pum
pum copied to clipboard
Pum 1.0.0, questions and ideas
Hi to all, Pum users and developers. Pum is growing and I think it's time to do some cleaning and refactoring to the code and define what the future of Pum should be.
I started doing a global refactoring of the Pum's code on a new branch https://github.com/opengisch/pum/tree/to_version_one (at the moment it is very incomplete).
Here are my ideas:
Goals
- improve the code structure
- fix the issue about the config file #5, difficult to solve with the current code structure
- fix some illogical things (for example, the current
baseline
command require the delta directory as an option), difficult to fix with the current code structure - make adding new commands much easier
- improve the tests
- improve the documentation (for users and developers)
Questions and ideas
- divide the
test-and-upgrade
command into 2 separatetest
andupgrade
(already existing) commands - improve the output of the check command, but how?
- add other useful commands (for example, a command that generate a delta file from the difference between two databases)
- ...
Feel free to add comments and ideas to this issue
Sounds great to me. I agree with all the goals you listed. And I agree that the output of the check command should be improved – quite hard to read currently.
I am not sure about removing the test-and-upgrade
command though. As long as the underlying commands (check
and upgrade
) are available I find it good to provide a higher-level command that just does the right thing.
One thing that we may need in the future is support for multiple delta dirs. This can be added later, but I think the new design should not make it hard to support that in the future.
Another thing I'd be interested improving: the test framework. Running the automated tests is not convenient today. I'd suggest to improve the situation and use pytest.
@marioba Custom SRS for deltas could be great otherwise its a mess for internationalization https://github.com/qwat/qwat-data-model/issues/199
I'd say "custom parameters" rather than "custom SRS".
Pum could use cursor.execute(delta_file.read(), vars)
, with vars
being set based on custom user parameters.
I'd say "custom parameters" rather than "custom SRS". +1
I have a pr almost ready for this.
Le mar. 24 avr. 2018 à 03:12, Régis Haubourg [email protected] a écrit :
I'd say "custom parameters" rather than "custom SRS".
+1
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/opengisch/pum/issues/25#issuecomment-383827661, or mute the thread https://github.com/notifications/unsubscribe-auth/AAHxG2qXJeSYC9R-JqfA86Ve3HQZaUXeks5trtBVgaJpZM4SPnGO .
done :)