Resolved Issue #1915#1916
Conversation
yawpitch
left a comment
There was a problem hiding this comment.
1- Dictionary entries are ALWAYS on separate lines, which makes them less easy to read when dealing with small entries. Like here
I’d consider that more readable, not less, however readability of the tests isn’t as high a priority as consistency.
2- These very same entries have a redundant comma after their last entry. Like here
The advantage of the comma is that it allows a future commit to add a new entry line and not create a 2 line diff. It also allows entries to be arbitrarily rearranged / sorted without a potential SyntaxError.
I don’t think either of these are significant issues.
|
Great! Thank you for your swift review. |
|
Yes the formatting is black’s doing. We can expect it to make some changes relative to the hand-made templates. We could reduce that diff noise by going through and doing a PR that ran black on every non-auto-generated test file, but we’ve been handling it exercise by exercise as we introduce generator templates. As for the failure that appears to be a bit of a race condition with #1914 and #1912 not being completed. I’ll try and clear that up today if possible. |
Co-Authored-By: Corey McCandless <cmccandless@users.noreply.github.com>
* added jinja2 template for etl auto test generation * updated etl tests to v2.0.1 canonical data * minor update to template * Update exercises/etl/.meta/template.j2 Co-Authored-By: Corey McCandless <cmccandless@users.noreply.github.com>
Currently, there are two minor issues with the generated test file which appear to be with the black auto formatter:
1- Dictionary entries are ALWAYS on separate lines, which makes them less easy to read when dealing with small entries. Like here
2- These very same entries have a redundant comma after their last entry. Like here
edit:
Resolved #1915