GLIDE Number and Humanitarian Plan codelists need updating

[IsabelBirds]

GLIDE Number | Replicated Codelists and
Humanitarian Plan | Replicated Codelists haven't been synced for a while and are causing some issues for users.

cc Bee Webb, Steven Flower

[stevieflow]

@markbrough

• the GLIDE list is via this scraper, which may have stopped: https://github.com/codeforIATI/humanitarian-emergency-codelists?
• the Humanitarian Plan list seems to be up-to-date here https://github.com/codeforIATI/Unofficial-Codelists/blob/master/xml/HumanitarianPlan.xml - but then out of synch on the website: https://codelists.codeforiati.org/HumanitarianPlan/

[markbrough]

Thanks for flagging these issues @stevieflow @IsabelBirds

It looks like @andylolz merged a couple of PRs a few days ago - thank you!!

NB the process is:

• the scraper repo (one repo per codelist) gets data into a CSV file stored in the gh-pages branch of that repo
• the codelists-updater repo checks these files and makes them into XML files, taking into account any existing and withdrawn codes: https://github.com/codeforiati/codelists-updater
• the codelists repo builds the website: https://github.com/codeforiati/codelists

IIRC there was some issue with the humanitarian response plan codelists… I am not sure if the API had changed or if the scraper was not correctly handling withdrawn codes, similar to this comment:

codeforIATI/codelist-updater#79

I think that is why I didn’t merge those PRs (but I can’t remember)

However, it also looks like the GitHub Actions dependencies need to be updated for the codelists website, as it is not currently building the website:

https://github.com/codeforIATI/codelists/actions/runs/17250792424

I will take a look at that issue tomorrow!

Would also be great to know what people are using these codelists for if you are able to share more information?

[stevieflow]

Great, thanks @markbrough

Yes - this is a IATI publisher using their data to generate reports such as FTS --> they have been incorporating these lists into their workflow, and hence asked why the 2025 plans were not there

[markbrough]

Unfortunately, the issue is that the version of nodejs the Github Actions dependencies are running is too old. That means that we have to upgrade nodejs; but then we run into all kinds of other issues with vuepress. After spending a bit of time on it, I think my preference is just to go ahead and make the jump to NuxtJS, also because using vuepress so far has involved a bunch of hacks, and it is not very well supported anymore. Switching to NuxtJS should be quite quick to do but needs a bit of time. I think I can get this done on Friday.

#22

[markbrough]

I just finished rewriting to NuxtJS in #22 - the Github Action is now running again and the codes should now all be updated. Please let me know if you have any other issues, otherwise feel free to close this issue!

NB, I think this transition has a few other advantages, aside from the fact that most dependencies are now more up to date:

• the code is cleaner -- there is a clearer separation between the Python scripts (which build all the data in /api/) and the scripts which build the frontend
• the search bar now also searches in the descriptions of codelists; it is quite easy to extend that to look elsewhere, e.g. Search bar only searches codelist titles #134
• I think it would also be easier to handle partial translations of codelists (e.g. we have some Spanish and Portuguese codelists, and it would be nice to make them discoverable)

[stevieflow]

Amazing - many thanks @markbrough - it's really great

Let's also discuss how this could / should replace https://iatistandard.org/en/iati-standard/203/codelists/ - as ^ is far superior, in many ways imho

Thanks

[stevieflow]

@markbrough I think the GLIDE is slightly out of synch

c4iati: https://codelists.codeforiati.org/GLIDENumber/ (updated 25th Aug - first entry TC-2025-000144-VNM

GLIDE: https://glidenumber.net/glide/public/search/search.jsp - has 14 newer entries

Is this just about when the synch-up is set, you think?

Thanks

[markbrough]

Hi @stevieflow sorry, I had forgotten the workflow a bit! It involves:

1. Scraper automatically (daily) gets data for this codelist and makes CSVs: https://github.com/codeforIATI/humanitarian-emergency-codelists
2. Codelists updater automatically (daily) gets data and makes IATI codelist XML: https://github.com/codeforiati/codelists-updater
3. Codelists are manually merged in codelist repos, e.g.: https://github.com/codeforIATI/Unofficial-Codelists/ => this triggers the codelist site to rebuild
4. Codelists site is immediately rebuilt and deployed as a static sitehttps://github.com/codeforIATI/codelists/

Step 3 is manual but at this point I think we could probably skip the manual review if we add in a couple of sanity checks?

[markbrough]

The site is now updated, anyway! If anyone else would like to be able to merge PRs in the Unofficial Codelists repo, I'm happy to add you on?

[andylolz]

Step 3 is manual but at this point I think we could probably skip the manual review if we add in a couple of sanity checks?

^^ Yeah, I think this would def be worth doing. I had a feeling there was an issue for that, but I don’t see one? I’ve created codeforIATI/codelist-updater#82, just so we have somewhere to put ideas for what those sanity checks should include.