search

SuLab / scheduled-bots

GeneWiki Scheduled Bots
MIT License
9 stars 15 forks source link

homosapiens genebot does not function due to a missing key "src-version" in mygene.info #52

Closed andrawaag closed 4 years ago

andrawaag commented 4 years ago

genebot_homosapiens does not work. It reports a missing src_version

GeneBot.py On line 858 expects a value in ["src"]["src_version"] that does not seems to exist.

["src"] in https://mygene.info/v3/metadata indeed does not have ["src-version"] Has this recently been replaced and if so by what?

andrewsu commented 4 years ago

Looks like this issue is related to this older issue https://github.com/SuLab/scheduled-bots/issues/36. It looks like in the commit that addressed that previous issue (https://github.com/SuLab/scheduled-bots/commit/edc952666eb0c55cb7bf40a1ba4423f6535bf9fb), the src_version info was simply moved to a different location in the json object. However, as described in this blog post from around that same time https://mygene.info/blog/data-release-updates-and-metadata-deprecations the intent was the completely remove that node in favor of versions that are nested within each source, e.g.:

"src": {
    "PantherDB": {
      "code": {
        "branch": "v3",
        "commit": "2a4aeca",
        "folder": "src/plugins/PantherDB",
        "repo": "https://github.com/biothings/mygene.info.git",
        "url":  "https://github.com/biothings/..."
      },
      "stats": {
        "PantherDB": 156054
      },
      "version": "2017-12-11"
 }

Anyway, bottom line, it looks like we need to refactor all the bots that depend on that src_version to extract the info out of new object structure...

andrawaag commented 4 years ago

As per run #91 on jenkins the issue is fixed with the above commits