Develop a graphical ZIM explorer

[benoit74]

While checking ZIM quality or diagnosing issues, the QA team often needs to:

• look at the list of entries of a given ZIM
• search for a specific text or patterns in this list (and count entries, get total size in ZIM)
• get details (mainly size) of a given entry and observe content (when its text)
• get list of duplicated entries

This is especially important for Zimit ZIMs where the ZIM content is way less predicatable than other scrapers.

While as a developer I do this with a combination of zimcheck / zimdump / grep / wc / ..., this is not possible for the rest of the content team and this is not convenient at all for me, consuming too much of my time.

Should we develop a graphical ZIM explorer?

@RavanJAltaie @Popolechien does it make any sense for you? Do you wanna add more?

[Popolechien]

That is a good question. Feels like this would be needed, but I also suspect the tool would still be too complex for us to use as we wouldn't know what it is we're looking for in a zim.
Other than that, isn't there a bit of overlap with the CMS (or couldn't some of the features you mention be integrated into it)?

[rgaudin]

• I still believe training a couple people to use a few tools would be a better way.
• there is indeed overlap with the CMS but CMS is a (to-be) publication & QA tool. I can see use cases that it wont cover in your workflow.
• a good part of the needs is available via fuse module. Maybe packaging it properly for windows would help

[mgautierfr]

a good part of the needs is available via fuse module. Maybe packaging it properly for windows would help

I may have miss something. But we have that ?

[rgaudin]

I may have miss something. But we have that ?

I though you had someone working on one…

[mgautierfr]

We have this openzim/zim-tools#400
But I don't know if juzz0 is actually working on it.

[kelson42]

Definitly a ticket for ZIM tools, I would recommend to make one ticket per question. Current tools should be able to do the job. If people can not use the command line and follow a howto then they are not a fit to diagnose anything for a broken ZIM IMHO. ZIM fuse might be indeed part of the response, but this is Linux only.

[rgaudin]

ZIM fuse might be indeed part of the response, but this is Linux only.

It's not ; via third party tools which is why I mentioned packaging. Not sure setup can be simplified enough though.

[benoit74]

I had something way simpler than command line tools, and way more straightforward than downloading ZIM, issuing proper command line, pushing result to a text editor because I still suck at using less, forgetting to delete the ZIM and finally having a disk full, ... Even for me it would save me time. Is it worth it? I don't have measures to bring to the table, but I know it would increase significantly my confort.

v1 could be just a web UI running on storage server so that ZIMs are already available locally, and displaying a raw list of files, and once a ZIM is clicked displaying the list of entries.

I just realize this is indeed part of the CMS whishlist "Book Overview with basic search/filter". Let's close this issue then, CMS feature will be enough from my perspective if I understand it correctly.

[rgaudin]

For some reason I read Desktop while it's not written anywhere 😅