Skip to content

[pull] master from robotframework:master#145

Open
pull[bot] wants to merge 2032 commits into
testautomation:masterfrom
robotframework:master
Open

[pull] master from robotframework:master#145
pull[bot] wants to merge 2032 commits into
testautomation:masterfrom
robotframework:master

Conversation

@pull

@pull pull Bot commented May 10, 2022

Copy link
Copy Markdown

See Commits and Changes for more details.


Created by pull[bot]

Can you help keep this open source service alive? 💖 Please sponsor : )

@pull pull Bot added the ⤵️ pull label May 10, 2022
@squash-labs

squash-labs Bot commented Feb 14, 2023

Copy link
Copy Markdown

Manage this branch in Squash

Test this branch here: https://master-n02a3.squash.io

yanne and others added 28 commits November 27, 2025 18:06
Hopefully this doesn't make tests flakey...
- Disable current timeouts when running keyword teardown to avoid them
  stopping execution.
- After teardown, check has any of timeouts exceeded during it.
- Timeouts set inside teardown are used normally.

Fixes #5474.
Also round time left in the "timeout active" debug message to three
decimals.
Part of #5373. Fixes #5558.

---------

Co-authored-by: Clemens Otto <clemens.otto@rohde-schwarz.com>
Co-authored-by: Pekka Klärck <peke@iki.fi>
The value should be omitted altogether or '0' used instead.

Also add default value to `Get Substring` keyword's `start` argument
to actually allow omitting the value.

Fixes #5537.
Use newish Robot features to simplify tests.
Argument conversion handles conversion for us.
- Use templates to avoid duplication
- Make template keywords more generic
This was already done with `Split To Lines` and `Get Substring`, but
now it is also done with `Split String`, `Split String From Right` and
`Generate Random String`.

Fixes #5537.
Fixes #5568.

Also enhance comments about other deprecations in RF 7.4. Now all have
a TODO about removing in RF 9. Possibly those that affect signatures
(`Literal[""]`) could be removed already earlier, with others there's
no real hurry.
Also some fine-tuning.
Earlier no conversion was attempted if argument was typed like `arg:
int | object`. Nowadays conversion to types before `object` is
attempted. Fixes #5529.

We plant to change `Any` conversion this way as well. See #5571 for
details.

Also simplify `Should Be Equal` typing. The old approach would have
caused issues if `Any` handling was changed.
- Whitespace
- Better test names
pekkaklarck and others added 30 commits July 6, 2026 02:30
Includes tests for:
- parsing docstring
- saving results to XML and JSON specs using both HTML and RAW docs
- parsing XML and JSON specs

Part if #5604.
isinstance(item, IOBase) triggered ABC machinery that accessed the
object's __class__, which raised AttributeError for non-Python objects
from some vendor SDKs (e.g. Squish). Check issubclass(type(item), IOBase)
against the type instead, so type_name degrades gracefully and the real
conversion error surfaces instead of a masking AttributeError.
- Allow bold/italics formatting with headers
- Allow code-style formatting with argument names
- Ignore positional-only marker similarly as kwo marker
- Strip spaces from return documentation
- Refactor
Includes adding longer example using Markdown. Libdoc HTML files
generated based on the examples still need to be updated/added.

Part of #5604.
Some localizations were invalidated due to layout changes. These values are now set to `null` and English translations are used until they are updated.
Needed because table of contents does not contain
importing section anymore (#5697)
Part of argument and return value documentation (#5604)
Part of #5703. HTML UI code still needs to be adjusted.
There was a bug in docstring info parsing (#5604) if docstring for
`__init__` was get via `get_keyword_names`.

Parsing `__init__` docstrings was not tested at all, so tests were
added also for the non-dynamic case.
Use `inspect.cleandoc` instead of `textwrap.dedent` for
dedentation. The former automatically handles the first line like we
want and it's also three times faster.
Tags are now handled by the same parser that handles argument, return
value and exception documentation (#5604). This means all such
information is handled consistently (#5707).

Not having an empty row before `Tags:` is deprecated.
`robot.utils.split_tags_from_doc` became unnecessary and was
deprecated as well.
Mostly related to #5707.

- Add `update_docs` to TestLibrary and ResourceFile to provide more
  convenient API for updating all docstring info (also #5604)
- `KeywordImplementation.get_tags()` -> `all_tags`
- Faster regex for searching does doc have tags at all
- Document enhanced support for `Tags:` in keyword documentation.
  Fixes #5707.

- Finalize documentation related to argument, return value and
  excemption documentation. Fixes #5604.

- Mention Markdown support in few more places. Related to #5304.
This happens when a value without `__class__` is tested against
`collections.abc` classes. See #5699 about problems caused by this in
type conversion.

Such objects could surface in other places than in type conversion as
well. Handling them in `is_list_like`, `is_dict_like` and in few
other places will likely help, but problems can still occur. I really
hope `collections.abc` classes are fixed so that they don't fail like
this. See python/cpython#153772.
Also document in API docs that RF 7.5 doesn't actually support
reporting deprecation warnings yet.

Fixes #5668.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.