This PR attempts to improve the placement of own-line comments between
branches in the setting where the comment is more indented than the
preceding node.
There are two main changes.
### First change: Preceding node has leading content
If the preceding node has leading content, we now regard the comment as
automatically _less_ indented than the preceding node, and format
accordingly.
For example,
```python
if True: preceding_node
# leading on `else`, not trailing on `preceding_node`
else: ...
```
This is more compatible with `black`, although there is a (presumably
very uncommon) edge case:
```python
if True:
this;that
# leading on `else`, but trailing in `black`
else: ...
```
I'm sort of okay with this - presumably if one wanted a comment for
those semi-colon separated statements, one should have put it _above_
them, and one wanted a comment only for `that` then it ought to have
been on the same line?
### Second change: searching for last child in body
While searching for the (recursively) last child in the body of the
preceding _branch_, we implicitly assumed that the preceding node had to
have a body to begin the recursion. But actually, in the base case, the
preceding node _is_ the last child in the body of the preceding branch.
So, for example:
```python
if True:
something
last_child_but_no_body
# leading on else for `main` but trailing in this PR
else: ...
```
### More examples
The table below is an attempt to summarize the changes in behavior. The
rows alternate between an example snippet with `while` and the same
example with `if` - in the former case we do _not_ have an `else` node
and in the latter we do.
Notice that:
1. On `main` our handling of `if` vs. `while` is not consistent, whereas
it is consistent in the present PR
2. We disagree with `black` in all cases except that last example on
`main`, but agree in all cases for the present PR (though see above for
a wonky edge case where we disagree).
<table>
<tr>
<th>Original
</th>
<th><code>main</code> </th>
<th>This
PR </th>
<th><code>black</code> </th>
</tr>
<tr>
<td>
<pre lang="python">
while True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
while True:
pass
else:
# comment
pass
</pre>
</td>
<td>
<pre lang="python">
while True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
while True:
pass
# comment
else:
pass
</pre>
</td>
</tr>
<tr>
<td>
<pre lang="python">
if True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
if True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
if True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
if True:
pass
# comment
else:
pass
</pre>
</td>
</tr>
<tr>
<td>
<pre lang="python">
while True: pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
while True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
while True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
while True:
pass
# comment
else:
pass
</pre>
</td>
</tr>
<tr>
<td>
<pre lang="python">
if True: pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
if True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
if True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
if True:
pass
# comment
else:
pass
</pre>
</td>
</tr>
<tr>
<td>
<pre lang="python">
while True: pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
while True:
pass
else:
# comment
pass
</pre>
</td>
<td>
<pre lang="python">
while True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
while True:
pass
# comment
else:
pass
</pre>
</td>
</tr>
<tr>
<td>
<pre lang="python">
if True: pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
if True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
if True:
pass
# comment
else:
pass
</pre>
</td>
<td>
<pre lang="python">
if True:
pass
# comment
else:
pass
</pre>
</td>
</tr>
</table>
Summary
--
This PR fixes#17796 by taking the approach mentioned in
https://github.com/astral-sh/ruff/issues/17796#issuecomment-2847943862
of simply recursing into the `MatchAs` patterns when checking if we need
parentheses. This allows us to reuse the parentheses in the inner
pattern before also breaking the `MatchAs` pattern itself:
```diff
match class_pattern:
case Class(xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx) as capture:
pass
- case (
- Class(xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx) as capture
- ):
+ case Class(
+ xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+ ) as capture:
pass
- case (
- Class(
- xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
- ) as capture
- ):
+ case Class(
+ xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+ ) as capture:
pass
case (
Class(
@@ -685,13 +683,11 @@
match sequence_pattern_brackets:
case [xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx] as capture:
pass
- case (
- [xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx] as capture
- ):
+ case [
+ xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+ ] as capture:
pass
- case (
- [
- xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
- ] as capture
- ):
+ case [
+ xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
+ ] as capture:
pass
```
I haven't really resolved the question of whether or not it's okay
always to recurse, but I'm hoping the ecosystem check on this PR might
shed some light on that.
Test Plan
--
New tests based on the issue and then reviewing the ecosystem check here
Summary
--
This is a first step toward fixing #9745. After reviewing our open
issues and several Black issues and PRs, I personally found the function
case the most compelling, especially with very long argument lists:
```py
def func(
self,
arg1: int,
arg2: bool,
arg3: bool,
arg4: float,
arg5: bool,
) -> tuple[...]:
if arg2 and arg3:
raise ValueError
```
or many annotations:
```py
def function(
self, data: torch.Tensor | tuple[torch.Tensor, ...], other_argument: int
) -> torch.Tensor | tuple[torch.Tensor, ...]:
do_something(data)
return something
```
I think docstrings help the situation substantially both because syntax
highlighting will usually give a very clear separation between the
annotations and the docstring and because we already allow a blank line
_after_ the docstring:
```py
def function(
self, data: torch.Tensor | tuple[torch.Tensor, ...], other_argument: int
) -> torch.Tensor | tuple[torch.Tensor, ...]:
"""
A function doing something.
And a longer description of the things it does.
"""
do_something(data)
return something
```
There are still other comments on #9745, such as [this one] with 9
upvotes, where users specifically request blank lines in all block
types, or at least including conditionals and loops. I'm sympathetic to
that case as well, even if personally I don't find an [example] like
this:
```py
if blah:
# Do some stuff that is logically related
data = get_data()
# Do some different stuff that is logically related
results = calculate_results()
return results
```
to be much more readable than:
```py
if blah:
# Do some stuff that is logically related
data = get_data()
# Do some different stuff that is logically related
results = calculate_results()
return results
```
I'm probably just used to the latter from the formatters I've used, but
I do prefer it. I also think that functions are the least susceptible to
the accidental introduction of a newline after refactoring described in
Micha's [comment] on #8893.
I actually considered further restricting this change to functions with
multiline headers. I don't think very short functions like:
```py
def foo():
return 1
```
benefit nearly as much from the allowed newline, but I just went with
any function without a docstring for now. I guess a marginal case like:
```py
def foo(a_long_parameter: ALongType, b_long_parameter: BLongType) -> CLongType:
return 1
```
might be a good argument for not restricting it.
I caused a couple of syntax errors before adding special handling for
the ellipsis-only case, so I suspect that there are some other
interesting edge cases that may need to be handled better.
Test Plan
--
Existing tests, plus a few simple new ones. As noted above, I suspect
that we may need a few more for edge cases I haven't considered.
[this one]:
https://github.com/astral-sh/ruff/issues/9745#issuecomment-2876771400
[example]:
https://github.com/psf/black/issues/902#issuecomment-1562154809
[comment]:
https://github.com/astral-sh/ruff/issues/8893#issuecomment-1867259744
When formatting clause headers for clauses that are not their own node,
like an `else` clause or `finally` clause, we begin searching for the
keyword at the end of the previous statement. However, if the previous
statement ended in a semicolon this caused a panic because we only
expected trivia between the end of the last statement and the keyword.
This PR adjusts the starting point of our search for the keyword to
begin after the optional semicolon in these cases.
Closes#21065
## Summary
I spun this out from #21005 because I thought it might be helpful
separately. It just renders a nice `Diagnostic` for syntax errors
pointing to the source of the error. This seemed a bit more helpful to
me than just the byte offset when working on #21005, and we had most of
the code around after #20443 anyway.
## Test Plan
This doesn't actually affect any passing tests, but here's an example of
the additional output I got when I broke the spacing after the `in`
token:
```
error[internal-error]: Expected 'in', found name
--> /home/brent/astral/ruff/crates/ruff_python_formatter/resources/test/fixtures/black/cases/cantfit.py:50:79
|
48 | need_more_to_make_the_line_long_enough,
49 | )
50 | del ([], name_1, name_2), [(), [], name_4, name_3], name_1[[name_2 for name_1 inname_0]]
| ^^^^^^^^
51 | del ()
|
```
I just appended this to the other existing output for now.
## Summary
Fixes#20774 by tracking whether an `InterpolatedStringState` element is
nested inside of another interpolated element. This feels like kind of a
naive fix, so I'm welcome to other ideas. But it resolves the problem in
the issue and clears up the syntax error in the black compatibility
test, without affecting many other cases.
The other affected case is actually interesting too because the
[input](96b156303b/crates/ruff_python_formatter/resources/test/fixtures/ruff/expression/fstring.py (L707))
is invalid, but the previous quote selection fixed the invalid syntax:
```pycon
Python 3.11.13 (main, Sep 2 2025, 14:20:25) [Clang 20.1.4 ] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> f'{1: abcd "{'aa'}" }' # input
File "<stdin>", line 1
f'{1: abcd "{'aa'}" }'
^^
SyntaxError: f-string: expecting '}'
>>> f'{1: abcd "{"aa"}" }' # old output
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
ValueError: Invalid format specifier ' abcd "aa" ' for object of type 'int'
>>> f'{1: abcd "{'aa'}" }' # new output
File "<stdin>", line 1
f'{1: abcd "{'aa'}" }'
^^
SyntaxError: f-string: expecting '}'
```
We now preserve the invalid syntax in the input.
Unfortunately, this also seems to be another edge case I didn't consider
in https://github.com/astral-sh/ruff/pull/20867 because we don't flag
this as a syntax error after 0.14.1:
<details><summary>Shell output</summary>
<p>
```
> uvx ruff@0.14.0 check --ignore ALL --target-version py311 - <<EOF
f'{1: abcd "{'aa'}" }'
EOF
invalid-syntax: Cannot reuse outer quote character in f-strings on Python 3.11 (syntax was added in Python 3.12)
--> -:1:14
|
1 | f'{1: abcd "{'aa'}" }'
| ^
|
Found 1 error.
> uvx ruff@0.14.1 check --ignore ALL --target-version py311 - <<EOF
f'{1: abcd "{'aa'}" }'
EOF
All checks passed!
> uvx python@3.11 -m ast <<EOF
f'{1: abcd "{'aa'}" }'
EOF
Traceback (most recent call last):
File "<frozen runpy>", line 198, in _run_module_as_main
File "<frozen runpy>", line 88, in _run_code
File "/home/brent/.local/share/uv/python/cpython-3.11.13-linux-x86_64-gnu/lib/python3.11/ast.py", line 1752, in <module>
main()
File "/home/brent/.local/share/uv/python/cpython-3.11.13-linux-x86_64-gnu/lib/python3.11/ast.py", line 1748, in main
tree = parse(source, args.infile.name, args.mode, type_comments=args.no_type_comments)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/brent/.local/share/uv/python/cpython-3.11.13-linux-x86_64-gnu/lib/python3.11/ast.py", line 50, in parse
return compile(source, filename, mode, flags,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "<stdin>", line 1
f'{1: abcd "{'aa'}" }'
^^
SyntaxError: f-string: expecting '}'
```
</p>
</details>
I assumed that was the same `ParseError` as the one caused by
`f"{1:""}"`, but this is a nested interpolation inside of the format
spec.
## Test Plan
New test copied from the black compatibility test. I guess this is a
duplicate now, I started working on this branch before the new black
tests were imported, so I could delete the separate test in our fixtures
if that's preferable.
Summary
--
Fixes#20844 by refining the unsupported syntax error check for [PEP
701]
f-strings before Python 3.12 to allow backslash escapes and escaped
outer quotes
in the format spec part of f-strings. These are only disallowed within
the
f-string expression part on earlier versions. Using the examples from
the PR:
```pycon
>>> f"{1:\x64}"
'1'
>>> f"{1:\"d\"}"
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
ValueError: Invalid format specifier '"d"' for object of type 'int'
```
Note that the second case is a runtime error, but this is actually
avoidable if
you override `__format__`, so despite being pretty weird, this could
actually be
a valid use case.
```pycon
>>> class C:
... def __format__(*args, **kwargs): return "<C>"
...
>>> f"{C():\"d\"}"
'<C>'
```
At first I thought narrowing the range we check to exclude the format
spec would
only work for escapes, but it turns out that cases like `f"{1:""}"` are
already
covered by an existing `ParseError`, so we can just narrow the range of
both our
escape and quote checks.
Our comment check also seems to be working correctly because it's based
on the
actual tokens. A case like
[this](https://play.ruff.rs/9f1c2ff2-cd8e-4ad7-9f40-56c0a524209f):
```python
f"""{1:# }"""
```
doesn't include a comment token, instead the `#` is part of an
`InterpolatedStringLiteralElement`.
Test Plan
--
New inline parser tests
[PEP 701]: https://peps.python.org/pep-0701/
Summary
--
This PR implements the black preview style from
https://github.com/psf/black/pull/4720. As of Python 3.14, you're
allowed to omit the parentheses around groups of exceptions, as long as
there's no `as` binding:
**3.13**
```pycon
Python 3.13.4 (main, Jun 4 2025, 17:37:06) [Clang 20.1.4 ] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> try: ...
... except (Exception, BaseException): ...
...
Ellipsis
>>> try: ...
... except Exception, BaseException: ...
...
File "<python-input-1>", line 2
except Exception, BaseException: ...
^^^^^^^^^^^^^^^^^^^^^^^^
SyntaxError: multiple exception types must be parenthesized
```
**3.14**
```pycon
Python 3.14.0rc2 (main, Sep 2 2025, 14:20:56) [Clang 20.1.4 ] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> try: ...
... except Exception, BaseException: ...
...
Ellipsis
>>> try: ...
... except (Exception, BaseException): ...
...
Ellipsis
>>> try: ...
... except Exception, BaseException as e: ...
...
File "<python-input-2>", line 2
except Exception, BaseException as e: ...
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
SyntaxError: multiple exception types must be parenthesized when using 'as'
```
I think this ended up being pretty straightforward, at least once Micha
showed me where to start :)
Test Plan
--
New tests
At first I thought we were deviating from black in how we handle
comments within the exception type tuple, but I think this applies to
how we format all tuples, not specifically with the new preview style.
Summary
--
```shell
git clone git@github.com:psf/black.git ../other/black
crates/ruff_python_formatter/resources/test/fixtures/import_black_tests.py ../other/black
```
Then ran our tests and accepted the snapshots
I had to make a small fix to our tuple normalization logic for `del`
statements
in the second commit, otherwise the tests were panicking at a changed
AST. I
think the new implementation is closer to the intention described in the
nearby
comment anyway, though.
The first commit adds the new Python, settings, and `.expect` files, the
next three commits make some small
fixes to help get the tests running, and then the fifth commit accepts
all but one of the new snapshots. The last commit includes the new
unsupported syntax error for one f-string example, tracked in #20774.
Test Plan
--
Newly imported tests. I went through all of the new snapshots and added
review comments below. I think they're all expected, except a few cases
I wasn't 100% sure about.
This PR resolves the issue noticed in
https://github.com/astral-sh/ruff/pull/20777#discussion_r2417233227.
Namely, cases like this were being flagged as syntax errors despite
being perfectly valid on Python 3.8:
```pycon
Python 3.8.20 (default, Oct 2 2024, 16:34:12)
[Clang 18.1.8 ] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> with (open("foo.txt", "w")): ...
...
Ellipsis
>>> with (open("foo.txt", "w")) as f: print(f)
...
<_io.TextIOWrapper name='foo.txt' mode='w' encoding='UTF-8'>
```
The second of these was already allowed but not the first:
```shell
> ruff check --target-version py38 --ignore ALL - <<EOF
with (open("foo.txt", "w")): ...
with (open("foo.txt", "w")) as f: print(f)
EOF
invalid-syntax: Cannot use parentheses within a `with` statement on Python 3.8 (syntax was added in Python 3.9)
--> -:1:6
|
1 | with (open("foo.txt", "w")): ...
| ^
2 | with (open("foo.txt", "w")) as f: print(f)
|
Found 1 error.
```
There was some discussion of related cases in
https://github.com/astral-sh/ruff/pull/16523#discussion_r1984657793, but
it seems I overlooked the single-element case when flagging tuples. As
suggested in the other thread, we can just check if there's more than
one element or a trailing comma, which will cause the tuple parsing on
<=3.8 and avoid the false positives.
## Summary
Based on the suggestion in
https://github.com/astral-sh/ruff/issues/20774#issuecomment-3383153511,
I added rendering of unsupported syntax errors in our `format` test.
In support of this, I added a `DummyFileResolver` type to `ruff_db` to
pass to `DisplayDiagnostics::new` (first commit). Another option would
obviously be implementing this directly in the fixtures, but we'd have
to import a `NotebookIndex` somehow; either by depending directly on
`ruff_notebook` or re-exporting it from `ruff_db`. I thought it might be
convenient elsewhere to have a dummy resolver, for example in the
parser, where we currently have a separate rendering pipeline
[copied](https://github.com/astral-sh/ruff/blob/main/crates/ruff_python_parser/tests/fixtures.rs#L321)
from our old rendering code in `ruff_linter`. I also briefly tried
implementing a `TestDb` in the formatter since I noticed the
`ruff_python_formatter::db` module, but that was turning into a lot more
code than the dummy resolver.
We could also push this a bit further if we wanted. I didn't add the new
snapshots to the black compatibility tests or to the preview snapshots,
for example. I thought it was kind of noisy enough (and helpful enough)
already, though. We could also use a shorter diagnostic format, but the
full output seems most useful once we accept this initial large batch of
changes.
## Test Plan
I went through the baseline snapshots pretty quickly, but they all
looked reasonable to me, with one exception I noted below. I also tested
that the case from #20774 produces a new unsupported syntax error.
Summary
--
Closes#19467 and also removes the warning about using Python 3.14
without
preview enabled.
I also bumped `PythonVersion::default` to 3.9 because it reaches EOL
this month,
but we could also defer that for now if we wanted.
The first three commits are related to the `latest` bump to 3.14; the
fourth commit
bumps the default to 3.10.
Note that this PR also bumps the default Python version for ty to 3.10
because
there was a test asserting that it stays in sync with
`ast::PythonVersion`.
Test Plan
--
Existing tests
I spot-checked the ecosystem report, and I believe these are all
expected. Inbits doesn't specify a target Python version, so I guess
we're applying the default. UP007, UP035, and UP045 all use the new
default value to emit new diagnostics.
## Summary
Removes the `module_ptr` field from `AstNodeRef` in release mode, and
change `NodeIndex` to a `NonZeroU32` to reduce the size of
`Option<AstNodeRef<_>>` fields.
I believe CI runs in debug mode, so this won't show up in the memory
report, but this reduces memory by ~2% in release mode.
As of [this cpython PR](https://github.com/python/cpython/pull/135996),
it is not allowed to concatenate t-strings with non-t-strings,
implicitly or explicitly. Expressions such as `"foo" t"{bar}"` are now
syntax errors.
This PR updates some AST nodes and parsing to reflect this change.
The structural change is that `TStringPart` is no longer needed, since,
as in the case of `BytesStringLiteral`, the only possibilities are that
we have a single `TString` or a vector of such (representing an implicit
concatenation of t-strings). This removes a level of nesting from many
AST expressions (which is what all the snapshot changes reflect), and
simplifies some logic in the implementation of visitors, for example.
The other change of note is in the parser. When we meet an implicit
concatenation of string-like literals, we now count the number of
t-string literals. If these do not exhaust the total number of
implicitly concatenated pieces, then we emit a syntax error. To recover
from this syntax error, we encode any t-string pieces as _invalid_
string literals (which means we flag them as invalid, record their
range, and record the value as `""`). Note that if at least one of the
pieces is an f-string we prefer to parse the entire string as an
f-string; otherwise we parse it as a string.
This logic is exactly the same as how we currently treat
`BytesStringLiteral` parsing and error recovery - and carries with it
the same pros and cons.
Finally, note that I have not implemented any changes in the
implementation of the formatter. As far as I can tell, none are needed.
I did change a few of the fixtures so that we are always concatenating
t-strings with t-strings.
Closes#18671
Note that while this has, I believe, always been invalid syntax, it was
reported as a different syntax error until Python 3.12:
Python 3.11:
```pycon
>>> x = 1
>>> f"{x! s}"
File "<stdin>", line 1
f"{x! s}"
^
SyntaxError: f-string: invalid conversion character: expected 's', 'r', or 'a'
```
Python 3.12:
```pycon
>>> x = 1
>>> f"{x! s}"
File "<stdin>", line 1
f"{x! s}"
^^^
SyntaxError: f-string: conversion type must come right after the exclamanation mark
```
## Summary
Garbage collect ASTs once we are done checking a given file. Queries
with a cross-file dependency on the AST will reparse the file on demand.
This reduces ty's peak memory usage by ~20-30%.
The primary change of this PR is adding a `node_index` field to every
AST node, that is assigned by the parser. `ParsedModule` can use this to
create a flat index of AST nodes any time the file is parsed (or
reparsed). This allows `AstNodeRef` to simply index into the current
instance of the `ParsedModule`, instead of storing a pointer directly.
The indices are somewhat hackily (using an atomic integer) assigned by
the `parsed_module` query instead of by the parser directly. Assigning
the indices in source-order in the (recursive) parser turns out to be
difficult, and collecting the nodes during semantic indexing is
impossible as `SemanticIndex` does not hold onto a specific
`ParsedModuleRef`, which the pointers in the flat AST are tied to. This
means that we have to do an extra AST traversal to assign and collect
the nodes into a flat index, but the small performance impact (~3% on
cold runs) seems worth it for the memory savings.
Part of https://github.com/astral-sh/ty/issues/214.
This PR implements template strings (t-strings) in the parser and
formatter for Ruff.
Minimal changes necessary to compile were made in other parts of the code (e.g. ty, the linter, etc.). These will be covered properly in follow-up PRs.
## Summary
This PR stabilizies the fix for
https://github.com/astral-sh/ruff/issues/14001
We try to only make breaking formatting changes once a year. However,
the plan was to release this fix as part of Ruff 0.9 but I somehow
missed it when promoting all other formatter changes.
I think it's worth making an exception here considering that this is a
bug fix, it improves readability, and it should be rare
(very few files in a single project). Our version policy explicitly
allows breaking formatter changes in any minor release and the idea of
only making breaking formatter changes once a year is mainly to avoid
multiple releases throughout the year that introduce large formatter
changes
Closes https://github.com/astral-sh/ruff/issues/14001
## Test Plan
Updated snapshot
## Summary
This should give us better coverage for the unsupported syntax error
features and
increases our confidence that the formatter doesn't accidentially
introduce new unsupported
syntax errors.
A feature like this would have been very useful when working on f-string
formatting
where it took a lot of iteration to find all Python 3.11 or older
incompatibilities.
## Test Plan
I applied my changes on top of
https://github.com/astral-sh/ruff/pull/16523 and
removed the target version check in the with-statement formatting code.
As expected,
the integration tests now failed
## Summary
This is part of the preparation for detecting syntax errors in the
parser from https://github.com/astral-sh/ruff/pull/16090/. As suggested
in [this
comment](https://github.com/astral-sh/ruff/pull/16090/#discussion_r1953084509),
I started working on a `ParseOptions` struct that could be stored in the
parser. For this initial refactor, I only made it hold the existing
`Mode` option, but for syntax errors, we will also need it to have a
`PythonVersion`. For that use case, I'm picturing something like a
`ParseOptions::with_python_version` method, so you can extend the
current calls to something like
```rust
ParseOptions::from(mode).with_python_version(settings.target_version)
```
But I thought it was worth adding `ParseOptions` alone without changing
any other behavior first.
Most of the diff is just updating call sites taking `Mode` to take
`ParseOptions::from(Mode)` or those taking `PySourceType`s to take
`ParseOptions::from(PySourceType)`. The interesting changes are in the
new `parser/options.rs` file and smaller parts of `parser/mod.rs` and
`ruff_python_parser/src/lib.rs`.
## Test Plan
Existing tests, this should not change any behavior.
## Summary
This PR updates the formatter and linter to use the `PythonVersion`
struct from the `ruff_python_ast` crate internally. While this doesn't
remove the need for the `linter::PythonVersion` enum, it does remove the
`formatter::PythonVersion` enum and limits the use in the linter to
deserializing from CLI arguments and config files and moves most of the
remaining methods to the `ast::PythonVersion` struct.
## Test Plan
Existing tests, with some inputs and outputs updated to reflect the new
(de)serialization format. I think these are test-specific and shouldn't
affect any external (de)serialization.
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
## Summary
This is another follow-up to #15726 and #15778, extending the
quote-preserving behavior to f-strings and deleting the now-unused
`Generator::quote` field.
## Details
I also made one unrelated change to `rules/flynt/helpers.rs` to remove a
`to_string` call for making a `Box<str>` and tweaked some arguments to
some of the `Generator::unparse_f_string` methods to make the code
easier to follow, in my opinion. Happy to revert especially the latter
of these if needed.
Unfortunately this still does not fix the issue in #9660, which appears
to be more of an escaping issue than a quote-preservation issue. After
#15726, the result is now `a = f'# {"".join([])}' if 1 else ""` instead
of `a = f"# {''.join([])}" if 1 else ""` (single quotes on the outside
now), but we still don't have the desired behavior of double quotes
everywhere on Python 3.12+. I added a test for this but split it off
into another branch since it ended up being unaddressed here, but my
`dbg!` statements showed the correct preferred quotes going into
[`UnicodeEscape::with_preferred_quote`](https://github.com/astral-sh/ruff/blob/main/crates/ruff_python_literal/src/escape.rs#L54).
## Test Plan
Existing rule and `Generator` tests.
---------
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
## Summary
This is a very closely related follow-up to #15726, adding the same
quote-preserving behavior to bytestrings. Only one rule (UP018) was
affected this time, and it was easy to mirror the plain string changes.
## Test Plan
Existing tests
## Summary
This is a first step toward fixing #7799 by using the quoting style
stored in the `flags` field on `ast::StringLiteral`s to select a quoting
style. This PR does not include support for f-strings or byte strings.
Several rules also needed small updates to pass along existing quoting
styles instead of using `StringLiteralFlags::default()`. The remaining
snapshot changes are intentional and should preserve the quotes from the
input strings.
## Test Plan
Existing tests with some accepted updates, plus a few new RUF055 tests
for raw strings.
---------
Co-authored-by: Alex Waygood <alex.waygood@gmail.com>
## Summary
Fixes https://github.com/astral-sh/ruff/issues/14778
The formatter incorrectly removed the inner implicitly concatenated
string for following single-line f-string:
```py
f"{'aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa' 'a' if True else ""}"
# formatted
f"{ if True else ''}"
```
This happened because I changed the `RemoveSoftlinesBuffer` in
https://github.com/astral-sh/ruff/pull/14489 to remove any content
wrapped in `if_group_breaks`. After all, it emulates an *all flat*
layout. This works fine when `if_group_breaks` is only used to **add**
content if the gorup breaks. It doesn't work if the same content is
rendered differently depending on if the group fits using
`if_group_breaks` and `if_groups_fits` because the enclosing `group`
might still *break* if the entire content exceeds the line-length limit.
This PR fixes this by unwrapping any `if_group_fits` content by removing
the `if_group_fits` start and end tags.
## Test Plan
added test
## Summary
This PR fixes a bug in the f-string formatting to not consider the
escaped newlines for `is_multiline`. This is done by checking if the
f-string is triple-quoted or not similar to normal string literals.
This is not required to be gated behind preview because the logic change
for `is_multiline` was added in
https://github.com/astral-sh/ruff/pull/14454.
## Test Plan
Add a test case which formats differently on `main`:
https://play.ruff.rs/ea3c55c2-f0fe-474e-b6b8-e3365e0ede5e
## Summary
fixes: #14608
The logic that was only applied for 3.12+ target version needs to be
applied for other versions as well.
## Test Plan
I've moved the existing test cases for 3.12 only to `f_string.py` so
that it's tested against the default target version.
I think we should probably enabled testing for two target version (pre
3.12 and 3.12) but it won't highlight any issue because the parser
doesn't consider this. Maybe we should enable this once we have target
version specific syntax errors in place
(https://github.com/astral-sh/ruff/issues/6591).
## Summary
fixes: #13813
This PR fixes a bug in the formatting assignment statement when the
value is an f-string.
This is resolved by using custom best fit layouts if the f-string is (a)
not already a flat f-string (thus, cannot be multiline) and (b) is not a
multiline string (thus, cannot be flattened). So, it is used in cases
like the following:
```py
aaaaaaaaaaaaaaaaaa = f"testeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee{
expression}moreeeeeeeeeeeeeeeee"
```
Which is (a) `FStringLayout::Multiline` and (b) not a multiline.
There are various other examples in the PR diff along with additional
explanation and context as code comments.
## Test Plan
Add multiple test cases for various scenarios.
When there is a function or class definition at the end of a suite
followed by the beginning of an alternative block, we have to insert a
single empty line between them.
In the if-else-statement example below, we insert an empty line after
the `foo` in the if-block, but none after the else-block `foo`, since in
the latter case the enclosing suite already adds empty lines.
```python
if sys.version_info >= (3, 10):
def foo():
return "new"
else:
def foo():
return "old"
class Bar:
pass
```
To do so, we track whether the current suite is the last one in the
current statement with a new option on the suite kind.
Fixes#12199
---------
Co-authored-by: Micha Reiser <micha@reiser.io>
## Summary
This PR updates the entire parser stack in multiple ways:
### Make the lexer lazy
* https://github.com/astral-sh/ruff/pull/11244
* https://github.com/astral-sh/ruff/pull/11473
Previously, Ruff's lexer would act as an iterator. The parser would
collect all the tokens in a vector first and then process the tokens to
create the syntax tree.
The first task in this project is to update the entire parsing flow to
make the lexer lazy. This includes the `Lexer`, `TokenSource`, and
`Parser`. For context, the `TokenSource` is a wrapper around the `Lexer`
to filter out the trivia tokens[^1]. Now, the parser will ask the token
source to get the next token and only then the lexer will continue and
emit the token. This means that the lexer needs to be aware of the
"current" token. When the `next_token` is called, the current token will
be updated with the newly lexed token.
The main motivation to make the lexer lazy is to allow re-lexing a token
in a different context. This is going to be really useful to make the
parser error resilience. For example, currently the emitted tokens
remains the same even if the parser can recover from an unclosed
parenthesis. This is important because the lexer emits a
`NonLogicalNewline` in parenthesized context while a normal `Newline` in
non-parenthesized context. This different kinds of newline is also used
to emit the indentation tokens which is important for the parser as it's
used to determine the start and end of a block.
Additionally, this allows us to implement the following functionalities:
1. Checkpoint - rewind infrastructure: The idea here is to create a
checkpoint and continue lexing. At a later point, this checkpoint can be
used to rewind the lexer back to the provided checkpoint.
2. Remove the `SoftKeywordTransformer` and instead use lookahead or
speculative parsing to determine whether a soft keyword is a keyword or
an identifier
3. Remove the `Tok` enum. The `Tok` enum represents the tokens emitted
by the lexer but it contains owned data which makes it expensive to
clone. The new `TokenKind` enum just represents the type of token which
is very cheap.
This brings up a question as to how will the parser get the owned value
which was stored on `Tok`. This will be solved by introducing a new
`TokenValue` enum which only contains a subset of token kinds which has
the owned value. This is stored on the lexer and is requested by the
parser when it wants to process the data. For example:
8196720f80/crates/ruff_python_parser/src/parser/expression.rs (L1260-L1262)
[^1]: Trivia tokens are `NonLogicalNewline` and `Comment`
### Remove `SoftKeywordTransformer`
* https://github.com/astral-sh/ruff/pull/11441
* https://github.com/astral-sh/ruff/pull/11459
* https://github.com/astral-sh/ruff/pull/11442
* https://github.com/astral-sh/ruff/pull/11443
* https://github.com/astral-sh/ruff/pull/11474
For context,
https://github.com/RustPython/RustPython/pull/4519/files#diff-5de40045e78e794aa5ab0b8aacf531aa477daf826d31ca129467703855408220
added support for soft keywords in the parser which uses infinite
lookahead to classify a soft keyword as a keyword or an identifier. This
is a brilliant idea as it basically wraps the existing Lexer and works
on top of it which means that the logic for lexing and re-lexing a soft
keyword remains separate. The change here is to remove
`SoftKeywordTransformer` and let the parser determine this based on
context, lookahead and speculative parsing.
* **Context:** The transformer needs to know the position of the lexer
between it being at a statement position or a simple statement position.
This is because a `match` token starts a compound statement while a
`type` token starts a simple statement. **The parser already knows
this.**
* **Lookahead:** Now that the parser knows the context it can perform
lookahead of up to two tokens to classify the soft keyword. The logic
for this is mentioned in the PR implementing it for `type` and `match
soft keyword.
* **Speculative parsing:** This is where the checkpoint - rewind
infrastructure helps. For `match` soft keyword, there are certain cases
for which we can't classify based on lookahead. The idea here is to
create a checkpoint and keep parsing. Based on whether the parsing was
successful and what tokens are ahead we can classify the remaining
cases. Refer to #11443 for more details.
If the soft keyword is being parsed in an identifier context, it'll be
converted to an identifier and the emitted token will be updated as
well. Refer
8196720f80/crates/ruff_python_parser/src/parser/expression.rs (L487-L491).
The `case` soft keyword doesn't require any special handling because
it'll be a keyword only in the context of a match statement.
### Update the parser API
* https://github.com/astral-sh/ruff/pull/11494
* https://github.com/astral-sh/ruff/pull/11505
Now that the lexer is in sync with the parser, and the parser helps to
determine whether a soft keyword is a keyword or an identifier, the
lexer cannot be used on its own. The reason being that it's not
sensitive to the context (which is correct). This means that the parser
API needs to be updated to not allow any access to the lexer.
Previously, there were multiple ways to parse the source code:
1. Passing the source code itself
2. Or, passing the tokens
Now that the lexer and parser are working together, the API
corresponding to (2) cannot exists. The final API is mentioned in this
PR description: https://github.com/astral-sh/ruff/pull/11494.
### Refactor the downstream tools (linter and formatter)
* https://github.com/astral-sh/ruff/pull/11511
* https://github.com/astral-sh/ruff/pull/11515
* https://github.com/astral-sh/ruff/pull/11529
* https://github.com/astral-sh/ruff/pull/11562
* https://github.com/astral-sh/ruff/pull/11592
And, the final set of changes involves updating all references of the
lexer and `Tok` enum. This was done in two-parts:
1. Update all the references in a way that doesn't require any changes
from this PR i.e., it can be done independently
* https://github.com/astral-sh/ruff/pull/11402
* https://github.com/astral-sh/ruff/pull/11406
* https://github.com/astral-sh/ruff/pull/11418
* https://github.com/astral-sh/ruff/pull/11419
* https://github.com/astral-sh/ruff/pull/11420
* https://github.com/astral-sh/ruff/pull/11424
2. Update all the remaining references to use the changes made in this
PR
For (2), there were various strategies used:
1. Introduce a new `Tokens` struct which wraps the token vector and add
methods to query a certain subset of tokens. These includes:
1. `up_to_first_unknown` which replaces the `tokenize` function
2. `in_range` and `after` which replaces the `lex_starts_at` function
where the former returns the tokens within the given range while the
latter returns all the tokens after the given offset
2. Introduce a new `TokenFlags` which is a set of flags to query certain
information from a token. Currently, this information is only limited to
any string type token but can be expanded to include other information
in the future as needed. https://github.com/astral-sh/ruff/pull/11578
3. Move the `CommentRanges` to the parsed output because this
information is common to both the linter and the formatter. This removes
the need for `tokens_and_ranges` function.
## Test Plan
- [x] Update and verify the test snapshots
- [x] Make sure the entire test suite is passing
- [x] Make sure there are no changes in the ecosystem checks
- [x] Run the fuzzer on the parser
- [x] Run this change on dozens of open-source projects
### Running this change on dozens of open-source projects
Refer to the PR description to get the list of open source projects used
for testing.
Now, the following tests were done between `main` and this branch:
1. Compare the output of `--select=E999` (syntax errors)
2. Compare the output of default rule selection
3. Compare the output of `--select=ALL`
**Conclusion: all output were same**
## What's next?
The next step is to introduce re-lexing logic and update the parser to
feed the recovery information to the lexer so that it can emit the
correct token. This moves us one step closer to having error resilience
in the parser and provides Ruff the possibility to lint even if the
source code contains syntax errors.
## Summary
This PR fixes the bug where the formatter would format an f-string and
could potentially change the AST.
For a triple-quoted f-string, the element can't be formatted into
multiline if it has a format specifier because otherwise the newline
would be treated as part of the format specifier.
Given the following f-string:
```python
f"""aaaaaaaaaaaaaaaa bbbbbbbbbbbbbbbbbb ccccccccccc {
variable:.3f} ddddddddddddddd eeeeeeee"""
```
The formatter sees that the f-string is already multiline so it assumes
that it can contain line breaks i.e., broken into multiple lines. But,
in this specific case we can't format it as:
```python
f"""aaaaaaaaaaaaaaaa bbbbbbbbbbbbbbbbbb ccccccccccc {
variable:.3f
} ddddddddddddddd eeeeeeee"""
```
Because the format specifier string would become ".3f\n", which is not
the original string (`.3f`).
If the original source code already contained a newline, they'll be
preserved. For example:
```python
f"""aaaaaaaaaaaaaaaa bbbbbbbbbbbbbbbbbb ccccccccccc {
variable:.3f
} ddddddddddddddd eeeeeeee"""
```
The above will be formatted as:
```py
f"""aaaaaaaaaaaaaaaa bbbbbbbbbbbbbbbbbb ccccccccccc {variable:.3f
} ddddddddddddddd eeeeeeee"""
```
Note that the newline after `.3f` is part of the format specifier which
needs to be preserved.
The Python version is irrelevant in this case.
fixes: #10040
## Test Plan
Add some test cases to verify this behavior.
(Supersedes #9152, authored by @LaBatata101)
## Summary
This PR replaces the current parser generated from LALRPOP to a
hand-written recursive descent parser.
It also updates the grammar for [PEP
646](https://peps.python.org/pep-0646/) so that the parser outputs the
correct AST. For example, in `data[*x]`, the index expression is now a
tuple with a single starred expression instead of just a starred
expression.
Beyond the performance improvements, the parser is also error resilient
and can provide better error messages. The behavior as seen by any
downstream tools isn't changed. That is, the linter and formatter can
still assume that the parser will _stop_ at the first syntax error. This
will be updated in the following months.
For more details about the change here, refer to the PR corresponding to
the individual commits and the release blog post.
## Test Plan
Write _lots_ and _lots_ of tests for both valid and invalid syntax and
verify the output.
## Acknowledgements
- @MichaReiser for reviewing 100+ parser PRs and continuously providing
guidance throughout the project
- @LaBatata101 for initiating the transition to a hand-written parser in
#9152
- @addisoncrump for implementing the fuzzer which helped
[catch](https://github.com/astral-sh/ruff/pull/10903)
[a](https://github.com/astral-sh/ruff/pull/10910)
[lot](https://github.com/astral-sh/ruff/pull/10966)
[of](https://github.com/astral-sh/ruff/pull/10896)
[bugs](https://github.com/astral-sh/ruff/pull/10877)
---------
Co-authored-by: Victor Hugo Gomes <labatata101@linuxmail.org>
Co-authored-by: Micha Reiser <micha@reiser.io>
## Summary
This is a follow up on https://github.com/astral-sh/ruff/pull/10492
I incorrectly assumed that `subscript.value.end()` always points past
the value. However, this isn't the case for parenthesized values where
the end "ends" before the parentheses.
## Test Plan
I added new tests for the parenthesized case.
## Summary
This PR fixes an instability where formatting a subscribt
where the `slice` is not an `ExprSlice` and it has a trailing
end-of-line comment after its opening `[` required two formatting passes
to be stable.
The fix is to associate the trailing end-of-line comment as dangling
comment on `[` to preserve its position, similar to how Ruff does it for
other parenthesized expressions.
This also matches how trailing end-of-line subscript comments are
handled when the `slice` is an `ExprSlice`.
Fixes https://github.com/astral-sh/ruff/issues/10355
## Versioning
Shipping this as part of a patch release is fine because:
* It fixes a stability issue
* It doesn't impact already formatted code because Ruff would already
have moved to the comment to the end of the line (instead of preserving
it)
## Test Plan
Added tests
## Summary
I used `codespell` and `gramma` to identify mispellings and grammar
errors throughout the codebase and fixed them. I tried not to make any
controversial changes, but feel free to revert as you see fit.
## Summary
This PR changes how we format `with` statements with a single with item
for Python 3.8 or older. This change is not compatible with Black.
This is how we format a single-item with statement today
```python
def run(data_path, model_uri):
with pyspark.sql.SparkSession.builder.config(
key="spark.python.worker.reuse", value=True
).config(key="spark.ui.enabled", value=False).master(
"local-cluster[2, 1, 1024]"
).getOrCreate():
# ignore spark log output
spark.sparkContext.setLogLevel("OFF")
print(score_model(spark, data_path, model_uri))
```
This is different than how we would format the same expression if it is
inside any other clause header (`while`, `if`, ...):
```python
def run(data_path, model_uri):
while (
pyspark.sql.SparkSession.builder.config(
key="spark.python.worker.reuse", value=True
)
.config(key="spark.ui.enabled", value=False)
.master("local-cluster[2, 1, 1024]")
.getOrCreate()
):
# ignore spark log output
spark.sparkContext.setLogLevel("OFF")
print(score_model(spark, data_path, model_uri))
```
Which seems inconsistent to me.
This PR changes the formatting of the single-item with Python 3.8 or
older to match that of other clause headers.
```python
def run(data_path, model_uri):
with (
pyspark.sql.SparkSession.builder.config(
key="spark.python.worker.reuse", value=True
)
.config(key="spark.ui.enabled", value=False)
.master("local-cluster[2, 1, 1024]")
.getOrCreate()
):
# ignore spark log output
spark.sparkContext.setLogLevel("OFF")
print(score_model(spark, data_path, model_uri))
```
According to our versioning policy, this style change is gated behind a
preview flag.
## Test Plan
See added tests.
Added
## Summary
Fixes https://github.com/astral-sh/ruff/issues/10267
The issue with the current formatting is that the formatter flips
between the `SingleParenthesizedContextManager` and
`ParenthesizeIfExpands` or `SingleWithTarget` because the layouts use
incompatible formatting ( `SingleParenthesizedContextManager`:
`maybe_parenthesize_expression(context)` vs `ParenthesizeIfExpands`:
`parenthesize_if_expands(item)`, `SingleWithTarget`:
`optional_parentheses(item)`.
The fix is to ensure that the layouts between which the formatter flips
when adding or removing parentheses are the same. I do this by
introducing a new `SingleWithoutTarget` layout that uses the same
formatting as `SingleParenthesizedContextManager` if it has no target
and prefer `SingleWithoutTarget` over using `ParenthesizeIfExpands` or
`SingleWithTarget`.
## Formatting change
The downside is that we now use `maybe_parenthesize_expression` over
`parenthesize_if_expands` for expressions where
`can_omit_optional_parentheses` returns `false`. This can lead to stable
formatting changes. I only found one formatting change in our ecosystem
check and, unfortunately, this is necessary to fix the instability (and
instability fixes are okay to have as part of minor changes according to
our versioning policy)
The benefit of the change is that `with` items with a single context
manager and without a target are now formatted identically to how the
same expression would be formatted in other clause headers.
## Test Plan
I ran the ecosystem check locally
## Summary
Fixes the handling end of line comments that belong to `**kwargs` when
the `**kwargs` come after a slash.
The issue was that we missed to include the `**kwargs` start position
when determining the start of the next node coming after the `/`.
Fixes https://github.com/astral-sh/ruff/issues/10281
## Test Plan
Added test
## Summary
_This is preview only feature and is available using the `--preview`
command-line flag._
With the implementation of [PEP 701] in Python 3.12, f-strings can now
be broken into multiple lines, can contain comments, and can re-use the
same quote character. Currently, no other Python formatter formats the
f-strings so there's some discussion which needs to happen in defining
the style used for f-string formatting. Relevant discussion:
https://github.com/astral-sh/ruff/discussions/9785
The goal for this PR is to add minimal support for f-string formatting.
This would be to format expression within the replacement field without
introducing any major style changes.
### Newlines
The heuristics for adding newline is similar to that of
[Prettier](https://prettier.io/docs/en/next/rationale.html#template-literals)
where the formatter would only split an expression in the replacement
field across multiple lines if there was already a line break within the
replacement field.
In other words, the formatter would not add any newlines unless they
were already present i.e., they were added by the user. This makes
breaking any expression inside an f-string optional and in control of
the user. For example,
```python
# We wouldn't break this
aaaaaaaaaaa = f"asaaaaaaaaaaaaaaaa { aaaaaaaaaaaa + bbbbbbbbbbbb + ccccccccccccccc } cccccccccc"
# But, we would break the following as there's already a newline
aaaaaaaaaaa = f"asaaaaaaaaaaaaaaaa {
aaaaaaaaaaaa + bbbbbbbbbbbb + ccccccccccccccc } cccccccccc"
```
If there are comments in any of the replacement field of the f-string,
then it will always be a multi-line f-string in which case the formatter
would prefer to break expressions i.e., introduce newlines. For example,
```python
x = f"{ # comment
a }"
```
### Quotes
The logic for formatting quotes remains unchanged. The existing logic is
used to determine the necessary quote char and is used accordingly.
Now, if the expression inside an f-string is itself a string like, then
we need to make sure to preserve the existing quote and not change it to
the preferred quote unless it's 3.12. For example,
```python
f"outer {'inner'} outer"
# For pre 3.12, preserve the single quote
f"outer {'inner'} outer"
# While for 3.12 and later, the quotes can be changed
f"outer {"inner"} outer"
```
But, for triple-quoted strings, we can re-use the same quote char unless
the inner string is itself a triple-quoted string.
```python
f"""outer {"inner"} outer""" # valid
f"""outer {'''inner'''} outer""" # preserve the single quote char for the inner string
```
### Debug expressions
If debug expressions are present in the replacement field of a f-string,
then the whitespace needs to be preserved as they will be rendered as it
is (for example, `f"{ x = }"`. If there are any nested f-strings, then
the whitespace in them needs to be preserved as well which means that
we'll stop formatting the f-string as soon as we encounter a debug
expression.
```python
f"outer { x = !s :.3f}"
# ^^
# We can remove these whitespaces
```
Now, the whitespace doesn't need to be preserved around conversion spec
and format specifiers, so we'll format them as usual but we won't be
formatting any nested f-string within the format specifier.
### Miscellaneous
- The
[`hug_parens_with_braces_and_square_brackets`](https://github.com/astral-sh/ruff/issues/8279)
preview style isn't implemented w.r.t. the f-string curly braces.
- The
[indentation](https://github.com/astral-sh/ruff/discussions/9785#discussioncomment-8470590)
is always relative to the f-string containing statement
## Test Plan
* Add new test cases
* Review existing snapshot changes
* Review the ecosystem changes
[PEP 701]: https://peps.python.org/pep-0701/
## Summary
This PR implements the `blank_line_after_nested_stub_class` preview
style in the formatter.
The logic is divided into 3 parts:
1. In between preceding and following nodes at top level and nested
suite
2. When there's a trailing comment after the class
3. When there is no following node from (1) which is the case when it's
the last or the only node in a suite
We handle (3) with `FormatLeadingAlternateBranchComments`.
## Test Plan
- Add new test cases and update existing snapshots
- Checked the `typeshed` diff
fixes: #8891
Dylan
Brent Westbrook
Micha Reiser
Ibraheem Ahmed
K
Max Mynter
Alex Waygood
Dimitri Papadopoulos Orfanos
Dhruv Manilawala
yataka
Hamir Mahal
konsti
Jelle Zijlstra
Auguste Lalande
Shaygan Hooshyari