the program only generates a standalone attachment for the file (as in your 1st screenshot), and most features of Zotero will not work very well with those - included exports, citations, API, and more.

What you are trying to do will only work the way you want if you your web pages are saved as full Zotero records with a parent metadata item. 2 ways to do it:
1. (probably the best) open the webpage of recorded in your file.url and just capture it using the connector; then delete the standalone attachment.
2. right-click on the item and create a parent attachement. Change its type to Web page, paste the URL into the proper field, add a title, etc. => probably more work in the end.

grguthrie

Thanks - sounds like I need to then revise how I import all my desktop legacy links.
I was mislead by seeing that both types of items open like a url, so presumed that the URL would be (easily found) in the available data.

grguthrie

I would note that Zotero can find the original URL even for an attachment item; obvious since it can directly open it.
And also in the API dump, it gives a [link]"alternative" url that gives a page with a button with an encoded url to the original page (i.e. the url) - but I couldn't figure out how to decode that button linked in that page.
So in theory, if one knew these encodings, one could get the API json dump and have a small program to chase down the original url's.

aborel

> Zotero can find the original URL even for an attachment item; obvious since it can directly open it.

I think Zotero just triggering the default behaviour of your operating system, it doesn't need to know the content of the file.

grguthrie

Yes, but it must have a reference to that file, or its saved version, and that file has the url in it in a readable form.
If I could find a way to get that link from the "attachment" item, I could chase it myself!
Thanks.