What it does?
GRI lists reviews in a compressed format which also allows you to click either review number of topic to open them in your browser, if your terminal supports hyperlink extension.
Since I created that tool the number of open reviews I had went down considerably and I suspect that its ability to focus on reviews that are likely to merge or to be abandoned helps cleaning up the queus.
Why not improving GerTTY?
GerTTY is lovely and improving it would have being my first choice but at this moment I would evaluate it as “unmaintained”, or at least not an active project. I would not have a chance of introducing major changes like consolidating reviews from multiple servers.
If you ever used GerTTY you probably know that it supports multiple servers, but it does this in a way that requires you to start a different instance for each server. There is effectively no way to see all reviews in a single terminal window. This was a deal-breaker for me.
Shortly, at this moment there is no configuration file for
gri, as it loads Gerrit servers from
~/gertty.yaml config, with one caveat credentials are loaded from
~/netrc file. That was my decision as I found the idea of storing credentials in a single place, different than tool config safer.
To install it just run
pip install gri
gri will list only your own reviews. To display incoming ones add
-i . Debug mode will print
json received from gerrit.
$ gri --help [11:33:34] Usage: gri [OPTIONS] Options: -d, --debug Debug mode -i, --incoming Incoming reviews (not mine) -s, --server TEXT Query a single server instead of all --help Show this message and exit.
-s option is the most recent addition, and allows you to query only one server. If you do
-s0it will query only the first server from your config. It proved to be quick way to focus on reviews from a single server.
The sorting order is not configurable yet and is based on a scoring system which values more reviews that are closer to being merged and that penalizes reviews that are unlikely to be ready to merge. So down-votes from users of CI, draft/wip/dnm statuses or merge conflicts do all lower the score value.
You are welcomed to contribute new features to it, in the end is mare out out about 100 lines of python code. Before implementing a new feature is would better to create a bug that describes the problem and the desired solution, based on this we can see if we can change its default behavior or make the new feature optional.