Open feydreva opened 4 months ago
Nice @feydreva , first step is add as comaintaner the user AutoUpdateBot
like this
Then, submit a pull request and create a file over config/feydreva/<package-name>.yaml
for example, here you can share your nvchecker entry to help to modify. For example
nvchecker:
gitlab: libxc/libxc
source: gitlab
use_max_tag: true
test: true
Hello, So after some effort, lots of reading about git and how it works, I made a pull request. please be kind :) if there is any mistake or other things I need to do, let me know ! Thank you for your help and patience
Thanks, let's run the pipeline for the PR #42
Hello, Autobot doesn't seem to update the package with new version. Is there anything I have to do ? how can I verify this ? Thank you
True, winamax-bin.yaml
needs python-lxml
in order to support for htmlparser.
Many nvchecker configs here using regex like this
nvchecker:
regex: pkgname-([\d.]+).tar.gz
source: regex
url: https://www.example.com
pkgver: get-pkgver
test:
enable: true
We can switch to regex instead of xml parser.
Autobot doesn't seem to update the package with new version.
https://github.com/arch4edu/aur-auto-update/actions/runs/8306337444/job/22734274209
ModuleNotFoundError: No module named 'typing_extensions'
Error: Process completed with exit code 1.
This workflow is failing. Fixing now.
In https://github.com/arch4edu/aur-auto-update/actions/runs/8308392492/job/22738576498
[I 03-16 14:21:15.467 core:374] winamax-bin: updated to 0;url=https://dl.winamax.fr/client/electron/linux/all/winamax-1.53.0.deb
No action is configured for winamax-bin.
Please check the parsed version. When the version is correct, please set test: true
to enable the build test.
hmm trying to make a new config file, but do not succeed to make it work :/ This only place I see the name of the file is in the source code of the page :
regex doesn't seem to work here
No worries, if you have some progress, we could take a look.
There is another problem. The page is hosted on cloudfront and could not be accessed from some country/region. So it may return 403 sometime in GitHub Actions.
This only place I see the name of the file is in the source code of the page : regex doesn't seem to work here
regex will work here.
nvchecker:
source: "regex"
url: "https://www.winamax.fr/jouer-au-poker?dl=1&platform=linux"
regex: 'winamax-([\d.]+).deb'
There is another problem. The page is hosted on cloudfront and could not be accessed from some country/region. So it may return 403 sometime in GitHub Actions.
A workaround could be creating a proxy to access the content. The following is an example to do it with Cloudflare Worker (generated by GPT). You can also try other cloud function services.
export default {
async fetch(request, env, ctx) {
const url = "https://www.winamax.fr/jouer-au-poker?dl=1&platform=linux";
const regex = /winamax-([\d.]+).deb/;
try {
const response = await fetch(url);
const text = await response.text();
const match = text.match(regex);
if (match && match[1]) {
return new Response(match[1]);
} else {
return new Response("No match found");
}
} catch (error) {
return new Response("Error: " + error);
}
},
};
Hello, I am really interested in this. I maintain an AUR package. I got nvcheker working, so I get a notification when a new version is there. But I need to manually update the PKGBUILD and manually git push to aur.
I would love to have a bot doing version update automatically, but reading this, I don't understand how to implement it. Could you guide me with more instruction ?
Thank you