fix: Fixed a bug where imports would not check reexports for shortest path #112

Masara · 2024-04-13T19:29:33Z

Closes #82

Summary of Changes

Fixed the bug where imports would not use the shortest path from existing reexports.

codecov · 2024-04-13T19:30:58Z

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 99.55%. Comparing base (cb061ab) to head (9ea4421).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #112      +/-   ##
==========================================
+ Coverage   99.53%   99.55%   +0.01%     
==========================================
  Files          23       23              
  Lines        2373     2462      +89     
==========================================
+ Hits         2362     2451      +89     
  Misses         11       11

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

github-actions · 2024-04-13T19:31:04Z

🦙 MegaLinter status: ✅ SUCCESS

Descriptor	Linter	Files	Fixed	Errors	Elapsed time
✅ PYTHON	black	5	0	0	1.23s
✅ PYTHON	mypy	5		0	5.57s
✅ PYTHON	ruff	5	0	0	0.04s
✅ REPOSITORY	git_diff	yes		no	0.02s

See detailed report in MegaLinter reports
Set VALIDATE_ALL_CODEBASE: true in mega-linter.yml to validate all sources, not only the diff

MegaLinter is graciously provided by

…s' into 82-wrong-imports-in-created-stubs

...pshots__/test_generate_stubs/TestStubFileGeneration.test_stub_creation[class_module].sdsstub

tests/data/various_modules_package/__init__.py

…ate the stubs for them at the directory where they were reexported from

…s' into 82-wrong-imports-in-created-stubs

… & Alias names are now correctly taken over to stubs and whitelist reexports are now handled correctly

# Conflicts: # src/safeds_stubgen/stubs_generator/_generate_stubs.py

lars-reimann

On this branch, no stubs are created at all for Safe-DS:

Masara · 2024-04-25T17:23:04Z

For me stubs are created. Did you remove the -p argument from the call? The last PR merge removed the --package argument for the api call

lars-reimann · 2024-04-25T17:27:53Z

Yep, same call as for main, where it works fine: safe-ds-stubgen -o "out" -nc -s "../library/src" --docstyle NUMPYDOC -v

Did you delete your out folder to ensure that new files are actually being created?

…e the imports names where converted too soon

Masara · 2024-04-26T08:44:20Z

I did find and fix a bug here, which wouldn't let the parser create stubs for the latest main (same bug as mentioned here.)
Could you try again?

lars-reimann · 2024-05-02T13:40:17Z

@Masara Please take a look at the merge conflicts. Then this should also be good to go.

# Conflicts: # src/safeds_stubgen/stubs_generator/_generate_stubs.py # tests/safeds_stubgen/stubs_generator/test_generate_stubs.py

lars-reimann · 2024-05-02T17:47:21Z

Unfortunately, this is broken for me again. I'm running safe-ds-stubgen -o "./out" -nc -s "E:/Repositories/safe-ds/Library/src" --docstyle NUMPYDOC -v, as before, but now my output looks like this:

safeds/data/image/containers/containers.sdsstub contains the full classes Image and ImageList.
safeds/data/tabular/containers/containers.sdsstub only contains this:

package safeds.data.tabular.containers

class Table

Imports aren't being created at all, so I cannot check whether that works now.

…ctory

Masara · 2024-05-03T19:31:00Z

@lars-reimann Should be fixed now, could you take another look?

lars-reimann · 2024-05-03T19:49:31Z

Code is being generated again, and the imports look good as well.

I greatly preferred the previous folder structure, though. Previously, there were folders for the package parts (e.g. safeds/data/tabular/containers) and then individual files inside them (e.g. table.sdsstub, column.sdsstub, ...). Now the files are combined into one, leading to very long stub files like safeds/data/tabular/containers.sdsstub). The point of having packages is that more than one file can belong to them.

The behavior from #81 should be restored. If it's easier, I'd also be fine with generating one file for each global declaration (class, global function, ...) with the same name as the declaration.

Masara · 2024-05-03T20:51:01Z

@lars-reimann How do you mean should it look like? Instead of writing every declaration into, for example, the safeds/data/tabular/containers.sdsstub file, I should create files like safeds/data/tabular/containers/Column.sdsstubs which only contains one class / global function?

lars-reimann · 2024-05-03T21:30:14Z

@lars-reimann How do you mean should it look like? Instead of writing every declaration into, for example, the safeds/data/tabular/containers.sdsstub file, I should create files like safeds/data/tabular/containers/Column.sdsstubs which only contains one class / global function?

That would be OK, yes.

…tions: Now they are created as extra stub files

Masara · 2024-05-04T00:06:46Z

I changed the way the directories are build, is this all right now?

lars-reimann · 2024-05-04T07:16:41Z

The file structure is great now. However, now all packages are wrong again. Example (safeds/data/tabular/containers/Column.sdsstub):

Actual:

package safeds.data.tabular.containers.Column

Expected:

package safeds.data.tabular.containers

Another issue is that files import their own declaration. Example (safeds/data/tabular/containers/Column.sdsstub):

from safeds.data.tabular.containers import Column

These imports should be removed.

lars-reimann · 2024-05-04T07:23:26Z

We also need tests to detect these regressions or look more closely at updated snapshots before overwriting them.

…s' into 82-wrong-imports-in-created-stubs

… and a bug where imports for themselves would be created

Masara · 2024-05-04T11:12:49Z

That was on me, I didn't check the snapshots thoroughly last time. It should work correctly now.

lars-reimann

Awesome, works very well.

## [0.3.0](v0.2.0...v0.3.0) (2024-05-04) ### Features * Added handling for sequence classes ([#127](#127)) ([cb061ab](cb061ab)), closes [#126](#126) * DocString result names for Safe-DS stub results ([#101](#101)) ([fe163e3](fe163e3)), closes [#100](#100) * Examples from docstrings are also taken over to stub docstrings ([#116](#116)) ([6665186](6665186)), closes [#115](#115) * Replace the docstring_parser library with Griffe ([#79](#79)) ([9b2f802](9b2f802)) ### Bug Fixes * `Self` types as results are translated to class names ([#110](#110)) ([4554a56](4554a56)), closes [#86](#86) * Creating stubs with relative paths for source and output directories ([#128](#128)) ([b4493c9](b4493c9)), closes [#125](#125) * Docstrings have the correct indentation for nested classes (stubs) ([#114](#114)) ([c7b8550](c7b8550)), closes [#113](#113) * Fixed a bug where double ? would be generated for stubs ([#103](#103)) ([c35c6ac](c35c6ac)), closes [#87](#87) [#87](#87) * Fixed a bug where imports would not check reexports for shortest path ([#112](#112)) ([48c5367](48c5367)), closes [#82](#82) * Fixed a bug where results in stubs would not be named ([#131](#131)) ([4408c84](4408c84)), closes [#100](#100) * Fixed a bug which prevented mypy version update ([#107](#107)) ([501d2cd](501d2cd)) * Fixed the stubs generator ([#108](#108)) ([9ad6df6](9ad6df6)), closes [#80](#80) * Generated names of callback results start with result, not with param ([#104](#104)) ([6e696e9](6e696e9)), closes [#85](#85) * Include lines of examples that start with `...` ([#130](#130)) ([3477b4a](3477b4a)), closes [#129](#129) * No "// TODO ..." if return type is explicitly `None` ([#111](#111)) ([08e345f](08e345f)), closes [#83](#83) * Removed the Epydoc parser ([#89](#89)) ([684a101](684a101)) * Replaced tabs with 4 spaces ([#105](#105)) ([8e7aa5d](8e7aa5d)), closes [#84](#84) * The file structure of stubs resembles the "package" path. ([#106](#106)) ([ff1800e](ff1800e)), closes [#81](#81) * Translation of callable ([#102](#102)) ([c581e6a](c581e6a)), closes [#88](#88) [#88](#88)

lars-reimann · 2024-05-04T15:36:57Z

🎉 This PR is included in version 0.3.0 🎉

The release is available on:

v0.3.0
GitHub release

Your semantic-release bot 📦🚀

fixed a bug where imports would not check reexports for shortest path

50bc623

Masara requested a review from a team as a code owner April 13, 2024 19:29

Masara linked an issue Apr 13, 2024 that may be closed by this pull request

Wrong imports in created stubs #82

Closed

megalinter-bot and others added 3 commits April 13, 2024 19:31

style: apply automated linter fixes

52d4974

New test data to test shortest import path for reexports

33cf801

Merge remote-tracking branch 'origin/82-wrong-imports-in-created-stub…

a9e632c

…s' into 82-wrong-imports-in-created-stubs

Masara assigned lars-reimann Apr 13, 2024

megalinter-bot and others added 2 commits April 13, 2024 19:34

style: apply automated linter fixes

74a21c9

Snapshot fixes

536e9f5

lars-reimann reviewed Apr 15, 2024

View reviewed changes

...pshots__/test_generate_stubs/TestStubFileGeneration.test_stub_creation[class_module].sdsstub Outdated Show resolved Hide resolved

lars-reimann reviewed Apr 15, 2024

View reviewed changes

tests/data/various_modules_package/__init__.py Show resolved Hide resolved

Masara and others added 6 commits April 20, 2024 19:54

Merge branch 'main' into 82-wrong-imports-in-created-stubs

5027518

reworked how reexported classes and functions are handled. Now we cre…

511b2b3

…ate the stubs for them at the directory where they were reexported from

style: apply automated linter fixes

81296df

style: apply automated linter fixes

94b951d

Adjusting test and snapshot data [Part 1]

2a72e82

Merge remote-tracking branch 'origin/82-wrong-imports-in-created-stub…

c1377a3

…s' into 82-wrong-imports-in-created-stubs

Masara force-pushed the 82-wrong-imports-in-created-stubs branch from e2356d7 to 7641bc3 Compare April 22, 2024 00:15

fix for reexported classes per wildcard import that are used as types…

49c668a

… & Alias names are now correctly taken over to stubs and whitelist reexports are now handled correctly

Masara force-pushed the 82-wrong-imports-in-created-stubs branch from a2bd81a to 49c668a Compare April 22, 2024 12:24

Masara added 3 commits April 22, 2024 14:29

test fix

cf825ca

Merge branch 'main' into 82-wrong-imports-in-created-stubs

2988f93

Merge branch 'main' into 82-wrong-imports-in-created-stubs

8cef177

# Conflicts: # src/safeds_stubgen/stubs_generator/_generate_stubs.py

lars-reimann requested changes Apr 25, 2024

View reviewed changes

fixed a bug where imports in parsed modules could not be found becaus…

63f884c

…e the imports names where converted too soon

Merge branch 'main' into 82-wrong-imports-in-created-stubs

c690209

# Conflicts: # src/safeds_stubgen/stubs_generator/_generate_stubs.py # tests/safeds_stubgen/stubs_generator/test_generate_stubs.py

lars-reimann mentioned this pull request May 2, 2024

Wrong imports in created stubs #82

Closed

fixed a bug where package modules would be created with an extra dire…

47730e2

…ctory

Masara force-pushed the 82-wrong-imports-in-created-stubs branch from 8d7ef3b to 47730e2 Compare May 3, 2024 19:30

Masara and others added 2 commits May 3, 2024 22:36

linter fix

efa2a3a

style: apply automated linter fixes

04e171e

Masara and others added 2 commits May 4, 2024 01:53

Changed the logic of stub directories for reexported classes and func…

1881972

…tions: Now they are created as extra stub files

style: apply automated linter fixes

b4e5da1

Masara added 3 commits May 4, 2024 11:46

Merge branch 'main' into 82-wrong-imports-in-created-stubs

8c1f4ec

Merge remote-tracking branch 'origin/82-wrong-imports-in-created-stub…

9dcc070

…s' into 82-wrong-imports-in-created-stubs

Fixed an error where the package info of reexported stubs where wrong…

4a0db67

… and a bug where imports for themselves would be created

Masara force-pushed the 82-wrong-imports-in-created-stubs branch from cd98e08 to 4a0db67 Compare May 4, 2024 10:56

Merge branch 'main' into 82-wrong-imports-in-created-stubs

9ea4421

Masara mentioned this pull request May 4, 2024

Use result names in docstrings for result names in generated stubs #100

Closed

lars-reimann approved these changes May 4, 2024

View reviewed changes

lars-reimann merged commit 48c5367 into main May 4, 2024
8 checks passed

lars-reimann deleted the 82-wrong-imports-in-created-stubs branch May 4, 2024 12:18

lars-reimann added the released Included in a release label May 4, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix: Fixed a bug where imports would not check reexports for shortest path #112

fix: Fixed a bug where imports would not check reexports for shortest path #112

Masara commented Apr 13, 2024

codecov bot commented Apr 13, 2024 •

edited

github-actions bot commented Apr 13, 2024 •

edited

lars-reimann left a comment

Masara commented Apr 25, 2024

lars-reimann commented Apr 25, 2024

Masara commented Apr 26, 2024

lars-reimann commented May 2, 2024

lars-reimann commented May 2, 2024 •

edited

Masara commented May 3, 2024

lars-reimann commented May 3, 2024 •

edited

Masara commented May 3, 2024

lars-reimann commented May 3, 2024

Masara commented May 4, 2024

lars-reimann commented May 4, 2024 •

edited

lars-reimann commented May 4, 2024 •

edited

Masara commented May 4, 2024

lars-reimann left a comment

lars-reimann commented May 4, 2024

fix: Fixed a bug where imports would not check reexports for shortest path #112

fix: Fixed a bug where imports would not check reexports for shortest path #112

Conversation

Masara commented Apr 13, 2024

Summary of Changes

codecov bot commented Apr 13, 2024 • edited

Codecov Report

github-actions bot commented Apr 13, 2024 • edited

🦙 MegaLinter status: ✅ SUCCESS

lars-reimann left a comment

Choose a reason for hiding this comment

Masara commented Apr 25, 2024

lars-reimann commented Apr 25, 2024

Masara commented Apr 26, 2024

lars-reimann commented May 2, 2024

lars-reimann commented May 2, 2024 • edited

Masara commented May 3, 2024

lars-reimann commented May 3, 2024 • edited

Masara commented May 3, 2024

lars-reimann commented May 3, 2024

Masara commented May 4, 2024

lars-reimann commented May 4, 2024 • edited

lars-reimann commented May 4, 2024 • edited

Masara commented May 4, 2024

lars-reimann left a comment

Choose a reason for hiding this comment

lars-reimann commented May 4, 2024

codecov bot commented Apr 13, 2024 •

edited

github-actions bot commented Apr 13, 2024 •

edited

lars-reimann commented May 2, 2024 •

edited

lars-reimann commented May 3, 2024 •

edited

lars-reimann commented May 4, 2024 •

edited

lars-reimann commented May 4, 2024 •

edited