Empty commit to test performance

## Summary

Previously, we panicked on expressions like `f"{v:{f'0.2f'}}"` because
we did not infer types for expressions nested inside format spec
elements.

## Test Plan

```
cargo nextest run -p red_knot_workspace -- --ignored linter_af linter_gz
```

.editorconfig

@@ -17,4 +17,7 @@ indent_size = 4
 trim_trailing_whitespace = false
 
 [*.md]
-max_line_length = 100
+max_line_length = 100
+
+[*.toml]
+indent_size = 4

.github/workflows/ci.yaml

@@ -268,6 +268,7 @@ jobs:
       - uses: Swatinem/rust-cache@v2
         with:
           workspaces: "fuzz -> target"
+          cache-all-crates: "true"
       - name: "Install cargo-binstall"
         uses: cargo-bins/cargo-binstall@main
         with:

.github/workflows/publish-playground.yml

@@ -47,7 +47,7 @@ jobs:
         working-directory: playground
       - name: "Deploy to Cloudflare Pages"
         if: ${{ env.CF_API_TOKEN_EXISTS == 'true' }}
-        uses: cloudflare/wrangler-action@v3.11.0
+        uses: cloudflare/wrangler-action@v3.12.1
         with:
           apiToken: ${{ secrets.CF_API_TOKEN }}
           accountId: ${{ secrets.CF_ACCOUNT_ID }}

.pre-commit-config.yaml

@@ -17,7 +17,7 @@ exclude: |
 
 repos:
   - repo: https://github.com/abravalheri/validate-pyproject
-    rev: v0.22
+    rev: v0.23
     hooks:
       - id: validate-pyproject
 
@@ -53,13 +53,13 @@ repos:
         files: '^crates/.*/resources/mdtest/.*\.md'
         exclude: |
           (?x)^(
-            .*?invalid(_.+)_syntax.md
+            .*?invalid(_.+)*_syntax\.md
           )$
         additional_dependencies:
           - black==24.10.0
 
   - repo: https://github.com/crate-ci/typos
-    rev: v1.27.0
+    rev: v1.27.3
     hooks:
       - id: typos
 
@@ -73,7 +73,7 @@ repos:
         pass_filenames: false # This makes it a lot faster
 
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.7.2
+    rev: v0.7.4
     hooks:
       - id: ruff-format
       - id: ruff

CHANGELOG.md

@@ -1,5 +1,54 @@
 # Changelog
 
+## 0.7.4
+
+### Preview features
+
+- \[`flake8-datetimez`\] Detect usages of `datetime.max`/`datetime.min` (`DTZ901`) ([#14288](https://github.com/astral-sh/ruff/pull/14288))
+- \[`flake8-logging`\] Implement `root-logger-calls` (`LOG015`) ([#14302](https://github.com/astral-sh/ruff/pull/14302))
+- \[`flake8-no-pep420`\] Detect empty implicit namespace packages (`INP001`) ([#14236](https://github.com/astral-sh/ruff/pull/14236))
+- \[`flake8-pyi`\] Add "replace with `Self`" fix (`PYI019`) ([#14238](https://github.com/astral-sh/ruff/pull/14238))
+- \[`perflint`\] Implement quick-fix for `manual-list-comprehension` (`PERF401`) ([#13919](https://github.com/astral-sh/ruff/pull/13919))
+- \[`pylint`\] Implement `shallow-copy-environ` (`W1507`) ([#14241](https://github.com/astral-sh/ruff/pull/14241))
+- \[`ruff`\] Implement `none-not-at-end-of-union` (`RUF036`) ([#14314](https://github.com/astral-sh/ruff/pull/14314))
+- \[`ruff`\] Implementation `unsafe-markup-call` from `flake8-markupsafe` plugin (`RUF035`) ([#14224](https://github.com/astral-sh/ruff/pull/14224))
+- \[`ruff`\] Report problems for `attrs` dataclasses (`RUF008`, `RUF009`) ([#14327](https://github.com/astral-sh/ruff/pull/14327))
+
+### Rule changes
+
+- \[`flake8-boolean-trap`\] Exclude dunder methods that define operators (`FBT001`) ([#14203](https://github.com/astral-sh/ruff/pull/14203))
+- \[`flake8-pyi`\] Add "replace with `Self`" fix (`PYI034`) ([#14217](https://github.com/astral-sh/ruff/pull/14217))
+- \[`flake8-pyi`\] Always autofix `duplicate-union-members` (`PYI016`) ([#14270](https://github.com/astral-sh/ruff/pull/14270))
+- \[`flake8-pyi`\] Improve autofix for nested and mixed type unions for `unnecessary-type-union` (`PYI055`) ([#14272](https://github.com/astral-sh/ruff/pull/14272))
+- \[`flake8-pyi`\] Mark fix as unsafe when type annotation contains comments for `duplicate-literal-member` (`PYI062`) ([#14268](https://github.com/astral-sh/ruff/pull/14268))
+
+### Server
+
+- Use the current working directory to resolve settings from `ruff.configuration` ([#14352](https://github.com/astral-sh/ruff/pull/14352))
+
+### Bug fixes
+
+- Avoid conflicts between `PLC014` (`useless-import-alias`) and `I002` (`missing-required-import`) by considering `lint.isort.required-imports` for `PLC014` ([#14287](https://github.com/astral-sh/ruff/pull/14287))
+- \[`flake8-type-checking`\] Skip quoting annotation if it becomes invalid syntax (`TCH001`)
+- \[`flake8-pyi`\] Avoid using `typing.Self` in stub files pre-Python 3.11 (`PYI034`) ([#14230](https://github.com/astral-sh/ruff/pull/14230))
+- \[`flake8-pytest-style`\] Flag `pytest.raises` call with keyword argument `expected_exception` (`PT011`) ([#14298](https://github.com/astral-sh/ruff/pull/14298))
+- \[`flake8-simplify`\] Infer "unknown" truthiness for literal iterables whose items are all unpacks (`SIM222`) ([#14263](https://github.com/astral-sh/ruff/pull/14263))
+- \[`flake8-type-checking`\] Fix false positives for `typing.Annotated` (`TCH001`) ([#14311](https://github.com/astral-sh/ruff/pull/14311))
+- \[`pylint`\] Allow `await` at the top-level scope of a notebook (`PLE1142`) ([#14225](https://github.com/astral-sh/ruff/pull/14225))
+- \[`pylint`\] Fix miscellaneous issues in `await-outside-async` detection (`PLE1142`) ([#14218](https://github.com/astral-sh/ruff/pull/14218))
+- \[`pyupgrade`\] Avoid applying PEP 646 rewrites in invalid contexts (`UP044`) ([#14234](https://github.com/astral-sh/ruff/pull/14234))
+- \[`pyupgrade`\] Detect permutations in redundant open modes (`UP015`) ([#14255](https://github.com/astral-sh/ruff/pull/14255))
+- \[`refurb`\] Avoid triggering `hardcoded-string-charset` for reordered sets (`FURB156`) ([#14233](https://github.com/astral-sh/ruff/pull/14233))
+- \[`refurb`\] Further special cases added to `verbose-decimal-constructor` (`FURB157`) ([#14216](https://github.com/astral-sh/ruff/pull/14216))
+- \[`refurb`\] Use `UserString` instead of non-existent `UserStr` (`FURB189`) ([#14209](https://github.com/astral-sh/ruff/pull/14209))
+- \[`ruff`\] Avoid treating lowercase letters as `# noqa` codes (`RUF100`) ([#14229](https://github.com/astral-sh/ruff/pull/14229))
+- \[`ruff`\] Do not report when `Optional` has no type arguments (`RUF013`) ([#14181](https://github.com/astral-sh/ruff/pull/14181))
+
+### Documentation
+
+- Add "Notebook behavior" section for `F704`, `PLE1142` ([#14266](https://github.com/astral-sh/ruff/pull/14266))
+- Document comment policy around fix safety ([#14300](https://github.com/astral-sh/ruff/pull/14300))
+
 ## 0.7.3
 
 ### Preview features

Cargo.lock

@@ -2,12 +2,6 @@
 # It is not intended for manual editing.
 version = 3
 
-[[package]]
-name = "Inflector"
-version = "0.11.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fe438c63458706e03479442743baae6c88256498e6431708f6dfc520a26515d3"
-
 [[package]]
 name = "adler"
 version = "1.0.2"
@@ -123,9 +117,9 @@ dependencies = [
 
 [[package]]
 name = "anyhow"
-version = "1.0.92"
+version = "1.0.93"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "74f37166d7d48a0284b99dd824694c26119c700b53bf0d1540cdb147dbdaaf13"
+checksum = "4c95c10ba0b00a02636238b814946408b1322d5ac4760326e6fb8ec956d85775"
 
 [[package]]
 name = "append-only-vec"
@@ -176,6 +170,12 @@ version = "1.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "f1fdabc7756949593fe60f30ec81974b613357de856987752631dea1e3394c80"
 
+[[package]]
+name = "base64"
+version = "0.13.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9e1b586273c5702936fe7b7d6896644d8be71e6314cfe09d3167c95f712589e8"
+
 [[package]]
 name = "base64"
 version = "0.22.0"
@@ -214,9 +214,9 @@ dependencies = [
 
 [[package]]
 name = "bstr"
-version = "1.10.0"
+version = "1.11.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "40723b8fb387abc38f4f4a37c09073622e41dd12327033091ef8950659e6dc0c"
+checksum = "1a68f1f47cdf0ec8ee4b941b2eee2a80cb796db73118c0dd09ac63fbe405be22"
 dependencies = [
  "memchr",
  "regex-automata 0.4.8",
@@ -347,9 +347,9 @@ dependencies = [
 
 [[package]]
 name = "clap"
-version = "4.5.20"
+version = "4.5.21"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b97f376d85a664d5837dbae44bf546e6477a679ff6610010f17276f686d867e8"
+checksum = "fb3b4b9e5a7c7514dfa52869339ee98b3156b0bfb4e8a77c4ff4babb64b1604f"
 dependencies = [
  "clap_builder",
  "clap_derive",
@@ -357,9 +357,9 @@ dependencies = [
 
 [[package]]
 name = "clap_builder"
-version = "4.5.20"
+version = "4.5.21"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "19bc80abd44e4bed93ca373a0704ccbd1b710dc5749406201bb018272808dc54"
+checksum = "b17a95aa67cc7b5ebd32aa5370189aa0d79069ef1c64ce893bd30fb24bff20ec"
 dependencies = [
  "anstream",
  "anstyle",
@@ -424,7 +424,7 @@ checksum = "2f8c93eb5f77c9050c7750e14f13ef1033a40a0aac70c6371535b6763a01438c"
 dependencies = [
  "nix 0.28.0",
  "terminfo",
- "thiserror",
+ "thiserror 1.0.67",
  "which",
  "winapi",
 ]
@@ -812,6 +812,17 @@ dependencies = [
  "windows-sys 0.48.0",
 ]
 
+[[package]]
+name = "displaydoc"
+version = "0.2.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "97369cbbc041bc366949bc74d34658d6cda5621039731c6310521892a3a20ae0"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn 2.0.87",
+]
+
 [[package]]
 name = "doc-comment"
 version = "0.3.3"
@@ -824,6 +835,12 @@ version = "0.1.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "9bda8e21c04aca2ae33ffc2fd8c23134f3cac46db123ba97bd9d3f3b8a4a85e1"
 
+[[package]]
+name = "dunce"
+version = "1.0.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "92773504d58c093f6de2459af4af33faa518c13451eb8f2b5698ed3d36e7c813"
+
 [[package]]
 name = "dyn-clone"
 version = "1.0.17"
@@ -1045,9 +1062,9 @@ dependencies = [
 
 [[package]]
 name = "hashbrown"
-version = "0.15.0"
+version = "0.15.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1e087f84d4f86bf4b218b927129862374b72199ae7d8657835f1e89000eea4fb"
+checksum = "3a9bfc1af68b1726ea47d3d5109de126281def866b33970e10fbab11b5dafab3"
 
 [[package]]
 name = "hashlink"
@@ -1108,6 +1125,124 @@ dependencies = [
  "cc",
 ]
 
+[[package]]
+name = "icu_collections"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "db2fa452206ebee18c4b5c2274dbf1de17008e874b4dc4f0aea9d01ca79e4526"
+dependencies = [
+ "displaydoc",
+ "yoke",
+ "zerofrom",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_locid"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "13acbb8371917fc971be86fc8057c41a64b521c184808a698c02acc242dbf637"
+dependencies = [
+ "displaydoc",
+ "litemap",
+ "tinystr",
+ "writeable",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_locid_transform"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "01d11ac35de8e40fdeda00d9e1e9d92525f3f9d887cdd7aa81d727596788b54e"
+dependencies = [
+ "displaydoc",
+ "icu_locid",
+ "icu_locid_transform_data",
+ "icu_provider",
+ "tinystr",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_locid_transform_data"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "fdc8ff3388f852bede6b579ad4e978ab004f139284d7b28715f773507b946f6e"
+
+[[package]]
+name = "icu_normalizer"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "19ce3e0da2ec68599d193c93d088142efd7f9c5d6fc9b803774855747dc6a84f"
+dependencies = [
+ "displaydoc",
+ "icu_collections",
+ "icu_normalizer_data",
+ "icu_properties",
+ "icu_provider",
+ "smallvec",
+ "utf16_iter",
+ "utf8_iter",
+ "write16",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_normalizer_data"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f8cafbf7aa791e9b22bec55a167906f9e1215fd475cd22adfcf660e03e989516"
+
+[[package]]
+name = "icu_properties"
+version = "1.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "93d6020766cfc6302c15dbbc9c8778c37e62c14427cb7f6e601d849e092aeef5"
+dependencies = [
+ "displaydoc",
+ "icu_collections",
+ "icu_locid_transform",
+ "icu_properties_data",
+ "icu_provider",
+ "tinystr",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_properties_data"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "67a8effbc3dd3e4ba1afa8ad918d5684b8868b3b26500753effea8d2eed19569"
+
+[[package]]
+name = "icu_provider"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6ed421c8a8ef78d3e2dbc98a973be2f3770cb42b606e3ab18d6237c4dfde68d9"
+dependencies = [
+ "displaydoc",
+ "icu_locid",
+ "icu_provider_macros",
+ "stable_deref_trait",
+ "tinystr",
+ "writeable",
+ "yoke",
+ "zerofrom",
+ "zerovec",
+]
+
+[[package]]
+name = "icu_provider_macros"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1ec89e9337638ecdc08744df490b221a7399bf8d164eb52a665454e60e075ad6"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn 2.0.87",
+]
+
 [[package]]
 name = "ident_case"
 version = "1.0.1"
@@ -1116,12 +1251,23 @@ checksum = "b9e0384b61958566e926dc50660321d12159025e767c18e043daf26b70104c39"
 
 [[package]]
 name = "idna"
-version = "0.5.0"
+version = "1.0.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "634d9b1461af396cad843f47fdba5597a4f9e6ddd4bfb6ff5d85028c25cb12f6"
+checksum = "686f825264d630750a544639377bae737628043f20d38bbc029e8f29ea968a7e"
 dependencies = [
- "unicode-bidi",
- "unicode-normalization",
+ "idna_adapter",
+ "smallvec",
+ "utf8_iter",
+]
+
+[[package]]
+name = "idna_adapter"
+version = "1.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "daca1df1c957320b2cf139ac61e7bd64fed304c5040df000a745aa1de3b4ef71"
+dependencies = [
+ "icu_normalizer",
+ "icu_properties",
 ]
 
 [[package]]
@@ -1167,22 +1313,22 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "707907fe3c25f5424cce2cb7e1cbcafee6bdbe735ca90ef77c29e84591e5b9da"
 dependencies = [
  "equivalent",
- "hashbrown 0.15.0",
+ "hashbrown 0.15.1",
  "serde",
 ]
 
 [[package]]
 name = "indicatif"
-version = "0.17.8"
+version = "0.17.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "763a5a8f45087d6bcea4222e7b72c291a054edf80e4ef6efd2a4979878c7bea3"
+checksum = "cbf675b85ed934d3c67b5c5469701eec7db22689d0a2139d856e0925fa28b281"
 dependencies = [
  "console",
- "instant",
  "number_prefix",
  "portable-atomic",
- "unicode-width 0.1.13",
+ "unicode-width 0.2.0",
  "vt100",
+ "web-time",
 ]
 
 [[package]]
@@ -1224,6 +1370,7 @@ dependencies = [
  "pest",
  "pest_derive",
  "regex",
+ "ron",
  "serde",
  "similar",
  "walkdir",
@@ -1260,11 +1407,11 @@ dependencies = [
 
 [[package]]
 name = "is-macro"
-version = "0.3.6"
+version = "0.3.7"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2069faacbe981460232f880d26bf3c7634e322d49053aa48c27e3ae642f728f1"
+checksum = "1d57a3e447e24c22647738e4607f1df1e0ec6f72e16182c4cd199f647cdfb0e4"
 dependencies = [
- "Inflector",
+ "heck",
  "proc-macro2",
  "quote",
  "syn 2.0.87",
@@ -1367,9 +1514,9 @@ checksum = "e2abad23fbc42b3700f2f279844dc832adb2b2eb069b2df918f455c4e18cc646"
 
 [[package]]
 name = "libc"
-version = "0.2.161"
+version = "0.2.164"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8e9489c2807c139ffd9c1794f4af0ebe86a828db53ecdc7fea2111d0fed085d1"
+checksum = "433bfe06b8c75da9b2e3fbea6e5329ff87748f0b144ef75306e674c3f6f7c13f"
 
 [[package]]
 name = "libcst"
@@ -1383,7 +1530,7 @@ dependencies = [
  "paste",
  "peg",
  "regex",
- "thiserror",
+ "thiserror 1.0.67",
 ]
 
 [[package]]
@@ -1429,6 +1576,12 @@ version = "0.4.14"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "78b3ae25bc7c8c38cec158d1f2757ee79e9b3740fbc7ccf0e59e4b08d793fa89"
 
+[[package]]
+name = "litemap"
+version = "0.7.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "643cb0b8d4fcc284004d5fd0d67ccf61dfffadb7f75e1e71bc420f4688a3a704"
+
 [[package]]
 name = "lock_api"
 version = "0.4.11"
@@ -1486,9 +1639,9 @@ checksum = "2532096657941c2fea9c289d370a250971c689d4f143798ff67113ec042024a5"
 
 [[package]]
 name = "matchit"
-version = "0.8.4"
+version = "0.8.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "47e1ffaa40ddd1f3ed91f717a33c8c0ee23fff369e3aa8772b9605cc1d22f4c3"
+checksum = "bd0aa4b8ca861b08d68afc8702af3250776898c1508b278e1da9d01e01d4b45c"
 
 [[package]]
 name = "memchr"
@@ -1815,7 +1968,7 @@ dependencies = [
  "pep440_rs 0.4.0",
  "regex",
  "serde",
- "thiserror",
+ "thiserror 1.0.67",
  "tracing",
  "unicode-width 0.1.13",
  "url",
@@ -1834,7 +1987,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "cd53dff83f26735fdc1ca837098ccf133605d794cdae66acfc2bfac3ec809d95"
 dependencies = [
  "memchr",
- "thiserror",
+ "thiserror 1.0.67",
  "ucd-trie",
 ]
 
@@ -2004,7 +2157,7 @@ dependencies = [
  "newtype-uuid",
  "quick-xml",
  "strip-ansi-escapes",
- "thiserror",
+ "thiserror 1.0.67",
  "uuid",
 ]
 
@@ -2111,7 +2264,7 @@ dependencies = [
  "compact_str",
  "countme",
  "dir-test",
- "hashbrown 0.15.0",
+ "hashbrown 0.15.1",
  "indexmap",
  "insta",
  "itertools 0.13.0",
@@ -2129,11 +2282,12 @@ dependencies = [
  "ruff_text_size",
  "rustc-hash 2.0.0",
  "salsa",
+ "serde",
  "smallvec",
  "static_assertions",
  "tempfile",
  "test-case",
- "thiserror",
+ "thiserror 2.0.3",
  "tracing",
 ]
 
@@ -2213,7 +2367,10 @@ version = "0.0.0"
 dependencies = [
  "anyhow",
  "crossbeam",
+ "glob",
+ "insta",
  "notify",
+ "pep440_rs 0.7.2",
  "rayon",
  "red_knot_python_semantic",
  "red_knot_vendored",
@@ -2223,7 +2380,9 @@ dependencies = [
  "ruff_text_size",
  "rustc-hash 2.0.0",
  "salsa",
- "tempfile",
+ "serde",
+ "thiserror 2.0.3",
+ "toml",
  "tracing",
 ]
 
@@ -2253,7 +2412,7 @@ checksum = "bd283d9651eeda4b2a83a43c1c91b266c40fd76ecd39a50a8c630ae69dc72891"
 dependencies = [
  "getrandom",
  "libredox",
- "thiserror",
+ "thiserror 1.0.67",
 ]
 
 [[package]]
@@ -2315,9 +2474,20 @@ dependencies = [
  "windows-sys 0.52.0",
 ]
 
+[[package]]
+name = "ron"
+version = "0.7.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "88073939a61e5b7680558e6be56b419e208420c2adb92be54921fa6b72283f1a"
+dependencies = [
+ "base64 0.13.1",
+ "bitflags 1.3.2",
+ "serde",
+]
+
 [[package]]
 name = "ruff"
-version = "0.7.3"
+version = "0.7.4"
 dependencies = [
  "anyhow",
  "argfile",
@@ -2364,7 +2534,7 @@ dependencies = [
  "strum",
  "tempfile",
  "test-case",
- "thiserror",
+ "thiserror 2.0.3",
  "tikv-jemallocator",
  "toml",
  "tracing",
@@ -2416,7 +2586,9 @@ dependencies = [
  "camino",
  "countme",
  "dashmap 6.1.0",
+ "dunce",
  "filetime",
+ "glob",
  "ignore",
  "insta",
  "matchit",
@@ -2432,7 +2604,7 @@ dependencies = [
  "salsa",
  "serde",
  "tempfile",
- "thiserror",
+ "thiserror 2.0.3",
  "tracing",
  "tracing-subscriber",
  "tracing-tree",
@@ -2534,7 +2706,7 @@ dependencies = [
 
 [[package]]
 name = "ruff_linter"
-version = "0.7.3"
+version = "0.7.4"
 dependencies = [
  "aho-corasick",
  "annotate-snippets 0.9.2",
@@ -2584,7 +2756,7 @@ dependencies = [
  "strum",
  "strum_macros",
  "test-case",
- "thiserror",
+ "thiserror 2.0.3",
  "toml",
  "typed-arena",
  "unicode-normalization",
@@ -2618,7 +2790,7 @@ dependencies = [
  "serde_json",
  "serde_with",
  "test-case",
- "thiserror",
+ "thiserror 2.0.3",
  "uuid",
 ]
 
@@ -2638,7 +2810,6 @@ dependencies = [
  "ruff_source_file",
  "ruff_text_size",
  "rustc-hash 2.0.0",
- "salsa",
  "schemars",
  "serde",
 ]
@@ -2691,7 +2862,7 @@ dependencies = [
  "similar",
  "smallvec",
  "static_assertions",
- "thiserror",
+ "thiserror 2.0.3",
  "tracing",
 ]
 
@@ -2771,6 +2942,7 @@ dependencies = [
 name = "ruff_python_stdlib"
 version = "0.0.0"
 dependencies = [
+ "bitflags 2.6.0",
  "unicode-ident",
 ]
 
@@ -2823,7 +2995,7 @@ dependencies = [
  "serde",
  "serde_json",
  "shellexpand",
- "thiserror",
+ "thiserror 2.0.3",
  "tracing",
  "tracing-subscriber",
 ]
@@ -2849,7 +3021,7 @@ dependencies = [
 
 [[package]]
 name = "ruff_wasm"
-version = "0.7.3"
+version = "0.7.4"
 dependencies = [
  "console_error_panic_hook",
  "console_log",
@@ -2932,9 +3104,9 @@ checksum = "583034fd73374156e66797ed8e5b0d5690409c9226b22d87cb7f19821c05d152"
 
 [[package]]
 name = "rustix"
-version = "0.38.37"
+version = "0.38.40"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8acb788b847c24f28525660c4d7758620a7210875711f79e7f663cc152726811"
+checksum = "99e4ea3e1cdc4b559b8e5650f9c8e5998e3e5c1343b4eaf034565f32318d63c0"
 dependencies = [
  "bitflags 2.6.0",
  "errno",
@@ -3077,9 +3249,9 @@ checksum = "1c107b6f4780854c8b126e228ea8869f4d7b71260f962fefb57b996b8959ba6b"
 
 [[package]]
 name = "serde"
-version = "1.0.214"
+version = "1.0.215"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f55c3193aca71c12ad7890f1785d2b73e1b9f63a0bbc353c08ef26fe03fc56b5"
+checksum = "6513c1ad0b11a9376da888e3e0baa0077f1aed55c17f50e7b2397136129fb88f"
 dependencies = [
  "serde_derive",
 ]
@@ -3097,9 +3269,9 @@ dependencies = [
 
 [[package]]
 name = "serde_derive"
-version = "1.0.214"
+version = "1.0.215"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "de523f781f095e28fa605cdce0f8307e451cc0fd14e2eb4cd2e98a355b147766"
+checksum = "ad1e866f866923f252f05c889987993144fb74e722403468a4ebd70c3cd756c0"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -3119,9 +3291,9 @@ dependencies = [
 
 [[package]]
 name = "serde_json"
-version = "1.0.132"
+version = "1.0.133"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d726bfaff4b320266d395898905d0eba0345aae23b54aee3a737e260fd46db03"
+checksum = "c7fceb2473b9166b2294ef05efcb65a3db80803f0b03ef86a5fc88a2b85ee377"
 dependencies = [
  "itoa",
  "memchr",
@@ -3234,6 +3406,12 @@ version = "0.9.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "6980e8d7511241f8acf4aebddbb1ff938df5eebe98691418c4468d0b72a96a67"
 
+[[package]]
+name = "stable_deref_trait"
+version = "1.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a8f112729512f8e442d81f95a8a7ddf2b7c6b8a1a6f509a95864142b30cab2d3"
+
 [[package]]
 name = "static_assertions"
 version = "1.1.0"
@@ -3324,9 +3502,9 @@ dependencies = [
 
 [[package]]
 name = "tempfile"
-version = "3.13.0"
+version = "3.14.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f0f2c9fc62d0beef6951ccffd757e241266a2c833136efbe35af6cd2567dca5b"
+checksum = "28cce251fcbc87fac86a866eeb0d6c2d536fc16d06f184bb61aeae11aa4cee0c"
 dependencies = [
  "cfg-if",
  "fastrand",
@@ -3403,7 +3581,16 @@ version = "1.0.67"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "3b3c6efbfc763e64eb85c11c25320f0737cb7364c4b6336db90aa9ebe27a0bbd"
 dependencies = [
- "thiserror-impl",
+ "thiserror-impl 1.0.67",
+]
+
+[[package]]
+name = "thiserror"
+version = "2.0.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c006c85c7651b3cf2ada4584faa36773bd07bac24acfb39f3c431b36d7e667aa"
+dependencies = [
+ "thiserror-impl 2.0.3",
 ]
 
 [[package]]
@@ -3417,6 +3604,17 @@ dependencies = [
  "syn 2.0.87",
 ]
 
+[[package]]
+name = "thiserror-impl"
+version = "2.0.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f077553d607adc1caf65430528a576c757a71ed73944b66ebb58ef2bbd243568"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn 2.0.87",
+]
+
 [[package]]
 name = "thread_local"
 version = "1.1.8"
@@ -3447,6 +3645,16 @@ dependencies = [
  "tikv-jemalloc-sys",
 ]
 
+[[package]]
+name = "tinystr"
+version = "0.7.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9117f5d4db391c1cf6927e7bea3db74b9a1c1add8f7eda9ffd5364f40f57b82f"
+dependencies = [
+ "displaydoc",
+ "zerovec",
+]
+
 [[package]]
 name = "tinytemplate"
 version = "1.2.1"
@@ -3663,12 +3871,6 @@ dependencies = [
  "unic-common",
 ]
 
-[[package]]
-name = "unicode-bidi"
-version = "0.3.15"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "08f95100a766bf4f8f28f90d77e0a5461bbdb219042e7679bebe79004fed8d75"
-
 [[package]]
 name = "unicode-ident"
 version = "1.0.13"
@@ -3736,7 +3938,7 @@ version = "2.10.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b74fc6b57825be3373f7054754755f03ac3a8f5d70015ccad699ba2029956f4a"
 dependencies = [
- "base64",
+ "base64 0.22.0",
  "flate2",
  "log",
  "once_cell",
@@ -3748,9 +3950,9 @@ dependencies = [
 
 [[package]]
 name = "url"
-version = "2.5.2"
+version = "2.5.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "22784dbdf76fdde8af1aeda5622b546b422b6fc585325248a2bf9f5e41e94d6c"
+checksum = "8d157f1b96d14500ffdc1f10ba712e780825526c03d9a49b4d0324b0d9113ada"
 dependencies = [
  "form_urlencoded",
  "idna",
@@ -3758,6 +3960,18 @@ dependencies = [
  "serde",
 ]
 
+[[package]]
+name = "utf16_iter"
+version = "1.0.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c8232dd3cdaed5356e0f716d285e4b40b932ac434100fe9b7e0e8e935b9e6246"
+
+[[package]]
+name = "utf8_iter"
+version = "1.0.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b6c140620e7ffbb22c2dee59cafe6084a59b5ffc27a8859a5f0d494b5d52b6be"
+
 [[package]]
 name = "utf8parse"
 version = "0.2.1"
@@ -4194,6 +4408,18 @@ version = "0.0.19"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d135d17ab770252ad95e9a872d365cf3090e3be864a34ab46f48555993efc904"
 
+[[package]]
+name = "write16"
+version = "1.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d1890f4022759daae28ed4fe62859b1236caebfc61ede2f63ed4e695f3f6d936"
+
+[[package]]
+name = "writeable"
+version = "0.5.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1e9df38ee2d2c3c5948ea468a8406ff0db0b29ae1ffde1bcf20ef305bcc95c51"
+
 [[package]]
 name = "yansi"
 version = "1.0.1"
@@ -4209,6 +4435,30 @@ dependencies = [
  "winapi",
 ]
 
+[[package]]
+name = "yoke"
+version = "0.7.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6c5b1314b079b0930c31e3af543d8ee1757b1951ae1e1565ec704403a7240ca5"
+dependencies = [
+ "serde",
+ "stable_deref_trait",
+ "yoke-derive",
+ "zerofrom",
+]
+
+[[package]]
+name = "yoke-derive"
+version = "0.7.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "28cc31741b18cb6f1d5ff12f5b7523e3d6eb0852bbbad19d73905511d9849b95"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn 2.0.87",
+ "synstructure",
+]
+
 [[package]]
 name = "zerocopy"
 version = "0.7.32"
@@ -4229,12 +4479,55 @@ dependencies = [
  "syn 2.0.87",
 ]
 
+[[package]]
+name = "zerofrom"
+version = "0.1.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "91ec111ce797d0e0784a1116d0ddcdbea84322cd79e5d5ad173daeba4f93ab55"
+dependencies = [
+ "zerofrom-derive",
+]
+
+[[package]]
+name = "zerofrom-derive"
+version = "0.1.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0ea7b4a3637ea8669cedf0f1fd5c286a17f3de97b8dd5a70a6c167a1730e63a5"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn 2.0.87",
+ "synstructure",
+]
+
 [[package]]
 name = "zeroize"
 version = "1.7.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "525b4ec142c6b68a2d10f01f7bbf6755599ca3f81ea53b8431b7dd348f5fdb2d"
 
+[[package]]
+name = "zerovec"
+version = "0.10.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "aa2b893d79df23bfb12d5461018d408ea19dfafe76c2c7ef6d4eba614f8ff079"
+dependencies = [
+ "yoke",
+ "zerofrom",
+ "zerovec-derive",
+]
+
+[[package]]
+name = "zerovec-derive"
+version = "0.10.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6eafa6dfb17584ea3e2bd6e76e0cc15ad7af12b09abdd1ca55961bed9b1063c6"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn 2.0.87",
+]
+
 [[package]]
 name = "zip"
 version = "0.6.6"

Cargo.toml

@@ -66,6 +66,7 @@ criterion = { version = "0.5.1", default-features = false }
 crossbeam = { version = "0.8.4" }
 dashmap = { version = "6.0.1" }
 dir-test = { version = "0.3.0" }
+dunce = { version = "1.0.5" }
 drop_bomb = { version = "0.1.5" }
 env_logger = { version = "0.11.0" }
 etcetera = { version = "0.8.0" }
@@ -81,7 +82,7 @@ hashbrown = { version = "0.15.0", default-features = false, features = [
 ignore = { version = "0.4.22" }
 imara-diff = { version = "0.1.5" }
 imperative = { version = "1.0.4" }
-indexmap = {version = "2.6.0" }
+indexmap = { version = "2.6.0" }
 indicatif = { version = "0.17.8" }
 indoc = { version = "2.0.4" }
 insta = { version = "1.35.1" }
@@ -136,7 +137,7 @@ strum_macros = { version = "0.26.0" }
 syn = { version = "2.0.55" }
 tempfile = { version = "3.9.0" }
 test-case = { version = "3.3.1" }
-thiserror = { version = "1.0.58" }
+thiserror = { version = "2.0.0" }
 tikv-jemallocator = { version = "0.6.0" }
 toml = { version = "0.8.11" }
 tracing = { version = "0.1.40" }

README.md

@@ -136,8 +136,8 @@ curl -LsSf https://astral.sh/ruff/install.sh | sh
 powershell -c "irm https://astral.sh/ruff/install.ps1 | iex"
 
 # For a specific version.
-curl -LsSf https://astral.sh/ruff/0.7.3/install.sh | sh
-powershell -c "irm https://astral.sh/ruff/0.7.3/install.ps1 | iex"
+curl -LsSf https://astral.sh/ruff/0.7.4/install.sh | sh
+powershell -c "irm https://astral.sh/ruff/0.7.4/install.ps1 | iex"
 ```
 
 You can also install Ruff via [Homebrew](https://formulae.brew.sh/formula/ruff), [Conda](https://anaconda.org/conda-forge/ruff),
@@ -170,7 +170,7 @@ Ruff can also be used as a [pre-commit](https://pre-commit.com/) hook via [`ruff
 ```yaml
 - repo: https://github.com/astral-sh/ruff-pre-commit
   # Ruff version.
-  rev: v0.7.3
+  rev: v0.7.4
   hooks:
     # Run the linter.
     - id: ruff

_typos.toml

@@ -1,6 +1,11 @@
 [files]
 # https://github.com/crate-ci/typos/issues/868
-extend-exclude = ["crates/red_knot_vendored/vendor/**/*", "**/resources/**/*", "**/snapshots/**/*"]
+extend-exclude = [
+  "crates/red_knot_vendored/vendor/**/*",
+  "**/resources/**/*",
+  "**/snapshots/**/*",
+  "crates/red_knot_workspace/src/workspace/pyproject/package_name.rs"
+]
 
 [default.extend-words]
 "arange" = "arange"  # e.g. `numpy.arange`

crates/red_knot/Cargo.toml

@@ -34,6 +34,7 @@ tracing-tree = { workspace = true }
 [dev-dependencies]
 filetime = { workspace = true }
 tempfile = { workspace = true }
+ruff_db = { workspace = true, features = ["testing"] }
 
 [lints]
 workspace = true

crates/red_knot/src/main.rs

@@ -183,10 +183,10 @@ fn run() -> anyhow::Result<ExitStatus> {
 
     let system = OsSystem::new(cwd.clone());
     let cli_configuration = args.to_configuration(&cwd);
-    let workspace_metadata = WorkspaceMetadata::from_path(
+    let workspace_metadata = WorkspaceMetadata::discover(
         system.current_directory(),
         &system,
-        Some(cli_configuration.clone()),
+        Some(&cli_configuration),
     )?;
 
     // TODO: Use the `program_settings` to compute the key for the database's persistent

crates/red_knot/src/target_version.rs

@@ -4,8 +4,8 @@
 #[derive(Copy, Clone, Hash, Debug, PartialEq, Eq, PartialOrd, Ord, Default, clap::ValueEnum)]
 pub enum TargetVersion {
     Py37,
-    #[default]
     Py38,
+    #[default]
     Py39,
     Py310,
     Py311,
@@ -46,3 +46,17 @@ impl From<TargetVersion> for red_knot_python_semantic::PythonVersion {
         }
     }
 }
+
+#[cfg(test)]
+mod tests {
+    use crate::target_version::TargetVersion;
+    use red_knot_python_semantic::PythonVersion;
+
+    #[test]
+    fn same_default_as_python_version() {
+        assert_eq!(
+            PythonVersion::from(TargetVersion::default()),
+            PythonVersion::default()
+        );
+    }
+}

crates/red_knot/tests/file_watching.rs

@@ -6,7 +6,7 @@ use std::time::Duration;
 use anyhow::{anyhow, Context};
 
 use red_knot_python_semantic::{resolve_module, ModuleName, Program, PythonVersion, SitePackages};
-use red_knot_workspace::db::RootDatabase;
+use red_knot_workspace::db::{Db, RootDatabase};
 use red_knot_workspace::watch;
 use red_knot_workspace::watch::{directory_watcher, WorkspaceWatcher};
 use red_knot_workspace::workspace::settings::{Configuration, SearchPathConfiguration};
@@ -14,6 +14,7 @@ use red_knot_workspace::workspace::WorkspaceMetadata;
 use ruff_db::files::{system_path_to_file, File, FileError};
 use ruff_db::source::source_text;
 use ruff_db::system::{OsSystem, SystemPath, SystemPathBuf};
+use ruff_db::testing::setup_logging;
 use ruff_db::Upcast;
 
 struct TestCase {
@@ -69,7 +70,6 @@ impl TestCase {
         Some(all_events)
     }
 
-    #[cfg(unix)]
     fn take_watch_changes(&self) -> Vec<watch::ChangeEvent> {
         self.try_take_watch_changes(Duration::from_secs(10))
             .expect("Expected watch changes but observed none")
@@ -110,8 +110,8 @@ impl TestCase {
     ) -> anyhow::Result<()> {
         let program = Program::get(self.db());
 
-        self.configuration.search_paths = configuration.clone();
-        let new_settings = configuration.into_settings(self.db.workspace().root(&self.db));
+        let new_settings = configuration.to_settings(self.db.workspace().root(&self.db));
+        self.configuration.search_paths = configuration;
 
         program.update_search_paths(&mut self.db, &new_settings)?;
 
@@ -204,7 +204,9 @@ where
             .as_utf8_path()
             .canonicalize_utf8()
             .with_context(|| "Failed to canonicalize root path.")?,
-    );
+    )
+    .simplified()
+    .to_path_buf();
 
     let workspace_path = root_path.join("workspace");
 
@@ -241,8 +243,7 @@ where
         search_paths,
     };
 
-    let workspace =
-        WorkspaceMetadata::from_path(&workspace_path, &system, Some(configuration.clone()))?;
+    let workspace = WorkspaceMetadata::discover(&workspace_path, &system, Some(&configuration))?;
 
     let db = RootDatabase::new(workspace, system)?;
 
@@ -1311,3 +1312,138 @@ mod unix {
         Ok(())
     }
 }
+
+#[test]
+fn nested_packages_delete_root() -> anyhow::Result<()> {
+    let mut case = setup(|root: &SystemPath, workspace_root: &SystemPath| {
+        std::fs::write(
+            workspace_root.join("pyproject.toml").as_std_path(),
+            r#"
+            [project]
+            name = "inner"
+            "#,
+        )?;
+
+        std::fs::write(
+            root.join("pyproject.toml").as_std_path(),
+            r#"
+            [project]
+            name = "outer"
+            "#,
+        )?;
+
+        Ok(())
+    })?;
+
+    assert_eq!(
+        case.db().workspace().root(case.db()),
+        &*case.workspace_path("")
+    );
+
+    std::fs::remove_file(case.workspace_path("pyproject.toml").as_std_path())?;
+
+    let changes = case.stop_watch();
+
+    case.apply_changes(changes);
+
+    // It should now pick up the outer workspace.
+    assert_eq!(case.db().workspace().root(case.db()), case.root_path());
+
+    Ok(())
+}
+
+#[test]
+fn added_package() -> anyhow::Result<()> {
+    let _ = setup_logging();
+    let mut case = setup([
+        (
+            "pyproject.toml",
+            r#"
+            [project]
+            name = "inner"
+
+            [tool.knot.workspace]
+            members = ["packages/*"]
+            "#,
+        ),
+        (
+            "packages/a/pyproject.toml",
+            r#"
+            [project]
+            name = "a"
+            "#,
+        ),
+    ])?;
+
+    assert_eq!(case.db().workspace().packages(case.db()).len(), 2);
+
+    std::fs::create_dir(case.workspace_path("packages/b").as_std_path())
+        .context("failed to create folder for package 'b'")?;
+
+    // It seems that the file watcher won't pick up on file changes shortly after the folder
+    // was created... I suspect this is because most file watchers don't support recursive
+    // file watching. Instead, file-watching libraries manually implement recursive file watching
+    // by setting a watcher for each directory. But doing this obviously "lags" behind.
+    case.take_watch_changes();
+
+    std::fs::write(
+        case.workspace_path("packages/b/pyproject.toml")
+            .as_std_path(),
+        r#"
+            [project]
+            name = "b"
+            "#,
+    )
+    .context("failed to write pyproject.toml for package b")?;
+
+    let changes = case.stop_watch();
+
+    case.apply_changes(changes);
+
+    assert_eq!(case.db().workspace().packages(case.db()).len(), 3);
+
+    Ok(())
+}
+
+#[test]
+fn removed_package() -> anyhow::Result<()> {
+    let mut case = setup([
+        (
+            "pyproject.toml",
+            r#"
+            [project]
+            name = "inner"
+
+            [tool.knot.workspace]
+            members = ["packages/*"]
+            "#,
+        ),
+        (
+            "packages/a/pyproject.toml",
+            r#"
+            [project]
+            name = "a"
+            "#,
+        ),
+        (
+            "packages/b/pyproject.toml",
+            r#"
+            [project]
+            name = "b"
+            "#,
+        ),
+    ])?;
+
+    assert_eq!(case.db().workspace().packages(case.db()).len(), 3);
+
+    std::fs::remove_dir_all(case.workspace_path("packages/b").as_std_path())
+        .context("failed to remove package 'b'")?;
+
+    let changes = case.stop_watch();
+
+    case.apply_changes(changes);
+
+    assert_eq!(case.db().workspace().packages(case.db()).len(), 2);
+
+    Ok(())
+}

crates/red_knot_python_semantic/Cargo.toml

@@ -13,7 +13,8 @@ license = { workspace = true }
 [dependencies]
 ruff_db = { workspace = true }
 ruff_index = { workspace = true }
-ruff_python_ast = { workspace = true, features = ["salsa"] }
+ruff_python_ast = { workspace = true }
+ruff_python_parser = { workspace = true }
 ruff_python_stdlib = { workspace = true }
 ruff_source_file = { workspace = true }
 ruff_text_size = { workspace = true }
@@ -32,6 +33,7 @@ thiserror = { workspace = true }
 tracing = { workspace = true }
 rustc-hash = { workspace = true }
 hashbrown = { workspace = true }
+serde = { workspace = true, optional = true }
 smallvec = { workspace = true }
 static_assertions = { workspace = true }
 test-case = { workspace = true }

crates/red_knot_python_semantic/resources/mdtest/annotations/optional.md

@@ -0,0 +1,47 @@
+# Optional
+
+## Annotation
+
+`typing.Optional` is equivalent to using the type with a None in a Union.
+
+```py
+from typing import Optional
+
+a: Optional[int]
+a1: Optional[bool]
+a2: Optional[Optional[bool]]
+a3: Optional[None]
+
+def f():
+    # revealed: int | None
+    reveal_type(a)
+    # revealed: bool | None
+    reveal_type(a1)
+    # revealed: bool | None
+    reveal_type(a2)
+    # revealed: None
+    reveal_type(a3)
+```
+
+## Assignment
+
+```py
+from typing import Optional
+
+a: Optional[int] = 1
+a = None
+# error: [invalid-assignment] "Object of type `Literal[""]` is not assignable to `int | None`"
+a = ""
+```
+
+## Typing Extensions
+
+```py
+from typing_extensions import Optional
+
+a: Optional[int]
+
+def f():
+    # revealed: int | None
+    reveal_type(a)
+```

crates/red_knot_python_semantic/resources/mdtest/annotations/string.md

@@ -0,0 +1,191 @@
+# String annotations
+
+## Simple
+
+```py
+def f() -> "int":
+    return 1
+
+reveal_type(f())  # revealed: int
+```
+
+## Nested
+
+```py
+def f() -> "'int'":
+    return 1
+
+reveal_type(f())  # revealed: int
+```
+
+## Type expression
+
+```py
+def f1() -> "int | str":
+    return 1
+
+def f2() -> "tuple[int, str]":
+    return 1
+
+reveal_type(f1())  # revealed: int | str
+reveal_type(f2())  # revealed: tuple[int, str]
+```
+
+## Partial
+
+```py
+def f() -> tuple[int, "str"]:
+    return 1
+
+reveal_type(f())  # revealed: tuple[int, str]
+```
+
+## Deferred
+
+```py
+def f() -> "Foo":
+    return Foo()
+
+class Foo:
+    pass
+
+reveal_type(f())  # revealed: Foo
+```
+
+## Deferred (undefined)
+
+```py
+# error: [unresolved-reference]
+def f() -> "Foo":
+    pass
+
+reveal_type(f())  # revealed: Unknown
+```
+
+## Partial deferred
+
+```py
+def f() -> int | "Foo":
+    return 1
+
+class Foo:
+    pass
+
+reveal_type(f())  # revealed: int | Foo
+```
+
+## `typing.Literal`
+
+```py
+from typing import Literal
+
+def f1() -> Literal["Foo", "Bar"]:
+    return "Foo"
+
+def f2() -> 'Literal["Foo", "Bar"]':
+    return "Foo"
+
+class Foo:
+    pass
+
+reveal_type(f1())  # revealed: Literal["Foo", "Bar"]
+reveal_type(f2())  # revealed: Literal["Foo", "Bar"]
+```
+
+## Various string kinds
+
+```py
+# error: [annotation-raw-string] "Type expressions cannot use raw string literal"
+def f1() -> r"int":
+    return 1
+
+# error: [annotation-f-string] "Type expressions cannot use f-strings"
+def f2() -> f"int":
+    return 1
+
+# error: [annotation-byte-string] "Type expressions cannot use bytes literal"
+def f3() -> b"int":
+    return 1
+
+def f4() -> "int":
+    return 1
+
+# error: [annotation-implicit-concat] "Type expressions cannot span multiple string literals"
+def f5() -> "in" "t":
+    return 1
+
+# error: [annotation-escape-character] "Type expressions cannot contain escape characters"
+def f6() -> "\N{LATIN SMALL LETTER I}nt":
+    return 1
+
+# error: [annotation-escape-character] "Type expressions cannot contain escape characters"
+def f7() -> "\x69nt":
+    return 1
+
+def f8() -> """int""":
+    return 1
+
+# error: [annotation-byte-string] "Type expressions cannot use bytes literal"
+def f9() -> "b'int'":
+    return 1
+
+reveal_type(f1())  # revealed: Unknown
+reveal_type(f2())  # revealed: Unknown
+reveal_type(f3())  # revealed: Unknown
+reveal_type(f4())  # revealed: int
+reveal_type(f5())  # revealed: Unknown
+reveal_type(f6())  # revealed: Unknown
+reveal_type(f7())  # revealed: Unknown
+reveal_type(f8())  # revealed: int
+reveal_type(f9())  # revealed: Unknown
+```
+
+## Various string kinds in `typing.Literal`
+
+```py
+from typing import Literal
+
+def f() -> Literal["a", r"b", b"c", "d" "e", "\N{LATIN SMALL LETTER F}", "\x67", """h"""]:
+    return "normal"
+
+reveal_type(f())  # revealed: Literal["a", "b", "de", "f", "g", "h"] | Literal[b"c"]
+```
+
+## Class variables
+
+```py
+MyType = int
+
+class Aliases:
+    MyType = str
+
+    forward: "MyType"
+    not_forward: MyType
+
+reveal_type(Aliases.forward)  # revealed: str
+reveal_type(Aliases.not_forward)  # revealed: str
+```
+
+## Annotated assignment
+
+```py
+a: "int" = 1
+b: "'int'" = 1
+c: "Foo"
+# error: [invalid-assignment] "Object of type `Literal[1]` is not assignable to `Foo`"
+d: "Foo" = 1
+
+class Foo:
+    pass
+
+c = Foo()
+
+reveal_type(a)  # revealed: Literal[1]
+reveal_type(b)  # revealed: Literal[1]
+reveal_type(c)  # revealed: Foo
+reveal_type(d)  # revealed: Foo
+```
+
+## Parameter
+
+TODO: Add tests once parameter inference is supported

crates/red_knot_python_semantic/resources/mdtest/annotations/string_annotations.md

@@ -1,9 +0,0 @@
-# String annotations
-
-```py
-def f() -> "int":
-    return 1
-
-# TODO: We do not support string annotations, but we should not panic if we encounter them
-reveal_type(f())  # revealed: @Todo
-```

crates/red_knot_python_semantic/resources/mdtest/assignment/annotations.md

@@ -110,3 +110,29 @@ c: builtins.tuple[builtins.tuple[builtins.int, builtins.int], builtins.int] = ((
 # error: [invalid-assignment] "Object of type `Literal["foo"]` is not assignable to `tuple[tuple[int, int], int]`"
 c: builtins.tuple[builtins.tuple[builtins.int, builtins.int], builtins.int] = "foo"
 ```
+
+## Future annotations are deferred
+
+```py
+from __future__ import annotations
+
+x: Foo
+
+class Foo:
+    pass
+
+x = Foo()
+reveal_type(x)  # revealed: Foo
+```
+
+## Annotations in stub files are deferred
+
+```pyi path=main.pyi
+x: Foo
+
+class Foo:
+    pass
+
+x = Foo()
+reveal_type(x)  # revealed: Foo
+```

crates/red_knot_python_semantic/resources/mdtest/assignment/unbound.md

@@ -35,6 +35,7 @@ class C:
     if flag:
         x = 2
 
+# error: [possibly-unbound-attribute] "Attribute `x` on type `Literal[C]` is possibly unbound"
 reveal_type(C.x)  # revealed: Literal[2]
 reveal_type(C.y)  # revealed: Literal[1]
 ```

crates/red_knot_python_semantic/resources/mdtest/attributes.md

@@ -9,14 +9,21 @@ def bool_instance() -> bool:
 flag = bool_instance()
 
 if flag:
-    class C:
+    class C1:
         x = 1
 
 else:
-    class C:
+    class C1:
         x = 2
 
-reveal_type(C.x)  # revealed: Literal[1, 2]
+class C2:
+    if flag:
+        x = 3
+    else:
+        x = 4
+
+reveal_type(C1.x)  # revealed: Literal[1, 2]
+reveal_type(C2.x)  # revealed: Literal[3, 4]
 ```
 
 ## Inherited attributes
@@ -53,3 +60,77 @@ reveal_type(A.__mro__)
 # `E` is earlier in the MRO than `F`, so we should use the type of `E.X`
 reveal_type(A.X)  # revealed: Literal[42]
 ```
+
+## Unions with possibly unbound paths
+
+### Definite boundness within a class
+
+In this example, the `x` attribute is not defined in the `C2` element of the union:
+
+```py
+def bool_instance() -> bool:
+    return True
+
+class C1:
+    x = 1
+
+class C2: ...
+
+class C3:
+    x = 3
+
+flag1 = bool_instance()
+flag2 = bool_instance()
+
+C = C1 if flag1 else C2 if flag2 else C3
+
+# error: [possibly-unbound-attribute] "Attribute `x` on type `Literal[C1, C2, C3]` is possibly unbound"
+reveal_type(C.x)  # revealed: Literal[1, 3]
+```
+
+### Possibly-unbound within a class
+
+We raise the same diagnostic if the attribute is possibly-unbound in at least one element of the
+union:
+
+```py
+def bool_instance() -> bool:
+    return True
+
+class C1:
+    x = 1
+
+class C2:
+    if bool_instance():
+        x = 2
+
+class C3:
+    x = 3
+
+flag1 = bool_instance()
+flag2 = bool_instance()
+
+C = C1 if flag1 else C2 if flag2 else C3
+
+# error: [possibly-unbound-attribute] "Attribute `x` on type `Literal[C1, C2, C3]` is possibly unbound"
+reveal_type(C.x)  # revealed: Literal[1, 2, 3]
+```
+
+## Unions with all paths unbound
+
+If the symbol is unbound in all elements of the union, we detect that:
+
+```py
+def bool_instance() -> bool:
+    return True
+
+class C1: ...
+class C2: ...
+
+flag = bool_instance()
+
+C = C1 if flag else C2
+
+# error: [unresolved-attribute] "Type `Literal[C1, C2]` has no attribute `x`"
+reveal_type(C.x)  # revealed: Unknown
+```

crates/red_knot_python_semantic/resources/mdtest/call/function.md

@@ -19,6 +19,15 @@ async def get_int_async() -> int:
 reveal_type(get_int_async())  # revealed: @Todo
 ```
 
+## Generic
+
+```py
+def get_int[T]() -> int:
+    return 42
+
+reveal_type(get_int())  # revealed: int
+```
+
 ## Decorated
 
 ```py

crates/red_knot_python_semantic/resources/mdtest/exception/basic.md

@@ -41,11 +41,10 @@ except EXCEPTIONS as f:
 ## Dynamic exception types
 
 ```py
-# TODO: we should not emit these `call-possibly-unbound-method` errors for `tuple.__class_getitem__`
 def foo(
     x: type[AttributeError],
-    y: tuple[type[OSError], type[RuntimeError]],  # error: [call-possibly-unbound-method]
-    z: tuple[type[BaseException], ...],  # error: [call-possibly-unbound-method]
+    y: tuple[type[OSError], type[RuntimeError]],
+    z: tuple[type[BaseException], ...],
 ):
     try:
         help()

crates/red_knot_python_semantic/resources/mdtest/expression/attribute.md

@@ -0,0 +1,28 @@
+# Attribute access
+
+## Boundness
+
+```py
+def flag() -> bool: ...
+
+class A:
+    always_bound = 1
+
+    if flag():
+        union = 1
+    else:
+        union = "abc"
+
+    if flag():
+        possibly_unbound = "abc"
+
+reveal_type(A.always_bound)  # revealed: Literal[1]
+
+reveal_type(A.union)  # revealed: Literal[1] | Literal["abc"]
+
+# error: [possibly-unbound-attribute] "Attribute `possibly_unbound` on type `Literal[A]` is possibly unbound"
+reveal_type(A.possibly_unbound)  # revealed: Literal["abc"]
+
+# error: [unresolved-attribute] "Type `Literal[A]` has no attribute `non_existent`"
+reveal_type(A.non_existent)  # revealed: Unknown
+```

crates/red_knot_python_semantic/resources/mdtest/generics.md

@@ -6,13 +6,9 @@ Basic PEP 695 generics
 
 ```py
 class MyBox[T]:
-    # TODO: `T` is defined here
-    # error: [unresolved-reference] "Name `T` used when not defined"
     data: T
     box_model_number = 695
 
-    # TODO: `T` is defined here
-    # error: [unresolved-reference] "Name `T` used when not defined"
     def __init__(self, data: T):
         self.data = data
 
@@ -31,17 +27,12 @@ reveal_type(MyBox.box_model_number)  # revealed: Literal[695]
 
 ```py
 class MyBox[T]:
-    # TODO: `T` is defined here
-    # error: [unresolved-reference] "Name `T` used when not defined"
     data: T
 
-    # TODO: `T` is defined here
-    # error: [unresolved-reference] "Name `T` used when not defined"
     def __init__(self, data: T):
         self.data = data
 
-# TODO not error on the subscripting or the use of type param
-# error: [unresolved-reference] "Name `T` used when not defined"
+# TODO not error on the subscripting
 # error: [non-subscriptable]
 class MySecureBox[T](MyBox[T]): ...
 
@@ -66,3 +57,55 @@ class S[T](Seq[S]): ...  # error: [non-subscriptable]
 
 reveal_type(S)  # revealed: Literal[S]
 ```
+
+## Type params
+
+A PEP695 type variable defines a value of type `typing.TypeVar` with attributes `__name__`,
+`__bounds__`, `__constraints__`, and `__default__` (the latter three all lazily evaluated):
+
+```py
+def f[T, U: A, V: (A, B), W = A, X: A = A1]():
+    reveal_type(T)  # revealed: T
+    reveal_type(T.__name__)  # revealed: Literal["T"]
+    reveal_type(T.__bound__)  # revealed: None
+    reveal_type(T.__constraints__)  # revealed: tuple[()]
+    reveal_type(T.__default__)  # revealed: NoDefault
+
+    reveal_type(U)  # revealed: U
+    reveal_type(U.__name__)  # revealed: Literal["U"]
+    reveal_type(U.__bound__)  # revealed: type[A]
+    reveal_type(U.__constraints__)  # revealed: tuple[()]
+    reveal_type(U.__default__)  # revealed: NoDefault
+
+    reveal_type(V)  # revealed: V
+    reveal_type(V.__name__)  # revealed: Literal["V"]
+    reveal_type(V.__bound__)  # revealed: None
+    reveal_type(V.__constraints__)  # revealed: tuple[type[A], type[B]]
+    reveal_type(V.__default__)  # revealed: NoDefault
+
+    reveal_type(W)  # revealed: W
+    reveal_type(W.__name__)  # revealed: Literal["W"]
+    reveal_type(W.__bound__)  # revealed: None
+    reveal_type(W.__constraints__)  # revealed: tuple[()]
+    reveal_type(W.__default__)  # revealed: type[A]
+
+    reveal_type(X)  # revealed: X
+    reveal_type(X.__name__)  # revealed: Literal["X"]
+    reveal_type(X.__bound__)  # revealed: type[A]
+    reveal_type(X.__constraints__)  # revealed: tuple[()]
+    reveal_type(X.__default__)  # revealed: type[A1]
+
+class A: ...
+class B: ...
+class A1(A): ...
+```
+
+## Minimum two constraints
+
+A typevar with less than two constraints emits a diagnostic and is treated as unconstrained:
+
+```py
+# error: [invalid-typevar-constraints] "TypeVar must have at least two constrained types"
+def f[T: (int,)]():
+    reveal_type(T.__constraints__)  # revealed: tuple[()]
+```

crates/red_knot_python_semantic/resources/mdtest/import/conditional.md

@@ -21,6 +21,7 @@ reveal_type(y)
 ```
 
 ```py
+# error: [possibly-unbound-import] "Member `y` of module `maybe_unbound` is possibly unbound"
 from maybe_unbound import x, y
 
 reveal_type(x)  # revealed: Literal[3]
@@ -50,6 +51,7 @@ reveal_type(y)
 Importing an annotated name prefers the declared type over the inferred type:
 
 ```py
+# error: [possibly-unbound-import] "Member `y` of module `maybe_unbound_annotated` is possibly unbound"
 from maybe_unbound_annotated import x, y
 
 reveal_type(x)  # revealed: Literal[3]

crates/red_knot_python_semantic/resources/mdtest/literal/literal.md

@@ -51,6 +51,8 @@ invalid1: Literal[3 + 4]
 invalid2: Literal[4 + 3j]
 # error: [invalid-literal-parameter]
 invalid3: Literal[(3, 4)]
+
+hello = "hello"
 invalid4: Literal[
     1 + 2,  # error: [invalid-literal-parameter]
     "foo",

crates/red_knot_python_semantic/resources/mdtest/narrow/issubclass.md

@@ -102,6 +102,9 @@ else:
 ### Handling of `None`
 
 ```py
+# TODO: this error should ideally go away once we (1) understand `sys.version_info` branches,
+# and (2) set the target Python version for this test to 3.10.
+# error: [possibly-unbound-import] "Member `NoneType` of module `types` is possibly unbound"
 from types import NoneType
 
 def flag() -> bool: ...

crates/red_knot_python_semantic/resources/mdtest/narrow/type.md

@@ -0,0 +1,152 @@
+# Narrowing for checks involving `type(x)`
+
+## `type(x) is C`
+
+```py
+class A: ...
+class B: ...
+
+def get_a_or_b() -> A | B:
+    return A()
+
+x = get_a_or_b()
+
+if type(x) is A:
+    reveal_type(x)  # revealed: A
+else:
+    # It would be wrong to infer `B` here. The type
+    # of `x` could be a subclass of `A`, so we need
+    # to infer the full union type:
+    reveal_type(x)  # revealed: A | B
+```
+
+## `type(x) is not C`
+
+```py
+class A: ...
+class B: ...
+
+def get_a_or_b() -> A | B:
+    return A()
+
+x = get_a_or_b()
+
+if type(x) is not A:
+    # Same reasoning as above: no narrowing should occur here.
+    reveal_type(x)  # revealed: A | B
+else:
+    reveal_type(x)  # revealed: A
+```
+
+## `type(x) == C`, `type(x) != C`
+
+No narrowing can occur for equality comparisons, since there might be a custom `__eq__`
+implementation on the metaclass.
+
+TODO: Narrowing might be possible in some cases where the classes themselves are `@final` or their
+metaclass is `@final`.
+
+```py
+class IsEqualToEverything(type):
+    def __eq__(cls, other):
+        return True
+
+class A(metaclass=IsEqualToEverything): ...
+class B(metaclass=IsEqualToEverything): ...
+
+def get_a_or_b() -> A | B:
+    return B()
+
+x = get_a_or_b()
+
+if type(x) == A:
+    reveal_type(x)  # revealed: A | B
+
+if type(x) != A:
+    reveal_type(x)  # revealed: A | B
+```
+
+## No narrowing for custom `type` callable
+
+```py
+class A: ...
+class B: ...
+
+def type(x):
+    return int
+
+def get_a_or_b() -> A | B:
+    return A()
+
+x = get_a_or_b()
+
+if type(x) is A:
+    reveal_type(x)  # revealed: A | B
+else:
+    reveal_type(x)  # revealed: A | B
+```
+
+## No narrowing for multiple arguments
+
+No narrowing should occur if `type` is used to dynamically create a class:
+
+```py
+def get_str_or_int() -> str | int:
+    return "test"
+
+x = get_str_or_int()
+
+if type(x, (), {}) is str:
+    reveal_type(x)  # revealed: str | int
+else:
+    reveal_type(x)  # revealed: str | int
+```
+
+## No narrowing for keyword arguments
+
+`type` can't be used with a keyword argument:
+
+```py
+def get_str_or_int() -> str | int:
+    return "test"
+
+x = get_str_or_int()
+
+# TODO: we could issue a diagnostic here
+if type(object=x) is str:
+    reveal_type(x)  # revealed: str | int
+```
+
+## Narrowing if `type` is aliased
+
+```py
+class A: ...
+class B: ...
+
+alias_for_type = type
+
+def get_a_or_b() -> A | B:
+    return A()
+
+x = get_a_or_b()
+
+if alias_for_type(x) is A:
+    reveal_type(x)  # revealed: A
+```
+
+## Limitations
+
+```py
+class Base: ...
+class Derived(Base): ...
+
+def get_base() -> Base:
+    return Base()
+
+x = get_base()
+
+if type(x) is Base:
+    # Ideally, this could be narrower, but there is now way to
+    # express a constraint like `Base & ~ProperSubtypeOf[Base]`.
+    reveal_type(x)  # revealed: Base
+```

crates/red_knot_python_semantic/resources/mdtest/regression/14334_diagnostics_in_wrong_file.md

@@ -0,0 +1,13 @@
+# Regression test for #14334
+
+Regression test for [this issue](https://github.com/astral-sh/ruff/issues/14334).
+
+```py path=base.py
+# error: [invalid-base]
+class Base(2): ...
+```
+
+```py path=a.py
+# No error here
+from base import Base
+```

crates/red_knot_python_semantic/resources/mdtest/scopes/moduletype_attrs.md

@@ -74,6 +74,7 @@ we're dealing with:
 ```py path=__getattr__.py
 import typing
 
+# error: [unresolved-attribute]
 reveal_type(typing.__getattr__)  # revealed: Unknown
 ```
 

crates/red_knot_python_semantic/resources/mdtest/sys_version_info.md

@@ -0,0 +1,139 @@
+# `sys.version_info`
+
+## The type of `sys.version_info`
+
+The type of `sys.version_info` is `sys._version_info`, at least according to typeshed's stubs (which
+we treat as the single source of truth for the standard library). This is quite a complicated type
+in typeshed, so there are many things we don't fully understand about the type yet; this is the
+source of several TODOs in this test file. Many of these TODOs should be naturally fixed as we
+implement more type-system features in the future.
+
+```py
+import sys
+
+reveal_type(sys.version_info)  # revealed: _version_info
+```
+
+## Literal types from comparisons
+
+Comparing `sys.version_info` with a 2-element tuple of literal integers always produces a `Literal`
+type:
+
+```py
+import sys
+
+reveal_type(sys.version_info >= (3, 9))  # revealed: Literal[True]
+reveal_type((3, 9) <= sys.version_info)  # revealed: Literal[True]
+
+reveal_type(sys.version_info > (3, 9))  # revealed: Literal[True]
+reveal_type((3, 9) < sys.version_info)  # revealed: Literal[True]
+
+reveal_type(sys.version_info < (3, 9))  # revealed: Literal[False]
+reveal_type((3, 9) > sys.version_info)  # revealed: Literal[False]
+
+reveal_type(sys.version_info <= (3, 9))  # revealed: Literal[False]
+reveal_type((3, 9) >= sys.version_info)  # revealed: Literal[False]
+
+reveal_type(sys.version_info == (3, 9))  # revealed: Literal[False]
+reveal_type((3, 9) == sys.version_info)  # revealed: Literal[False]
+
+reveal_type(sys.version_info != (3, 9))  # revealed: Literal[True]
+reveal_type((3, 9) != sys.version_info)  # revealed: Literal[True]
+```
+
+## Non-literal types from comparisons
+
+Comparing `sys.version_info` with tuples of other lengths will sometimes produce `Literal` types,
+sometimes not:
+
+```py
+import sys
+
+reveal_type(sys.version_info >= (3, 9, 1))  # revealed: bool
+reveal_type(sys.version_info >= (3, 9, 1, "final", 0))  # revealed: bool
+
+# TODO: While this won't fail at runtime, the user has probably made a mistake
+# if they're comparing a tuple of length >5 with `sys.version_info`
+# (`sys.version_info` is a tuple of length 5). It might be worth
+# emitting a lint diagnostic of some kind warning them about the probable error?
+reveal_type(sys.version_info >= (3, 9, 1, "final", 0, 5))  # revealed: bool
+
+# TODO: this should be `Literal[False]`; see #14279
+reveal_type(sys.version_info == (3, 9, 1, "finallllll", 0))  # revealed: bool
+```
+
+## Imports and aliases
+
+Comparisons with `sys.version_info` still produce literal types, even if the symbol is aliased to
+another name:
+
+```py
+from sys import version_info
+from sys import version_info as foo
+
+reveal_type(version_info >= (3, 9))  # revealed: Literal[True]
+reveal_type(foo >= (3, 9))  # revealed: Literal[True]
+
+bar = version_info
+reveal_type(bar >= (3, 9))  # revealed: Literal[True]
+```
+
+## Non-stdlib modules named `sys`
+
+Only comparisons with the symbol `version_info` from the `sys` module produce literal types:
+
+```py path=package/__init__.py
+```
+
+```py path=package/sys.py
+version_info: tuple[int, int] = (4, 2)
+```
+
+```py path=package/script.py
+from .sys import version_info
+
+reveal_type(version_info >= (3, 9))  # revealed: bool
+```
+
+## Accessing fields by name
+
+The fields of `sys.version_info` can be accessed by name:
+
+```py path=a.py
+import sys
+
+reveal_type(sys.version_info.major >= 3)  # revealed: Literal[True]
+reveal_type(sys.version_info.minor >= 9)  # revealed: Literal[True]
+reveal_type(sys.version_info.minor >= 10)  # revealed: Literal[False]
+```
+
+But the `micro`, `releaselevel` and `serial` fields are inferred as `@Todo` until we support
+properties on instance types:
+
+```py path=b.py
+import sys
+
+reveal_type(sys.version_info.micro)  # revealed: @Todo
+reveal_type(sys.version_info.releaselevel)  # revealed: @Todo
+reveal_type(sys.version_info.serial)  # revealed: @Todo
+```
+
+## Accessing fields by index/slice
+
+The fields of `sys.version_info` can be accessed by index or by slice:
+
+```py
+import sys
+
+reveal_type(sys.version_info[0] < 3)  # revealed: Literal[False]
+reveal_type(sys.version_info[1] > 9)  # revealed: Literal[False]
+
+# revealed: tuple[Literal[3], Literal[9], int, Literal["alpha", "beta", "candidate", "final"], int]
+reveal_type(sys.version_info[:5])
+
+reveal_type(sys.version_info[:2] >= (3, 9))  # revealed: Literal[True]
+reveal_type(sys.version_info[0:2] >= (3, 10))  # revealed: Literal[False]
+reveal_type(sys.version_info[:3] >= (3, 10, 1))  # revealed: Literal[False]
+reveal_type(sys.version_info[3] == "final")  # revealed: bool
+reveal_type(sys.version_info[3] == "finalllllll")  # revealed: Literal[False]
+```

crates/red_knot_python_semantic/resources/mdtest/unary/invert_add_usub.md

@@ -1,4 +1,6 @@
-# Unary Operations
+# Invert, UAdd, USub
+
+## Instance
 
 ```py
 from typing import Literal

crates/red_knot_python_semantic/resources/mdtest/unary/not.md

@@ -10,8 +10,6 @@ reveal_type(not not None)  # revealed: Literal[False]
 ## Function
 
 ```py
-from typing import reveal_type
-
 def f():
     return 1
 
@@ -115,3 +113,101 @@ reveal_type(not ())  # revealed: Literal[True]
 reveal_type(not ("hello",))  # revealed: Literal[False]
 reveal_type(not (1, "hello"))  # revealed: Literal[False]
 ```
+
+## Instance
+
+Not operator is inferred based on
+<https://docs.python.org/3/library/stdtypes.html#truth-value-testing>. An instance is True or False
+if the `__bool__` method says so.
+
+At runtime, the `__len__` method is a fallback for `__bool__`, but we can't make use of that. If we
+have a class that defines `__len__` but not `__bool__`, it is possible that any subclass could add a
+`__bool__` method that would invalidate whatever conclusion we drew from `__len__`. So instances of
+classes without a `__bool__` method, with or without `__len__`, must be inferred as unknown
+truthiness.
+
+```py
+class AlwaysTrue:
+    def __bool__(self) -> Literal[True]:
+        return True
+
+# revealed: Literal[False]
+reveal_type(not AlwaysTrue())
+
+class AlwaysFalse:
+    def __bool__(self) -> Literal[False]:
+        return False
+
+# revealed: Literal[True]
+reveal_type(not AlwaysFalse())
+
+# We don't get into a cycle if someone sets their `__bool__` method to the `bool` builtin:
+class BoolIsBool:
+    __bool__ = bool
+
+# revealed: bool
+reveal_type(not BoolIsBool())
+
+# At runtime, no `__bool__` and no `__len__` means truthy, but we can't rely on that, because
+# a subclass could add a `__bool__` method.
+class NoBoolMethod: ...
+
+# revealed: bool
+reveal_type(not NoBoolMethod())
+
+# And we can't rely on `__len__` for the same reason: a subclass could add `__bool__`.
+class LenZero:
+    def __len__(self) -> Literal[0]:
+        return 0
+
+# revealed: bool
+reveal_type(not LenZero())
+
+class LenNonZero:
+    def __len__(self) -> Literal[1]:
+        return 1
+
+# revealed: bool
+reveal_type(not LenNonZero())
+
+class WithBothLenAndBool1:
+    def __bool__(self) -> Literal[False]:
+        return False
+
+    def __len__(self) -> Literal[2]:
+        return 2
+
+# revealed: Literal[True]
+reveal_type(not WithBothLenAndBool1())
+
+class WithBothLenAndBool2:
+    def __bool__(self) -> Literal[True]:
+        return True
+
+    def __len__(self) -> Literal[0]:
+        return 0
+
+# revealed: Literal[False]
+reveal_type(not WithBothLenAndBool2())
+
+# TODO: raise diagnostic when __bool__ method is not valid: [unsupported-operator] "Method __bool__ for type `MethodBoolInvalid` should return `bool`, returned type `int`"
+# https://docs.python.org/3/reference/datamodel.html#object.__bool__
+class MethodBoolInvalid:
+    def __bool__(self) -> int:
+        return 0
+
+# revealed: bool
+reveal_type(not MethodBoolInvalid())
+
+# Don't trust a possibly-unbound `__bool__` method:
+def get_flag() -> bool:
+    return True
+
+class PossiblyUnboundBool:
+    if get_flag():
+        def __bool__(self) -> Literal[False]:
+            return False
+
+# revealed: bool
+reveal_type(not PossiblyUnboundBool())
+```

crates/red_knot_python_semantic/src/module_resolver/typeshed.rs

@@ -459,11 +459,11 @@ foo: 3.8-   # trailing comment
 ";
         let parsed_versions = TypeshedVersions::from_str(VERSIONS).unwrap();
         assert_eq!(parsed_versions.len(), 3);
-        assert_snapshot!(parsed_versions.to_string(), @r###"
+        assert_snapshot!(parsed_versions.to_string(), @r"
         bar: 2.7-3.10
         bar.baz: 3.1-3.9
         foo: 3.8-
-        "###
+        "
         );
     }
 

crates/red_knot_python_semantic/src/node_key.rs

@@ -1,14 +1,14 @@
-use ruff_python_ast::{AnyNodeRef, NodeKind};
-use ruff_text_size::{Ranged, TextRange};
+use ruff_python_ast::AnyNodeRef;
 
 /// Compact key for a node for use in a hash map.
 ///
-/// Compares two nodes by their kind and text range.
+/// Stores the memory address of the node, because using the range and the kind
+/// of the node is not enough to uniquely identify them in ASTs resulting from
+/// invalid syntax. For example, parsing the input `for` results in a `StmtFor`
+/// AST node where both the `target` and the `iter` field are `ExprName` nodes
+/// with the same (empty) range `3..3`.
 #[derive(Copy, Clone, Debug, Eq, PartialEq, Hash)]
-pub(super) struct NodeKey {
-    kind: NodeKind,
-    range: TextRange,
-}
+pub(super) struct NodeKey(usize);
 
 impl NodeKey {
     pub(super) fn from_node<'a, N>(node: N) -> Self
@@ -16,9 +16,6 @@ impl NodeKey {
         N: Into<AnyNodeRef<'a>>,
     {
         let node = node.into();
-        NodeKey {
-            kind: node.kind(),
-            range: node.range(),
-        }
+        NodeKey(node.as_ptr().as_ptr() as usize)
     }
 }

crates/red_knot_python_semantic/src/program.rs

@@ -54,6 +54,7 @@ impl Program {
 }
 
 #[derive(Clone, Debug, Eq, PartialEq)]
+#[cfg_attr(feature = "serde", derive(serde::Serialize))]
 pub struct ProgramSettings {
     pub target_version: PythonVersion,
     pub search_paths: SearchPathSettings,
@@ -61,6 +62,7 @@ pub struct ProgramSettings {
 
 /// Configures the search paths for module resolution.
 #[derive(Eq, PartialEq, Debug, Clone)]
+#[cfg_attr(feature = "serde", derive(serde::Serialize))]
 pub struct SearchPathSettings {
     /// List of user-provided paths that should take first priority in the module resolution.
     /// Examples in other type checkers are mypy's MYPYPATH environment variable,
@@ -91,6 +93,7 @@ impl SearchPathSettings {
 }
 
 #[derive(Debug, Clone, Eq, PartialEq)]
+#[cfg_attr(feature = "serde", derive(serde::Serialize))]
 pub enum SitePackages {
     Derived {
         venv_path: SystemPathBuf,

crates/red_knot_python_semantic/src/python_version.rs

@@ -5,6 +5,7 @@ use std::fmt;
 /// Unlike the `TargetVersion` enums in the CLI crates,
 /// this does not necessarily represent a Python version that we actually support.
 #[derive(Debug, Clone, Copy, Eq, PartialEq, Ord, PartialOrd, Hash)]
+#[cfg_attr(feature = "serde", derive(serde::Serialize))]
 pub struct PythonVersion {
     pub major: u8,
     pub minor: u8,
@@ -38,7 +39,7 @@ impl PythonVersion {
 
 impl Default for PythonVersion {
     fn default() -> Self {
-        Self::PY38
+        Self::PY39
     }
 }
 

crates/red_knot_python_semantic/src/semantic_index/ast_ids.rs

@@ -49,64 +49,50 @@ fn ast_ids<'db>(db: &'db dyn Db, scope: ScopeId) -> &'db AstIds {
     semantic_index(db, scope.file(db)).ast_ids(scope.file_scope_id(db))
 }
 
-pub trait HasScopedUseId {
-    /// The type of the ID uniquely identifying the use.
-    type Id: Copy;
-
-    /// Returns the ID that uniquely identifies the use in `scope`.
-    fn scoped_use_id(&self, db: &dyn Db, scope: ScopeId) -> Self::Id;
-}
-
 /// Uniquely identifies a use of a name in a [`crate::semantic_index::symbol::FileScopeId`].
 #[newtype_index]
 pub struct ScopedUseId;
 
-impl HasScopedUseId for ast::ExprName {
-    type Id = ScopedUseId;
+pub trait HasScopedUseId {
+    /// Returns the ID that uniquely identifies the use in `scope`.
+    fn scoped_use_id(&self, db: &dyn Db, scope: ScopeId) -> ScopedUseId;
+}
 
-    fn scoped_use_id(&self, db: &dyn Db, scope: ScopeId) -> Self::Id {
+impl HasScopedUseId for ast::ExprName {
+    fn scoped_use_id(&self, db: &dyn Db, scope: ScopeId) -> ScopedUseId {
         let expression_ref = ExpressionRef::from(self);
         expression_ref.scoped_use_id(db, scope)
     }
 }
 
 impl HasScopedUseId for ast::ExpressionRef<'_> {
-    type Id = ScopedUseId;
-
-    fn scoped_use_id(&self, db: &dyn Db, scope: ScopeId) -> Self::Id {
+    fn scoped_use_id(&self, db: &dyn Db, scope: ScopeId) -> ScopedUseId {
         let ast_ids = ast_ids(db, scope);
         ast_ids.use_id(*self)
     }
 }
 
-pub trait HasScopedAstId {
-    /// The type of the ID uniquely identifying the node.
-    type Id: Copy;
-
-    /// Returns the ID that uniquely identifies the node in `scope`.
-    fn scoped_ast_id(&self, db: &dyn Db, scope: ScopeId) -> Self::Id;
-}
-
-impl<T: HasScopedAstId> HasScopedAstId for Box<T> {
-    type Id = <T as HasScopedAstId>::Id;
-
-    fn scoped_ast_id(&self, db: &dyn Db, scope: ScopeId) -> Self::Id {
-        self.as_ref().scoped_ast_id(db, scope)
-    }
-}
-
 /// Uniquely identifies an [`ast::Expr`] in a [`crate::semantic_index::symbol::FileScopeId`].
 #[newtype_index]
 pub struct ScopedExpressionId;
 
+pub trait HasScopedExpressionId {
+    /// Returns the ID that uniquely identifies the node in `scope`.
+    fn scoped_expression_id(&self, db: &dyn Db, scope: ScopeId) -> ScopedExpressionId;
+}
+
+impl<T: HasScopedExpressionId> HasScopedExpressionId for Box<T> {
+    fn scoped_expression_id(&self, db: &dyn Db, scope: ScopeId) -> ScopedExpressionId {
+        self.as_ref().scoped_expression_id(db, scope)
+    }
+}
+
 macro_rules! impl_has_scoped_expression_id {
     ($ty: ty) => {
-        impl HasScopedAstId for $ty {
-            type Id = ScopedExpressionId;
-
-            fn scoped_ast_id(&self, db: &dyn Db, scope: ScopeId) -> Self::Id {
+        impl HasScopedExpressionId for $ty {
+            fn scoped_expression_id(&self, db: &dyn Db, scope: ScopeId) -> ScopedExpressionId {
                 let expression_ref = ExpressionRef::from(self);
-                expression_ref.scoped_ast_id(db, scope)
+                expression_ref.scoped_expression_id(db, scope)
             }
         }
     };
@@ -146,29 +132,20 @@ impl_has_scoped_expression_id!(ast::ExprSlice);
 impl_has_scoped_expression_id!(ast::ExprIpyEscapeCommand);
 impl_has_scoped_expression_id!(ast::Expr);
 
-impl HasScopedAstId for ast::ExpressionRef<'_> {
-    type Id = ScopedExpressionId;
-
-    fn scoped_ast_id(&self, db: &dyn Db, scope: ScopeId) -> Self::Id {
+impl HasScopedExpressionId for ast::ExpressionRef<'_> {
+    fn scoped_expression_id(&self, db: &dyn Db, scope: ScopeId) -> ScopedExpressionId {
         let ast_ids = ast_ids(db, scope);
         ast_ids.expression_id(*self)
     }
 }
 
-#[derive(Debug)]
+#[derive(Debug, Default)]
 pub(super) struct AstIdsBuilder {
     expressions_map: FxHashMap<ExpressionNodeKey, ScopedExpressionId>,
     uses_map: FxHashMap<ExpressionNodeKey, ScopedUseId>,
 }
 
 impl AstIdsBuilder {
-    pub(super) fn new() -> Self {
-        Self {
-            expressions_map: FxHashMap::default(),
-            uses_map: FxHashMap::default(),
-        }
-    }
-
     /// Adds `expr` to the expression ids map and returns its id.
     pub(super) fn record_expression(&mut self, expr: &ast::Expr) -> ScopedExpressionId {
         let expression_id = self.expressions_map.len().into();

crates/red_knot_python_semantic/src/semantic_index/builder.rs

@@ -124,9 +124,9 @@ impl<'db> SemanticIndexBuilder<'db> {
         self.try_node_context_stack_manager.enter_nested_scope();
 
         let file_scope_id = self.scopes.push(scope);
-        self.symbol_tables.push(SymbolTableBuilder::new());
-        self.use_def_maps.push(UseDefMapBuilder::new());
-        let ast_id_scope = self.ast_ids.push(AstIdsBuilder::new());
+        self.symbol_tables.push(SymbolTableBuilder::default());
+        self.use_def_maps.push(UseDefMapBuilder::default());
+        let ast_id_scope = self.ast_ids.push(AstIdsBuilder::default());
 
         let scope_id = ScopeId::new(self.db, self.file, file_scope_id, countme::Count::default());
 
@@ -373,6 +373,11 @@ impl<'db> SemanticIndexBuilder<'db> {
                 if let Some(default) = default {
                     self.visit_expr(default);
                 }
+                match type_param {
+                    ast::TypeParam::TypeVar(node) => self.add_definition(symbol, node),
+                    ast::TypeParam::ParamSpec(node) => self.add_definition(symbol, node),
+                    ast::TypeParam::TypeVarTuple(node) => self.add_definition(symbol, node),
+                };
             }
         }
 
@@ -671,9 +676,18 @@ where
                 if let Some(value) = &node.value {
                     self.visit_expr(value);
                 }
-                self.push_assignment(node.into());
-                self.visit_expr(&node.target);
-                self.pop_assignment();
+
+                // See https://docs.python.org/3/library/ast.html#ast.AnnAssign
+                if matches!(
+                    *node.target,
+                    ast::Expr::Attribute(_) | ast::Expr::Subscript(_) | ast::Expr::Name(_)
+                ) {
+                    self.push_assignment(node.into());
+                    self.visit_expr(&node.target);
+                    self.pop_assignment();
+                } else {
+                    self.visit_expr(&node.target);
+                }
             }
             ast::Stmt::AugAssign(
                 aug_assign @ ast::StmtAugAssign {
@@ -685,9 +699,18 @@ where
             ) => {
                 debug_assert_eq!(&self.current_assignments, &[]);
                 self.visit_expr(value);
-                self.push_assignment(aug_assign.into());
-                self.visit_expr(target);
-                self.pop_assignment();
+
+                // See https://docs.python.org/3/library/ast.html#ast.AugAssign
+                if matches!(
+                    **target,
+                    ast::Expr::Attribute(_) | ast::Expr::Subscript(_) | ast::Expr::Name(_)
+                ) {
+                    self.push_assignment(aug_assign.into());
+                    self.visit_expr(target);
+                    self.pop_assignment();
+                } else {
+                    self.visit_expr(target);
+                }
             }
             ast::Stmt::If(node) => {
                 self.visit_expr(&node.test);
@@ -1067,9 +1090,15 @@ where
             ast::Expr::Named(node) => {
                 // TODO walrus in comprehensions is implicitly nonlocal
                 self.visit_expr(&node.value);
-                self.push_assignment(node.into());
-                self.visit_expr(&node.target);
-                self.pop_assignment();
+
+                // See https://peps.python.org/pep-0572/#differences-between-assignment-expressions-and-assignment-statements
+                if node.target.is_name_expr() {
+                    self.push_assignment(node.into());
+                    self.visit_expr(&node.target);
+                    self.pop_assignment();
+                } else {
+                    self.visit_expr(&node.target);
+                }
             }
             ast::Expr::Lambda(lambda) => {
                 if let Some(parameters) = &lambda.parameters {

crates/red_knot_python_semantic/src/semantic_index/definition.rs

@@ -92,6 +92,9 @@ pub(crate) enum DefinitionNodeRef<'a> {
     WithItem(WithItemDefinitionNodeRef<'a>),
     MatchPattern(MatchPatternDefinitionNodeRef<'a>),
     ExceptHandler(ExceptHandlerDefinitionNodeRef<'a>),
+    TypeVar(&'a ast::TypeParamTypeVar),
+    ParamSpec(&'a ast::TypeParamParamSpec),
+    TypeVarTuple(&'a ast::TypeParamTypeVarTuple),
 }
 
 impl<'a> From<&'a ast::StmtFunctionDef> for DefinitionNodeRef<'a> {
@@ -130,6 +133,24 @@ impl<'a> From<&'a ast::Alias> for DefinitionNodeRef<'a> {
     }
 }
 
+impl<'a> From<&'a ast::TypeParamTypeVar> for DefinitionNodeRef<'a> {
+    fn from(value: &'a ast::TypeParamTypeVar) -> Self {
+        Self::TypeVar(value)
+    }
+}
+
+impl<'a> From<&'a ast::TypeParamParamSpec> for DefinitionNodeRef<'a> {
+    fn from(value: &'a ast::TypeParamParamSpec) -> Self {
+        Self::ParamSpec(value)
+    }
+}
+
+impl<'a> From<&'a ast::TypeParamTypeVarTuple> for DefinitionNodeRef<'a> {
+    fn from(value: &'a ast::TypeParamTypeVarTuple) -> Self {
+        Self::TypeVarTuple(value)
+    }
+}
+
 impl<'a> From<ImportFromDefinitionNodeRef<'a>> for DefinitionNodeRef<'a> {
     fn from(node_ref: ImportFromDefinitionNodeRef<'a>) -> Self {
         Self::ImportFrom(node_ref)
@@ -317,6 +338,15 @@ impl<'db> DefinitionNodeRef<'db> {
                 handler: AstNodeRef::new(parsed, handler),
                 is_star,
             }),
+            DefinitionNodeRef::TypeVar(node) => {
+                DefinitionKind::TypeVar(AstNodeRef::new(parsed, node))
+            }
+            DefinitionNodeRef::ParamSpec(node) => {
+                DefinitionKind::ParamSpec(AstNodeRef::new(parsed, node))
+            }
+            DefinitionNodeRef::TypeVarTuple(node) => {
+                DefinitionKind::TypeVarTuple(AstNodeRef::new(parsed, node))
+            }
         }
     }
 
@@ -356,6 +386,9 @@ impl<'db> DefinitionNodeRef<'db> {
                 identifier.into()
             }
             Self::ExceptHandler(ExceptHandlerDefinitionNodeRef { handler, .. }) => handler.into(),
+            Self::TypeVar(node) => node.into(),
+            Self::ParamSpec(node) => node.into(),
+            Self::TypeVarTuple(node) => node.into(),
         }
     }
 }
@@ -412,6 +445,9 @@ pub enum DefinitionKind<'db> {
     WithItem(WithItemDefinitionKind),
     MatchPattern(MatchPatternDefinitionKind),
     ExceptHandler(ExceptHandlerDefinitionKind),
+    TypeVar(AstNodeRef<ast::TypeParamTypeVar>),
+    ParamSpec(AstNodeRef<ast::TypeParamParamSpec>),
+    TypeVarTuple(AstNodeRef<ast::TypeParamTypeVarTuple>),
 }
 
 impl DefinitionKind<'_> {
@@ -421,7 +457,10 @@ impl DefinitionKind<'_> {
             DefinitionKind::Function(_)
             | DefinitionKind::Class(_)
             | DefinitionKind::Import(_)
-            | DefinitionKind::ImportFrom(_) => DefinitionCategory::DeclarationAndBinding,
+            | DefinitionKind::ImportFrom(_)
+            | DefinitionKind::TypeVar(_)
+            | DefinitionKind::ParamSpec(_)
+            | DefinitionKind::TypeVarTuple(_) => DefinitionCategory::DeclarationAndBinding,
             // a parameter always binds a value, but is only a declaration if annotated
             DefinitionKind::Parameter(parameter) => {
                 if parameter.annotation.is_some() {
@@ -696,3 +735,21 @@ impl From<&ast::ExceptHandlerExceptHandler> for DefinitionNodeKey {
         Self(NodeKey::from_node(handler))
     }
 }
+
+impl From<&ast::TypeParamTypeVar> for DefinitionNodeKey {
+    fn from(value: &ast::TypeParamTypeVar) -> Self {
+        Self(NodeKey::from_node(value))
+    }
+}
+
+impl From<&ast::TypeParamParamSpec> for DefinitionNodeKey {
+    fn from(value: &ast::TypeParamParamSpec) -> Self {
+        Self(NodeKey::from_node(value))
+    }
+}
+
+impl From<&ast::TypeParamTypeVarTuple> for DefinitionNodeKey {
+    fn from(value: &ast::TypeParamTypeVarTuple) -> Self {
+        Self(NodeKey::from_node(value))
+    }
+}

crates/red_knot_python_semantic/src/semantic_index/symbol.rs

@@ -210,7 +210,7 @@ impl ScopeKind {
 }
 
 /// Symbol table for a specific [`Scope`].
-#[derive(Debug)]
+#[derive(Debug, Default)]
 pub struct SymbolTable {
     /// The symbols in this scope.
     symbols: IndexVec<ScopedSymbolId, Symbol>,
@@ -220,13 +220,6 @@ pub struct SymbolTable {
 }
 
 impl SymbolTable {
-    fn new() -> Self {
-        Self {
-            symbols: IndexVec::new(),
-            symbols_by_name: SymbolMap::default(),
-        }
-    }
-
     fn shrink_to_fit(&mut self) {
         self.symbols.shrink_to_fit();
     }
@@ -278,18 +271,12 @@ impl PartialEq for SymbolTable {
 
 impl Eq for SymbolTable {}
 
-#[derive(Debug)]
+#[derive(Debug, Default)]
 pub(super) struct SymbolTableBuilder {
     table: SymbolTable,
 }
 
 impl SymbolTableBuilder {
-    pub(super) fn new() -> Self {
-        Self {
-            table: SymbolTable::new(),
-        }
-    }
-
     pub(super) fn add_symbol(&mut self, name: Name) -> (ScopedSymbolId, bool) {
         let hash = SymbolTable::hash_name(&name);
         let entry = self

crates/red_knot_python_semantic/src/semantic_index/use_def.rs

@@ -277,7 +277,7 @@ impl<'db> UseDefMap<'db> {
 
     pub(crate) fn use_boundness(&self, use_id: ScopedUseId) -> Boundness {
         if self.bindings_by_use[use_id].may_be_unbound() {
-            Boundness::MayBeUnbound
+            Boundness::PossiblyUnbound
         } else {
             Boundness::Bound
         }
@@ -292,7 +292,7 @@ impl<'db> UseDefMap<'db> {
 
     pub(crate) fn public_boundness(&self, symbol: ScopedSymbolId) -> Boundness {
         if self.public_symbols[symbol].may_be_unbound() {
-            Boundness::MayBeUnbound
+            Boundness::PossiblyUnbound
         } else {
             Boundness::Bound
         }
@@ -459,10 +459,6 @@ pub(super) struct UseDefMapBuilder<'db> {
 }
 
 impl<'db> UseDefMapBuilder<'db> {
-    pub(super) fn new() -> Self {
-        Self::default()
-    }
-
     pub(super) fn add_symbol(&mut self, symbol: ScopedSymbolId) {
         let new_symbol = self.symbol_states.push(SymbolState::undefined());
         debug_assert_eq!(symbol, new_symbol);

crates/red_knot_python_semantic/src/semantic_model.rs

@@ -6,7 +6,7 @@ use ruff_source_file::LineIndex;
 
 use crate::module_name::ModuleName;
 use crate::module_resolver::{resolve_module, Module};
-use crate::semantic_index::ast_ids::HasScopedAstId;
+use crate::semantic_index::ast_ids::HasScopedExpressionId;
 use crate::semantic_index::semantic_index;
 use crate::types::{binding_ty, infer_scope_types, Type};
 use crate::Db;
@@ -54,7 +54,7 @@ impl HasTy for ast::ExpressionRef<'_> {
         let file_scope = index.expression_scope_id(*self);
         let scope = file_scope.to_scope_id(model.db, model.file);
 
-        let expression_id = self.scoped_ast_id(model.db, scope);
+        let expression_id = self.scoped_expression_id(model.db, scope);
         infer_scope_types(model.db, scope).expression_ty(expression_id)
     }
 }

crates/red_knot_python_semantic/src/site_packages.rs

@@ -732,7 +732,20 @@ mod tests {
         let system = TestSystem::default();
         assert!(matches!(
             VirtualEnvironment::new("/.venv", &system),
-            Err(SitePackagesDiscoveryError::VenvDirIsNotADirectory(_))
+            Err(SitePackagesDiscoveryError::VenvDirCanonicalizationError(..))
+        ));
+    }
+
+    #[test]
+    fn reject_venv_that_is_not_a_directory() {
+        let system = TestSystem::default();
+        system
+            .memory_file_system()
+            .write_file("/.venv", "")
+            .unwrap();
+        assert!(matches!(
+            VirtualEnvironment::new("/.venv", &system),
+            Err(SitePackagesDiscoveryError::VenvDirIsNotADirectory(..))
         ));
     }
 

crates/red_knot_python_semantic/src/stdlib.rs

@@ -8,32 +8,38 @@ use crate::Db;
 
 /// Enumeration of various core stdlib modules, for which we have dedicated Salsa queries.
 #[derive(Debug, Clone, Copy, PartialEq, Eq)]
-enum CoreStdlibModule {
+pub(crate) enum CoreStdlibModule {
     Builtins,
     Types,
     Typeshed,
     TypingExtensions,
     Typing,
+    Sys,
 }
 
 impl CoreStdlibModule {
-    fn name(self) -> ModuleName {
-        let module_name = match self {
+    pub(crate) const fn as_str(self) -> &'static str {
+        match self {
             Self::Builtins => "builtins",
             Self::Types => "types",
             Self::Typing => "typing",
             Self::Typeshed => "_typeshed",
             Self::TypingExtensions => "typing_extensions",
-        };
-        ModuleName::new_static(module_name)
-            .unwrap_or_else(|| panic!("{module_name} should be a valid module name!"))
+            Self::Sys => "sys",
+        }
+    }
+
+    pub(crate) fn name(self) -> ModuleName {
+        let self_as_str = self.as_str();
+        ModuleName::new_static(self_as_str)
+            .unwrap_or_else(|| panic!("{self_as_str} should be a valid module name!"))
     }
 }
 
 /// Lookup the type of `symbol` in a given core module
 ///
 /// Returns `Symbol::Unbound` if the given core module cannot be resolved for some reason
-fn core_module_symbol<'db>(
+pub(crate) fn core_module_symbol<'db>(
     db: &'db dyn Db,
     core_module: CoreStdlibModule,
     symbol: &str,
@@ -51,29 +57,14 @@ pub(crate) fn builtins_symbol<'db>(db: &'db dyn Db, symbol: &str) -> Symbol<'db>
     core_module_symbol(db, CoreStdlibModule::Builtins, symbol)
 }
 
-/// Lookup the type of `symbol` in the `types` module namespace.
-///
-/// Returns `Symbol::Unbound` if the `types` module isn't available for some reason.
-#[inline]
-pub(crate) fn types_symbol<'db>(db: &'db dyn Db, symbol: &str) -> Symbol<'db> {
-    core_module_symbol(db, CoreStdlibModule::Types, symbol)
-}
-
 /// Lookup the type of `symbol` in the `typing` module namespace.
 ///
 /// Returns `Symbol::Unbound` if the `typing` module isn't available for some reason.
 #[inline]
-#[allow(dead_code)] // currently only used in tests
+#[cfg(test)]
 pub(crate) fn typing_symbol<'db>(db: &'db dyn Db, symbol: &str) -> Symbol<'db> {
     core_module_symbol(db, CoreStdlibModule::Typing, symbol)
 }
-/// Lookup the type of `symbol` in the `_typeshed` module namespace.
-///
-/// Returns `Symbol::Unbound` if the `_typeshed` module isn't available for some reason.
-#[inline]
-pub(crate) fn typeshed_symbol<'db>(db: &'db dyn Db, symbol: &str) -> Symbol<'db> {
-    core_module_symbol(db, CoreStdlibModule::Typeshed, symbol)
-}
 
 /// Lookup the type of `symbol` in the `typing_extensions` module namespace.
 ///

crates/red_knot_python_semantic/src/symbol.rs

@@ -6,7 +6,16 @@ use crate::{
 #[derive(Debug, Clone, Copy, PartialEq, Eq)]
 pub(crate) enum Boundness {
     Bound,
-    MayBeUnbound,
+    PossiblyUnbound,
+}
+
+impl Boundness {
+    pub(crate) fn or(self, other: Boundness) -> Boundness {
+        match (self, other) {
+            (Boundness::Bound, _) | (_, Boundness::Bound) => Boundness::Bound,
+            (Boundness::PossiblyUnbound, Boundness::PossiblyUnbound) => Boundness::PossiblyUnbound,
+        }
+    }
 }
 
 /// The result of a symbol lookup, which can either be a (possibly unbound) type
@@ -17,14 +26,14 @@ pub(crate) enum Boundness {
 /// bound = 1
 ///
 /// if flag:
-///     maybe_unbound = 2
+///     possibly_unbound = 2
 /// ```
 ///
 /// If we look up symbols in this scope, we would get the following results:
 /// ```rs
-/// bound:          Symbol::Type(Type::IntLiteral(1), Boundness::Bound),
-/// maybe_unbound:  Symbol::Type(Type::IntLiteral(2), Boundness::MayBeUnbound),
-/// non_existent:   Symbol::Unbound,
+/// bound:             Symbol::Type(Type::IntLiteral(1), Boundness::Bound),
+/// possibly_unbound:  Symbol::Type(Type::IntLiteral(2), Boundness::PossiblyUnbound),
+/// non_existent:      Symbol::Unbound,
 /// ```
 #[derive(Debug, Clone, PartialEq)]
 pub(crate) enum Symbol<'db> {
@@ -37,21 +46,18 @@ impl<'db> Symbol<'db> {
         matches!(self, Symbol::Unbound)
     }
 
-    pub(crate) fn may_be_unbound(&self) -> bool {
+    pub(crate) fn possibly_unbound(&self) -> bool {
         match self {
-            Symbol::Type(_, Boundness::MayBeUnbound) | Symbol::Unbound => true,
+            Symbol::Type(_, Boundness::PossiblyUnbound) | Symbol::Unbound => true,
             Symbol::Type(_, Boundness::Bound) => false,
         }
     }
 
-    pub(crate) fn unwrap_or_unknown(&self) -> Type<'db> {
-        match self {
-            Symbol::Type(ty, _) => *ty,
-            Symbol::Unbound => Type::Unknown,
-        }
-    }
-
-    pub(crate) fn as_type(&self) -> Option<Type<'db>> {
+    /// Returns the type of the symbol, ignoring possible unboundness.
+    ///
+    /// If the symbol is *definitely* unbound, this function will return `None`. Otherwise,
+    /// if there is at least one control-flow path where the symbol is bound, return the type.
+    pub(crate) fn ignore_possibly_unbound(&self) -> Option<Type<'db>> {
         match self {
             Symbol::Type(ty, _) => Some(*ty),
             Symbol::Unbound => None,
@@ -61,28 +67,80 @@ impl<'db> Symbol<'db> {
     #[cfg(test)]
     #[track_caller]
     pub(crate) fn expect_type(self) -> Type<'db> {
-        self.as_type()
+        self.ignore_possibly_unbound()
             .expect("Expected a (possibly unbound) type, not an unbound symbol")
     }
 
     #[must_use]
-    pub(crate) fn replace_unbound_with(
-        self,
-        db: &'db dyn Db,
-        replacement: &Symbol<'db>,
-    ) -> Symbol<'db> {
-        match replacement {
-            Symbol::Type(replacement, _) => Symbol::Type(
-                match self {
-                    Symbol::Type(ty, Boundness::Bound) => ty,
-                    Symbol::Type(ty, Boundness::MayBeUnbound) => {
-                        UnionType::from_elements(db, [*replacement, ty])
-                    }
-                    Symbol::Unbound => *replacement,
-                },
-                Boundness::Bound,
-            ),
+    pub(crate) fn or_fall_back_to(self, db: &'db dyn Db, fallback: &Symbol<'db>) -> Symbol<'db> {
+        match fallback {
+            Symbol::Type(fallback_ty, fallback_boundness) => match self {
+                Symbol::Type(_, Boundness::Bound) => self,
+                Symbol::Type(ty, boundness @ Boundness::PossiblyUnbound) => Symbol::Type(
+                    UnionType::from_elements(db, [*fallback_ty, ty]),
+                    fallback_boundness.or(boundness),
+                ),
+                Symbol::Unbound => fallback.clone(),
+            },
             Symbol::Unbound => self,
         }
     }
 }
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use crate::types::tests::setup_db;
+
+    #[test]
+    fn test_symbol_or_fall_back_to() {
+        use Boundness::{Bound, PossiblyUnbound};
+
+        let db = setup_db();
+        let ty1 = Type::IntLiteral(1);
+        let ty2 = Type::IntLiteral(2);
+
+        // Start from an unbound symbol
+        assert_eq!(
+            Symbol::Unbound.or_fall_back_to(&db, &Symbol::Unbound),
+            Symbol::Unbound
+        );
+        assert_eq!(
+            Symbol::Unbound.or_fall_back_to(&db, &Symbol::Type(ty1, PossiblyUnbound)),
+            Symbol::Type(ty1, PossiblyUnbound)
+        );
+        assert_eq!(
+            Symbol::Unbound.or_fall_back_to(&db, &Symbol::Type(ty1, Bound)),
+            Symbol::Type(ty1, Bound)
+        );
+
+        // Start from a possibly unbound symbol
+        assert_eq!(
+            Symbol::Type(ty1, PossiblyUnbound).or_fall_back_to(&db, &Symbol::Unbound),
+            Symbol::Type(ty1, PossiblyUnbound)
+        );
+        assert_eq!(
+            Symbol::Type(ty1, PossiblyUnbound)
+                .or_fall_back_to(&db, &Symbol::Type(ty2, PossiblyUnbound)),
+            Symbol::Type(UnionType::from_elements(&db, [ty2, ty1]), PossiblyUnbound)
+        );
+        assert_eq!(
+            Symbol::Type(ty1, PossiblyUnbound).or_fall_back_to(&db, &Symbol::Type(ty2, Bound)),
+            Symbol::Type(UnionType::from_elements(&db, [ty2, ty1]), Bound)
+        );
+
+        // Start from a definitely bound symbol
+        assert_eq!(
+            Symbol::Type(ty1, Bound).or_fall_back_to(&db, &Symbol::Unbound),
+            Symbol::Type(ty1, Bound)
+        );
+        assert_eq!(
+            Symbol::Type(ty1, Bound).or_fall_back_to(&db, &Symbol::Type(ty2, PossiblyUnbound)),
+            Symbol::Type(ty1, Bound)
+        );
+        assert_eq!(
+            Symbol::Type(ty1, Bound).or_fall_back_to(&db, &Symbol::Type(ty2, Bound)),
+            Symbol::Type(ty1, Bound)
+        );
+    }
+}

crates/red_knot_python_semantic/src/types/builder.rs

@@ -128,7 +128,7 @@ impl<'db> IntersectionBuilder<'db> {
     pub(crate) fn new(db: &'db dyn Db) -> Self {
         Self {
             db,
-            intersections: vec![InnerIntersectionBuilder::new()],
+            intersections: vec![InnerIntersectionBuilder::default()],
         }
     }
 
@@ -231,10 +231,6 @@ struct InnerIntersectionBuilder<'db> {
 }
 
 impl<'db> InnerIntersectionBuilder<'db> {
-    fn new() -> Self {
-        Self::default()
-    }
-
     /// Adds a positive type to this intersection.
     fn add_positive(&mut self, db: &'db dyn Db, new_positive: Type<'db>) {
         if let Type::Intersection(other) = new_positive {
@@ -253,7 +249,7 @@ impl<'db> InnerIntersectionBuilder<'db> {
                         .iter()
                         .find(|element| element.is_boolean_literal())
                     {
-                        *self = Self::new();
+                        *self = Self::default();
                         self.positive.insert(Type::BooleanLiteral(!value));
                         return;
                     }
@@ -272,7 +268,7 @@ impl<'db> InnerIntersectionBuilder<'db> {
                 }
                 // A & B = Never    if A and B are disjoint
                 if new_positive.is_disjoint_from(db, *existing_positive) {
-                    *self = Self::new();
+                    *self = Self::default();
                     self.positive.insert(Type::Never);
                     return;
                 }
@@ -285,7 +281,7 @@ impl<'db> InnerIntersectionBuilder<'db> {
             for (index, existing_negative) in self.negative.iter().enumerate() {
                 // S & ~T = Never    if S <: T
                 if new_positive.is_subtype_of(db, *existing_negative) {
-                    *self = Self::new();
+                    *self = Self::default();
                     self.positive.insert(Type::Never);
                     return;
                 }
@@ -326,7 +322,7 @@ impl<'db> InnerIntersectionBuilder<'db> {
                     .iter()
                     .any(|pos| *pos == KnownClass::Bool.to_instance(db)) =>
             {
-                *self = Self::new();
+                *self = Self::default();
                 self.positive.insert(Type::BooleanLiteral(!bool));
             }
             _ => {
@@ -348,7 +344,7 @@ impl<'db> InnerIntersectionBuilder<'db> {
                 for existing_positive in &self.positive {
                     // S & ~T = Never    if S <: T
                     if existing_positive.is_subtype_of(db, new_negative) {
-                        *self = Self::new();
+                        *self = Self::default();
                         self.positive.insert(Type::Never);
                         return;
                     }
@@ -383,7 +379,7 @@ mod tests {
     use crate::program::{Program, SearchPathSettings};
     use crate::python_version::PythonVersion;
     use crate::stdlib::typing_symbol;
-    use crate::types::{global_symbol, KnownClass, StringLiteralType, UnionBuilder};
+    use crate::types::{global_symbol, KnownClass, UnionBuilder};
     use crate::ProgramSettings;
     use ruff_db::files::system_path_to_file;
     use ruff_db::system::{DbWithTestSystem, SystemPathBuf};
@@ -775,7 +771,7 @@ mod tests {
             .build();
         assert_eq!(ty, s);
 
-        let literal = Type::StringLiteral(StringLiteralType::new(&db, "a"));
+        let literal = Type::string_literal(&db, "a");
         let expected = IntersectionBuilder::new(&db)
             .add_positive(s)
             .add_negative(literal)
@@ -878,7 +874,7 @@ mod tests {
 
         let ty = IntersectionBuilder::new(&db)
             .add_positive(s)
-            .add_negative(Type::StringLiteral(StringLiteralType::new(&db, "a")))
+            .add_negative(Type::string_literal(&db, "a"))
             .add_negative(t)
             .build();
         assert_eq!(ty, Type::Never);
@@ -912,7 +908,7 @@ mod tests {
         let db = setup_db();
 
         let t_p = KnownClass::Int.to_instance(&db);
-        let t_n = Type::StringLiteral(StringLiteralType::new(&db, "t_n"));
+        let t_n = Type::string_literal(&db, "t_n");
 
         let ty = IntersectionBuilder::new(&db)
             .add_positive(t_p)

crates/red_knot_python_semantic/src/types/diagnostic.rs

@@ -73,10 +73,6 @@ pub struct TypeCheckDiagnostics {
 }
 
 impl TypeCheckDiagnostics {
-    pub fn new() -> Self {
-        Self { inner: Vec::new() }
-    }
-
     pub(super) fn push(&mut self, diagnostic: TypeCheckDiagnostic) {
         self.inner.push(Arc::new(diagnostic));
     }
@@ -148,7 +144,7 @@ impl<'db> TypeCheckDiagnosticsBuilder<'db> {
         Self {
             db,
             file,
-            diagnostics: TypeCheckDiagnostics::new(),
+            diagnostics: TypeCheckDiagnostics::default(),
         }
     }
 

crates/red_knot_python_semantic/src/types/display.rs

@@ -1,13 +1,14 @@
 //! Display implementations for types.
 
-use std::fmt::{self, Display, Formatter};
+use std::fmt::{self, Display, Formatter, Write};
 
 use ruff_db::display::FormatterJoinExtension;
 use ruff_python_ast::str::Quote;
 use ruff_python_literal::escape::AsciiEscape;
 
 use crate::types::{
-    ClassLiteralType, InstanceType, IntersectionType, KnownClass, SubclassOfType, Type, UnionType,
+    ClassLiteralType, InstanceType, IntersectionType, KnownClass, StringLiteralType,
+    SubclassOfType, Type, UnionType,
 };
 use crate::Db;
 use rustc_hash::FxHashMap;
@@ -66,10 +67,13 @@ impl Display for DisplayRepresentation<'_> {
             Type::Any => f.write_str("Any"),
             Type::Never => f.write_str("Never"),
             Type::Unknown => f.write_str("Unknown"),
-            Type::Instance(InstanceType { class })
-                if class.is_known(self.db, KnownClass::NoneType) =>
-            {
-                f.write_str("None")
+            Type::Instance(InstanceType { class }) => {
+                let representation = match class.known(self.db) {
+                    Some(KnownClass::NoneType) => "None",
+                    Some(KnownClass::NoDefaultType) => "NoDefault",
+                    _ => class.name(self.db),
+                };
+                f.write_str(representation)
             }
             // `[Type::Todo]`'s display should be explicit that is not a valid display of
             // any other type
@@ -82,16 +86,13 @@ impl Display for DisplayRepresentation<'_> {
             Type::SubclassOf(SubclassOfType { class }) => {
                 write!(f, "type[{}]", class.name(self.db))
             }
-            Type::Instance(InstanceType { class }) => f.write_str(class.name(self.db)),
-            Type::KnownInstance(known_instance) => f.write_str(known_instance.as_str()),
+            Type::KnownInstance(known_instance) => f.write_str(known_instance.repr(self.db)),
             Type::FunctionLiteral(function) => f.write_str(function.name(self.db)),
             Type::Union(union) => union.display(self.db).fmt(f),
             Type::Intersection(intersection) => intersection.display(self.db).fmt(f),
             Type::IntLiteral(n) => n.fmt(f),
             Type::BooleanLiteral(boolean) => f.write_str(if boolean { "True" } else { "False" }),
-            Type::StringLiteral(string) => {
-                write!(f, r#""{}""#, string.value(self.db).replace('"', r#"\""#))
-            }
+            Type::StringLiteral(string) => string.display(self.db).fmt(f),
             Type::LiteralString => f.write_str("LiteralString"),
             Type::BytesLiteral(bytes) => {
                 let escape =
@@ -326,15 +327,40 @@ impl<'db> Display for DisplayTypeArray<'_, 'db> {
     }
 }
 
+impl<'db> StringLiteralType<'db> {
+    fn display(&'db self, db: &'db dyn Db) -> DisplayStringLiteralType<'db> {
+        DisplayStringLiteralType { db, ty: self }
+    }
+}
+
+struct DisplayStringLiteralType<'db> {
+    ty: &'db StringLiteralType<'db>,
+    db: &'db dyn Db,
+}
+
+impl Display for DisplayStringLiteralType<'_> {
+    fn fmt(&self, f: &mut Formatter<'_>) -> fmt::Result {
+        let value = self.ty.value(self.db);
+        f.write_char('"')?;
+        for ch in value.chars() {
+            match ch {
+                // `escape_debug` will escape even single quotes, which is not necessary for our
+                // use case as we are already using double quotes to wrap the string.
+                '\'' => f.write_char('\'')?,
+                _ => write!(f, "{}", ch.escape_debug())?,
+            }
+        }
+        f.write_char('"')
+    }
+}
+
 #[cfg(test)]
 mod tests {
     use ruff_db::files::system_path_to_file;
     use ruff_db::system::{DbWithTestSystem, SystemPathBuf};
 
     use crate::db::tests::TestDb;
-    use crate::types::{
-        global_symbol, BytesLiteralType, SliceLiteralType, StringLiteralType, Type, UnionType,
-    };
+    use crate::types::{global_symbol, SliceLiteralType, StringLiteralType, Type, UnionType};
     use crate::{Program, ProgramSettings, PythonVersion, SearchPathSettings};
 
     fn setup_db() -> TestDb {
@@ -380,12 +406,12 @@ mod tests {
             Type::Unknown,
             Type::IntLiteral(-1),
             global_symbol(&db, mod_file, "A").expect_type(),
-            Type::StringLiteral(StringLiteralType::new(&db, "A")),
-            Type::BytesLiteral(BytesLiteralType::new(&db, [0u8].as_slice())),
-            Type::BytesLiteral(BytesLiteralType::new(&db, [7u8].as_slice())),
+            Type::string_literal(&db, "A"),
+            Type::bytes_literal(&db, &[0u8]),
+            Type::bytes_literal(&db, &[7u8]),
             Type::IntLiteral(0),
             Type::IntLiteral(1),
-            Type::StringLiteral(StringLiteralType::new(&db, "B")),
+            Type::string_literal(&db, "B"),
             global_symbol(&db, mod_file, "foo").expect_type(),
             global_symbol(&db, mod_file, "bar").expect_type(),
             global_symbol(&db, mod_file, "B").expect_type(),
@@ -451,4 +477,28 @@ mod tests {
             "slice[None, None, Literal[2]]"
         );
     }
+
+    #[test]
+    fn string_literal_display() {
+        let db = setup_db();
+
+        assert_eq!(
+            Type::StringLiteral(StringLiteralType::new(&db, r"\n"))
+                .display(&db)
+                .to_string(),
+            r#"Literal["\\n"]"#
+        );
+        assert_eq!(
+            Type::StringLiteral(StringLiteralType::new(&db, "'"))
+                .display(&db)
+                .to_string(),
+            r#"Literal["'"]"#
+        );
+        assert_eq!(
+            Type::StringLiteral(StringLiteralType::new(&db, r#"""#))
+                .display(&db)
+                .to_string(),
+            r#"Literal["\""]"#
+        );
+    }
 }

crates/red_knot_python_semantic/src/types/mro.rs

@@ -1,7 +1,6 @@
 use std::collections::VecDeque;
 use std::ops::Deref;
 
-use indexmap::IndexSet;
 use itertools::Either;
 use rustc_hash::FxHashSet;
 
@@ -26,22 +25,30 @@ impl<'db> Mro<'db> {
     /// (We emit a diagnostic warning about the runtime `TypeError` in
     /// [`super::infer::TypeInferenceBuilder::infer_region_scope`].)
     pub(super) fn of_class(db: &'db dyn Db, class: Class<'db>) -> Result<Self, MroError<'db>> {
-        Self::of_class_impl(db, class).map_err(|error_kind| {
-            let fallback_mro = Self::from([
-                ClassBase::Class(class),
-                ClassBase::Unknown,
-                ClassBase::object(db),
-            ]);
-            MroError {
-                kind: error_kind,
-                fallback_mro,
-            }
+        Self::of_class_impl(db, class).map_err(|error_kind| MroError {
+            kind: error_kind,
+            fallback_mro: Self::from_error(db, class),
         })
     }
 
+    pub(super) fn from_error(db: &'db dyn Db, class: Class<'db>) -> Self {
+        Self::from([
+            ClassBase::Class(class),
+            ClassBase::Unknown,
+            ClassBase::object(db),
+        ])
+    }
+
     fn of_class_impl(db: &'db dyn Db, class: Class<'db>) -> Result<Self, MroErrorKind<'db>> {
         let class_bases = class.explicit_bases(db);
 
+        if !class_bases.is_empty() && class.is_cyclically_defined(db) {
+            // We emit errors for cyclically defined classes elsewhere.
+            // It's important that we don't even try to infer the MRO for a cyclically defined class,
+            // or we'll end up in an infinite loop.
+            return Ok(Mro::from_error(db, class));
+        }
+
         match class_bases {
             // `builtins.object` is the special case:
             // the only class in Python that has an MRO with length <2
@@ -77,11 +84,6 @@ impl<'db> Mro<'db> {
                         Err(MroErrorKind::InvalidBases(bases_info))
                     },
                     |single_base| {
-                        if let ClassBase::Class(class_base) = single_base {
-                            if class_is_cyclically_defined(db, class_base) {
-                                return Err(MroErrorKind::CyclicClassDefinition);
-                            }
-                        }
                         let mro = std::iter::once(ClassBase::Class(class))
                             .chain(single_base.mro(db))
                             .collect();
@@ -96,10 +98,6 @@ impl<'db> Mro<'db> {
             // what MRO Python will give this class at runtime
             // (if an MRO is indeed resolvable at all!)
             multiple_bases => {
-                if class_is_cyclically_defined(db, class) {
-                    return Err(MroErrorKind::CyclicClassDefinition);
-                }
-
                 let mut valid_bases = vec![];
                 let mut invalid_bases = vec![];
 
@@ -282,13 +280,6 @@ pub(super) enum MroErrorKind<'db> {
     /// Each index is the index of a node representing an invalid base.
     InvalidBases(Box<[(usize, Type<'db>)]>),
 
-    /// The class inherits from itself!
-    ///
-    /// This is very unlikely to happen in working real-world code,
-    /// but it's important to explicitly account for it.
-    /// If we don't, there's a possibility of an infinite loop and a panic.
-    CyclicClassDefinition,
-
     /// The class has one or more duplicate bases.
     ///
     /// This variant records the indices and [`Class`]es
@@ -349,7 +340,7 @@ impl<'db> ClassBase<'db> {
     /// Return a `ClassBase` representing the class `builtins.object`
     fn object(db: &'db dyn Db) -> Self {
         KnownClass::Object
-            .to_class(db)
+            .to_class_literal(db)
             .into_class_literal()
             .map_or(Self::Unknown, |ClassLiteralType { class }| {
                 Self::Class(class)
@@ -380,7 +371,9 @@ impl<'db> ClassBase<'db> {
             | Type::ModuleLiteral(_)
             | Type::SubclassOf(_) => None,
             Type::KnownInstance(known_instance) => match known_instance {
-                KnownInstanceType::Literal => None,
+                KnownInstanceType::TypeVar(_)
+                | KnownInstanceType::Literal
+                | KnownInstanceType::Optional => None,
             },
         }
     }
@@ -414,7 +407,7 @@ impl<'db> From<ClassBase<'db>> for Type<'db> {
             ClassBase::Any => Type::Any,
             ClassBase::Todo => Type::Todo,
             ClassBase::Unknown => Type::Unknown,
-            ClassBase::Class(class) => Type::ClassLiteral(ClassLiteralType { class }),
+            ClassBase::Class(class) => Type::class_literal(class),
         }
     }
 }
@@ -460,46 +453,3 @@ fn c3_merge(mut sequences: Vec<VecDeque<ClassBase>>) -> Option<Mro> {
         }
     }
 }
-
-/// Return `true` if this class appears to be a cyclic definition,
-/// i.e., it inherits either directly or indirectly from itself.
-///
-/// A class definition like this will fail at runtime,
-/// but we must be resilient to it or we could panic.
-fn class_is_cyclically_defined(db: &dyn Db, class: Class) -> bool {
-    fn is_cyclically_defined_recursive<'db>(
-        db: &'db dyn Db,
-        class: Class<'db>,
-        classes_to_watch: &mut IndexSet<Class<'db>>,
-    ) -> bool {
-        if !classes_to_watch.insert(class) {
-            return true;
-        }
-        for explicit_base_class in class
-            .explicit_bases(db)
-            .iter()
-            .copied()
-            .filter_map(Type::into_class_literal)
-            .map(|ClassLiteralType { class }| class)
-        {
-            // Each base must be considered in isolation.
-            // This is due to the fact that if a class uses multiple inheritance,
-            // there could easily be a situation where two bases have the same class in their MROs;
-            // that isn't enough to constitute the class being cyclically defined.
-            let classes_to_watch_len = classes_to_watch.len();
-            if is_cyclically_defined_recursive(db, explicit_base_class, classes_to_watch) {
-                return true;
-            }
-            classes_to_watch.truncate(classes_to_watch_len);
-        }
-        false
-    }
-
-    class
-        .explicit_bases(db)
-        .iter()
-        .copied()
-        .filter_map(Type::into_class_literal)
-        .map(|ClassLiteralType { class }| class)
-        .any(|base_class| is_cyclically_defined_recursive(db, base_class, &mut IndexSet::default()))
-}

crates/red_knot_python_semantic/src/types/narrow.rs

@@ -1,11 +1,11 @@
-use crate::semantic_index::ast_ids::HasScopedAstId;
+use crate::semantic_index::ast_ids::HasScopedExpressionId;
 use crate::semantic_index::constraint::{Constraint, ConstraintNode, PatternConstraint};
 use crate::semantic_index::definition::Definition;
 use crate::semantic_index::expression::Expression;
 use crate::semantic_index::symbol::{ScopeId, ScopedSymbolId, SymbolTable};
 use crate::semantic_index::symbol_table;
 use crate::types::{
-    infer_expression_types, ClassLiteralType, InstanceType, IntersectionBuilder, KnownClass,
+    infer_expression_types, ClassLiteralType, IntersectionBuilder, KnownClass,
     KnownConstraintFunction, KnownFunction, Truthiness, Type, UnionBuilder,
 };
 use crate::Db;
@@ -257,17 +257,26 @@ impl<'db> NarrowingConstraintsBuilder<'db> {
         expression: Expression<'db>,
         is_positive: bool,
     ) -> Option<NarrowingConstraints<'db>> {
+        fn is_narrowing_target_candidate(expr: &ast::Expr) -> bool {
+            matches!(expr, ast::Expr::Name(_) | ast::Expr::Call(_))
+        }
+
         let ast::ExprCompare {
             range: _,
             left,
             ops,
             comparators,
         } = expr_compare;
-        if !left.is_name_expr() && comparators.iter().all(|c| !c.is_name_expr()) {
-            // If none of the comparators are name expressions,
-            // we have no symbol to narrow down the type of.
+
+        // Performance optimization: early return if there are no potential narrowing targets.
+        if !is_narrowing_target_candidate(left)
+            && comparators
+                .iter()
+                .all(|c| !is_narrowing_target_candidate(c))
+        {
             return None;
         }
+
         if !is_positive && comparators.len() > 1 {
             // We can't negate a constraint made by a multi-comparator expression, since we can't
             // know which comparison part is the one being negated.
@@ -283,42 +292,85 @@ impl<'db> NarrowingConstraintsBuilder<'db> {
             .tuple_windows::<(&ruff_python_ast::Expr, &ruff_python_ast::Expr)>();
         let mut constraints = NarrowingConstraints::default();
         for (op, (left, right)) in std::iter::zip(&**ops, comparator_tuples) {
-            if let ast::Expr::Name(ast::ExprName {
-                range: _,
-                id,
-                ctx: _,
-            }) = left
-            {
-                // SAFETY: we should always have a symbol for every Name node.
-                let symbol = self.symbols().symbol_id_by_name(id).unwrap();
-                let rhs_ty = inference.expression_ty(right.scoped_ast_id(self.db, scope));
+            let rhs_ty = inference.expression_ty(right.scoped_expression_id(self.db, scope));
 
-                match if is_positive { *op } else { op.negate() } {
-                    ast::CmpOp::IsNot => {
-                        if rhs_ty.is_singleton(self.db) {
-                            let ty = IntersectionBuilder::new(self.db)
-                                .add_negative(rhs_ty)
-                                .build();
-                            constraints.insert(symbol, ty);
-                        } else {
-                            // Non-singletons cannot be safely narrowed using `is not`
+            match left {
+                ast::Expr::Name(ast::ExprName {
+                    range: _,
+                    id,
+                    ctx: _,
+                }) => {
+                    let symbol = self
+                        .symbols()
+                        .symbol_id_by_name(id)
+                        .expect("Should always have a symbol for every Name node");
+
+                    match if is_positive { *op } else { op.negate() } {
+                        ast::CmpOp::IsNot => {
+                            if rhs_ty.is_singleton(self.db) {
+                                let ty = IntersectionBuilder::new(self.db)
+                                    .add_negative(rhs_ty)
+                                    .build();
+                                constraints.insert(symbol, ty);
+                            } else {
+                                // Non-singletons cannot be safely narrowed using `is not`
+                            }
                         }
-                    }
-                    ast::CmpOp::Is => {
-                        constraints.insert(symbol, rhs_ty);
-                    }
-                    ast::CmpOp::NotEq => {
-                        if rhs_ty.is_single_valued(self.db) {
-                            let ty = IntersectionBuilder::new(self.db)
-                                .add_negative(rhs_ty)
-                                .build();
-                            constraints.insert(symbol, ty);
+                        ast::CmpOp::Is => {
+                            constraints.insert(symbol, rhs_ty);
+                        }
+                        ast::CmpOp::NotEq => {
+                            if rhs_ty.is_single_valued(self.db) {
+                                let ty = IntersectionBuilder::new(self.db)
+                                    .add_negative(rhs_ty)
+                                    .build();
+                                constraints.insert(symbol, ty);
+                            }
+                        }
+                        _ => {
+                            // TODO other comparison types
                         }
-                    }
-                    _ => {
-                        // TODO other comparison types
                     }
                 }
+                ast::Expr::Call(ast::ExprCall {
+                    range: _,
+                    func: callable,
+                    arguments:
+                        ast::Arguments {
+                            args,
+                            keywords,
+                            range: _,
+                        },
+                }) if rhs_ty.is_class_literal() && keywords.is_empty() => {
+                    let [ast::Expr::Name(ast::ExprName { id, .. })] = &**args else {
+                        continue;
+                    };
+
+                    let is_valid_constraint = if is_positive {
+                        op == &ast::CmpOp::Is
+                    } else {
+                        op == &ast::CmpOp::IsNot
+                    };
+
+                    if !is_valid_constraint {
+                        continue;
+                    }
+
+                    let callable_ty =
+                        inference.expression_ty(callable.scoped_expression_id(self.db, scope));
+
+                    if callable_ty
+                        .into_class_literal()
+                        .is_some_and(|c| c.class.is_known(self.db, KnownClass::Type))
+                    {
+                        let symbol = self
+                            .symbols()
+                            .symbol_id_by_name(id)
+                            .expect("Should always have a symbol for every Name node");
+                        constraints.insert(symbol, rhs_ty.to_instance(self.db));
+                    }
+                }
+                _ => {}
             }
         }
         Some(constraints)
@@ -336,7 +388,7 @@ impl<'db> NarrowingConstraintsBuilder<'db> {
         // TODO: add support for PEP 604 union types on the right hand side of `isinstance`
         // and `issubclass`, for example `isinstance(x, str | (int | float))`.
         match inference
-            .expression_ty(expr_call.func.scoped_ast_id(self.db, scope))
+            .expression_ty(expr_call.func.scoped_expression_id(self.db, scope))
             .into_function_literal()
             .and_then(|f| f.known(self.db))
             .and_then(KnownFunction::constraint_function)
@@ -348,19 +400,17 @@ impl<'db> NarrowingConstraintsBuilder<'db> {
                     let symbol = self.symbols().symbol_id_by_name(id).unwrap();
 
                     let class_info_ty =
-                        inference.expression_ty(class_info.scoped_ast_id(self.db, scope));
+                        inference.expression_ty(class_info.scoped_expression_id(self.db, scope));
 
                     let to_constraint = match function {
                         KnownConstraintFunction::IsInstance => {
                             |class_literal: ClassLiteralType<'db>| {
-                                Type::Instance(InstanceType {
-                                    class: class_literal.class,
-                                })
+                                Type::instance(class_literal.class)
                             }
                         }
                         KnownConstraintFunction::IsSubclass => {
                             |class_literal: ClassLiteralType<'db>| {
-                                Type::SubclassOf(class_literal.to_subclass_of_type())
+                                Type::subclass_of(class_literal.class)
                             }
                         }
                     };
@@ -416,7 +466,7 @@ impl<'db> NarrowingConstraintsBuilder<'db> {
             // filter our arms with statically known truthiness
             .filter(|expr| {
                 inference
-                    .expression_ty(expr.scoped_ast_id(self.db, scope))
+                    .expression_ty(expr.scoped_expression_id(self.db, scope))
                     .bool(self.db)
                     != match expr_bool_op.op {
                         BoolOp::And => Truthiness::AlwaysTrue,

crates/red_knot_python_semantic/src/types/signatures.rs

@@ -0,0 +1,480 @@
+#![allow(dead_code)]
+use super::{definition_expression_ty, Type};
+use crate::semantic_index::definition::Definition;
+use crate::Db;
+use ruff_python_ast::{self as ast, name::Name};
+
+/// A typed callable signature.
+#[derive(Clone, Debug, PartialEq, Eq)]
+pub(crate) struct Signature<'db> {
+    parameters: Parameters<'db>,
+
+    /// Annotated return type (Unknown if no annotation.)
+    pub(crate) return_ty: Type<'db>,
+}
+
+impl<'db> Signature<'db> {
+    /// Return a todo signature: (*args: Todo, **kwargs: Todo) -> Todo
+    pub(crate) fn todo() -> Self {
+        Self {
+            parameters: Parameters::todo(),
+            return_ty: Type::Todo,
+        }
+    }
+
+    /// Return a typed signature from a function definition.
+    pub(super) fn from_function(
+        db: &'db dyn Db,
+        definition: Definition<'db>,
+        function_node: &'db ast::StmtFunctionDef,
+    ) -> Self {
+        let return_ty = function_node
+            .returns
+            .as_ref()
+            .map(|returns| {
+                if function_node.is_async {
+                    // TODO: generic `types.CoroutineType`!
+                    Type::Todo
+                } else {
+                    definition_expression_ty(db, definition, returns.as_ref())
+                }
+            })
+            .unwrap_or(Type::Unknown);
+
+        Self {
+            parameters: Parameters::from_parameters(
+                db,
+                definition,
+                function_node.parameters.as_ref(),
+            ),
+            return_ty,
+        }
+    }
+}
+
+/// The parameters portion of a typed signature.
+///
+/// The ordering of parameters is always as given in this struct: first positional-only parameters,
+/// then positional-or-keyword, then optionally the variadic parameter, then keyword-only
+/// parameters, and last, optionally the variadic keywords parameter.
+#[derive(Clone, Debug, Default, PartialEq, Eq)]
+pub(super) struct Parameters<'db> {
+    /// Parameters which may only be filled by positional arguments.
+    positional_only: Box<[ParameterWithDefault<'db>]>,
+
+    /// Parameters which may be filled by positional or keyword arguments.
+    positional_or_keyword: Box<[ParameterWithDefault<'db>]>,
+
+    /// The `*args` variadic parameter, if any.
+    variadic: Option<Parameter<'db>>,
+
+    /// Parameters which may only be filled by keyword arguments.
+    keyword_only: Box<[ParameterWithDefault<'db>]>,
+
+    /// The `**kwargs` variadic keywords parameter, if any.
+    keywords: Option<Parameter<'db>>,
+}
+
+impl<'db> Parameters<'db> {
+    /// Return todo parameters: (*args: Todo, **kwargs: Todo)
+    fn todo() -> Self {
+        Self {
+            variadic: Some(Parameter {
+                name: Some(Name::new_static("args")),
+                annotated_ty: Type::Todo,
+            }),
+            keywords: Some(Parameter {
+                name: Some(Name::new_static("kwargs")),
+                annotated_ty: Type::Todo,
+            }),
+            ..Default::default()
+        }
+    }
+
+    fn from_parameters(
+        db: &'db dyn Db,
+        definition: Definition<'db>,
+        parameters: &'db ast::Parameters,
+    ) -> Self {
+        let ast::Parameters {
+            posonlyargs,
+            args,
+            vararg,
+            kwonlyargs,
+            kwarg,
+            range: _,
+        } = parameters;
+        let positional_only = posonlyargs
+            .iter()
+            .map(|arg| ParameterWithDefault::from_node(db, definition, arg))
+            .collect();
+        let positional_or_keyword = args
+            .iter()
+            .map(|arg| ParameterWithDefault::from_node(db, definition, arg))
+            .collect();
+        let variadic = vararg
+            .as_ref()
+            .map(|arg| Parameter::from_node(db, definition, arg));
+        let keyword_only = kwonlyargs
+            .iter()
+            .map(|arg| ParameterWithDefault::from_node(db, definition, arg))
+            .collect();
+        let keywords = kwarg
+            .as_ref()
+            .map(|arg| Parameter::from_node(db, definition, arg));
+        Self {
+            positional_only,
+            positional_or_keyword,
+            variadic,
+            keyword_only,
+            keywords,
+        }
+    }
+}
+
+/// A single parameter of a typed signature, with optional default value.
+#[derive(Clone, Debug, PartialEq, Eq)]
+pub(super) struct ParameterWithDefault<'db> {
+    parameter: Parameter<'db>,
+
+    /// Type of the default value, if any.
+    default_ty: Option<Type<'db>>,
+}
+
+impl<'db> ParameterWithDefault<'db> {
+    fn from_node(
+        db: &'db dyn Db,
+        definition: Definition<'db>,
+        parameter_with_default: &'db ast::ParameterWithDefault,
+    ) -> Self {
+        Self {
+            default_ty: parameter_with_default
+                .default
+                .as_deref()
+                .map(|default| definition_expression_ty(db, definition, default)),
+            parameter: Parameter::from_node(db, definition, &parameter_with_default.parameter),
+        }
+    }
+}
+
+/// A single parameter of a typed signature.
+#[derive(Clone, Debug, PartialEq, Eq)]
+pub(super) struct Parameter<'db> {
+    /// Parameter name.
+    ///
+    /// It is possible for signatures to be defined in ways that leave positional-only parameters
+    /// nameless (e.g. via `Callable` annotations).
+    name: Option<Name>,
+
+    /// Annotated type of the parameter (Unknown if no annotation.)
+    annotated_ty: Type<'db>,
+}
+
+impl<'db> Parameter<'db> {
+    fn from_node(
+        db: &'db dyn Db,
+        definition: Definition<'db>,
+        parameter: &'db ast::Parameter,
+    ) -> Self {
+        Parameter {
+            name: Some(parameter.name.id.clone()),
+            annotated_ty: parameter
+                .annotation
+                .as_deref()
+                .map(|annotation| definition_expression_ty(db, definition, annotation))
+                .unwrap_or(Type::Unknown),
+        }
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use crate::db::tests::TestDb;
+    use crate::program::{Program, SearchPathSettings};
+    use crate::python_version::PythonVersion;
+    use crate::types::{global_symbol, FunctionType};
+    use crate::ProgramSettings;
+    use ruff_db::system::{DbWithTestSystem, SystemPathBuf};
+
+    pub(crate) fn setup_db() -> TestDb {
+        let db = TestDb::new();
+
+        let src_root = SystemPathBuf::from("/src");
+        db.memory_file_system()
+            .create_directory_all(&src_root)
+            .unwrap();
+
+        Program::from_settings(
+            &db,
+            &ProgramSettings {
+                target_version: PythonVersion::default(),
+                search_paths: SearchPathSettings::new(src_root),
+            },
+        )
+        .expect("Valid search path settings");
+
+        db
+    }
+
+    #[track_caller]
+    fn get_function_f<'db>(db: &'db TestDb, file: &'static str) -> FunctionType<'db> {
+        let module = ruff_db::files::system_path_to_file(db, file).unwrap();
+        global_symbol(db, module, "f")
+            .expect_type()
+            .expect_function_literal()
+    }
+
+    #[track_caller]
+    fn assert_param_with_default<'db>(
+        db: &'db TestDb,
+        param_with_default: &ParameterWithDefault<'db>,
+        expected_name: &'static str,
+        expected_annotation_ty_display: &'static str,
+        expected_default_ty_display: Option<&'static str>,
+    ) {
+        assert_eq!(
+            param_with_default
+                .default_ty
+                .map(|ty| ty.display(db).to_string()),
+            expected_default_ty_display.map(ToString::to_string)
+        );
+        assert_param(
+            db,
+            &param_with_default.parameter,
+            expected_name,
+            expected_annotation_ty_display,
+        );
+    }
+
+    #[track_caller]
+    fn assert_param<'db>(
+        db: &'db TestDb,
+        param: &Parameter<'db>,
+        expected_name: &'static str,
+        expected_annotation_ty_display: &'static str,
+    ) {
+        assert_eq!(param.name.as_ref().unwrap(), expected_name);
+        assert_eq!(
+            param.annotated_ty.display(db).to_string(),
+            expected_annotation_ty_display
+        );
+    }
+
+    #[test]
+    fn empty() {
+        let mut db = setup_db();
+        db.write_dedented("/src/a.py", "def f(): ...").unwrap();
+        let func = get_function_f(&db, "/src/a.py");
+
+        let sig = func.internal_signature(&db);
+
+        assert_eq!(sig.return_ty.display(&db).to_string(), "Unknown");
+        let params = sig.parameters;
+        assert!(params.positional_only.is_empty());
+        assert!(params.positional_or_keyword.is_empty());
+        assert!(params.variadic.is_none());
+        assert!(params.keyword_only.is_empty());
+        assert!(params.keywords.is_none());
+    }
+
+    #[test]
+    #[allow(clippy::many_single_char_names)]
+    fn full() {
+        let mut db = setup_db();
+        db.write_dedented(
+            "/src/a.py",
+            "
+            def f(a, b: int, c = 1, d: int = 2, /,
+                  e = 3, f: Literal[4] = 4, *args: object,
+                  g = 5, h: Literal[6] = 6, **kwargs: str) -> bytes: ...
+            ",
+        )
+        .unwrap();
+        let func = get_function_f(&db, "/src/a.py");
+
+        let sig = func.internal_signature(&db);
+
+        assert_eq!(sig.return_ty.display(&db).to_string(), "bytes");
+        let params = sig.parameters;
+        let [a, b, c, d] = &params.positional_only[..] else {
+            panic!("expected four positional-only parameters");
+        };
+        let [e, f] = &params.positional_or_keyword[..] else {
+            panic!("expected two positional-or-keyword parameters");
+        };
+        let Some(args) = params.variadic else {
+            panic!("expected a variadic parameter");
+        };
+        let [g, h] = &params.keyword_only[..] else {
+            panic!("expected two keyword-only parameters");
+        };
+        let Some(kwargs) = params.keywords else {
+            panic!("expected a kwargs parameter");
+        };
+
+        assert_param_with_default(&db, a, "a", "Unknown", None);
+        assert_param_with_default(&db, b, "b", "int", None);
+        assert_param_with_default(&db, c, "c", "Unknown", Some("Literal[1]"));
+        assert_param_with_default(&db, d, "d", "int", Some("Literal[2]"));
+        assert_param_with_default(&db, e, "e", "Unknown", Some("Literal[3]"));
+        assert_param_with_default(&db, f, "f", "Literal[4]", Some("Literal[4]"));
+        assert_param_with_default(&db, g, "g", "Unknown", Some("Literal[5]"));
+        assert_param_with_default(&db, h, "h", "Literal[6]", Some("Literal[6]"));
+        assert_param(&db, &args, "args", "object");
+        assert_param(&db, &kwargs, "kwargs", "str");
+    }
+
+    #[test]
+    fn not_deferred() {
+        let mut db = setup_db();
+        db.write_dedented(
+            "/src/a.py",
+            "
+            class A: ...
+            class B: ...
+
+            alias = A
+
+            def f(a: alias): ...
+
+            alias = B
+            ",
+        )
+        .unwrap();
+        let func = get_function_f(&db, "/src/a.py");
+
+        let sig = func.internal_signature(&db);
+
+        let [a] = &sig.parameters.positional_or_keyword[..] else {
+            panic!("expected one positional-or-keyword parameter");
+        };
+        // Parameter resolution not deferred; we should see A not B
+        assert_param_with_default(&db, a, "a", "A", None);
+    }
+
+    #[test]
+    fn deferred_in_stub() {
+        let mut db = setup_db();
+        db.write_dedented(
+            "/src/a.pyi",
+            "
+            class A: ...
+            class B: ...
+
+            alias = A
+
+            def f(a: alias): ...
+
+            alias = B
+            ",
+        )
+        .unwrap();
+        let func = get_function_f(&db, "/src/a.pyi");
+
+        let sig = func.internal_signature(&db);
+
+        let [a] = &sig.parameters.positional_or_keyword[..] else {
+            panic!("expected one positional-or-keyword parameter");
+        };
+        // Parameter resolution deferred; we should see B
+        assert_param_with_default(&db, a, "a", "B", None);
+    }
+
+    #[test]
+    fn generic_not_deferred() {
+        let mut db = setup_db();
+        db.write_dedented(
+            "/src/a.py",
+            "
+            class A: ...
+            class B: ...
+
+            alias = A
+
+            def f[T](a: alias, b: T) -> T: ...
+
+            alias = B
+            ",
+        )
+        .unwrap();
+        let func = get_function_f(&db, "/src/a.py");
+
+        let sig = func.internal_signature(&db);
+
+        let [a, b] = &sig.parameters.positional_or_keyword[..] else {
+            panic!("expected two positional-or-keyword parameters");
+        };
+        // TODO resolution should not be deferred; we should see A not B
+        assert_param_with_default(&db, a, "a", "B", None);
+        assert_param_with_default(&db, b, "b", "T", None);
+    }
+
+    #[test]
+    fn generic_deferred_in_stub() {
+        let mut db = setup_db();
+        db.write_dedented(
+            "/src/a.pyi",
+            "
+            class A: ...
+            class B: ...
+
+            alias = A
+
+            def f[T](a: alias, b: T) -> T: ...
+
+            alias = B
+            ",
+        )
+        .unwrap();
+        let func = get_function_f(&db, "/src/a.pyi");
+
+        let sig = func.internal_signature(&db);
+
+        let [a, b] = &sig.parameters.positional_or_keyword[..] else {
+            panic!("expected two positional-or-keyword parameters");
+        };
+        // Parameter resolution deferred; we should see B
+        assert_param_with_default(&db, a, "a", "B", None);
+        assert_param_with_default(&db, b, "b", "T", None);
+    }
+
+    #[test]
+    fn external_signature_no_decorator() {
+        let mut db = setup_db();
+        db.write_dedented(
+            "/src/a.py",
+            "
+            def f(a: int) -> int: ...
+            ",
+        )
+        .unwrap();
+        let func = get_function_f(&db, "/src/a.py");
+
+        let expected_sig = func.internal_signature(&db);
+
+        // With no decorators, internal and external signature are the same
+        assert_eq!(func.signature(&db), &expected_sig);
+    }
+
+    #[test]
+    fn external_signature_decorated() {
+        let mut db = setup_db();
+        db.write_dedented(
+            "/src/a.py",
+            "
+            def deco(func): ...
+
+            @deco
+            def f(a: int) -> int: ...
+            ",
+        )
+        .unwrap();
+        let func = get_function_f(&db, "/src/a.py");
+
+        let expected_sig = Signature::todo();
+
+        // With no decorators, internal and external signature are the same
+        assert_eq!(func.signature(&db), &expected_sig);
+    }
+}

crates/red_knot_python_semantic/src/types/string_annotation.rs

@@ -0,0 +1,77 @@
+use ruff_db::files::File;
+use ruff_db::source::source_text;
+use ruff_python_ast::str::raw_contents;
+use ruff_python_ast::{self as ast, ModExpression, StringFlags};
+use ruff_python_parser::{parse_expression_range, Parsed};
+use ruff_text_size::Ranged;
+
+use crate::types::diagnostic::{TypeCheckDiagnostics, TypeCheckDiagnosticsBuilder};
+use crate::Db;
+
+type AnnotationParseResult = Result<Parsed<ModExpression>, TypeCheckDiagnostics>;
+
+/// Parses the given expression as a string annotation.
+pub(crate) fn parse_string_annotation(
+    db: &dyn Db,
+    file: File,
+    string_expr: &ast::ExprStringLiteral,
+) -> AnnotationParseResult {
+    let _span = tracing::trace_span!("parse_string_annotation", string=?string_expr.range(), file=%file.path(db)).entered();
+
+    let source = source_text(db.upcast(), file);
+    let node_text = &source[string_expr.range()];
+    let mut diagnostics = TypeCheckDiagnosticsBuilder::new(db, file);
+
+    if let [string_literal] = string_expr.value.as_slice() {
+        let prefix = string_literal.flags.prefix();
+        if prefix.is_raw() {
+            diagnostics.add(
+                string_literal.into(),
+                "annotation-raw-string",
+                format_args!("Type expressions cannot use raw string literal"),
+            );
+        // Compare the raw contents (without quotes) of the expression with the parsed contents
+        // contained in the string literal.
+        } else if raw_contents(node_text)
+            .is_some_and(|raw_contents| raw_contents == string_literal.as_str())
+        {
+            let range_excluding_quotes = string_literal
+                .range()
+                .add_start(string_literal.flags.opener_len())
+                .sub_end(string_literal.flags.closer_len());
+
+            // TODO: Support multiline strings like:
+            // ```py
+            // x: """
+            //     int
+            //     | float
+            // """ = 1
+            // ```
+            match parse_expression_range(source.as_str(), range_excluding_quotes) {
+                Ok(parsed) => return Ok(parsed),
+                Err(parse_error) => diagnostics.add(
+                    string_literal.into(),
+                    "forward-annotation-syntax-error",
+                    format_args!("Syntax error in forward annotation: {}", parse_error.error),
+                ),
+            }
+        } else {
+            // The raw contents of the string doesn't match the parsed content. This could be the
+            // case for annotations that contain escape sequences.
+            diagnostics.add(
+                string_expr.into(),
+                "annotation-escape-character",
+                format_args!("Type expressions cannot contain escape characters"),
+            );
+        }
+    } else {
+        // String is implicitly concatenated.
+        diagnostics.add(
+            string_expr.into(),
+            "annotation-implicit-concat",
+            format_args!("Type expressions cannot span multiple string literals"),
+        );
+    }
+
+    Err(diagnostics.finish())
+}

crates/red_knot_python_semantic/src/types/unpacker.rs

@@ -4,9 +4,9 @@ use ruff_db::files::File;
 use ruff_python_ast::{self as ast, AnyNodeRef};
 use rustc_hash::FxHashMap;
 
-use crate::semantic_index::ast_ids::{HasScopedAstId, ScopedExpressionId};
+use crate::semantic_index::ast_ids::{HasScopedExpressionId, ScopedExpressionId};
 use crate::semantic_index::symbol::ScopeId;
-use crate::types::{TupleType, Type, TypeCheckDiagnostics, TypeCheckDiagnosticsBuilder};
+use crate::types::{Type, TypeCheckDiagnostics, TypeCheckDiagnosticsBuilder};
 use crate::Db;
 
 /// Unpacks the value expression type to their respective targets.
@@ -29,7 +29,7 @@ impl<'db> Unpacker<'db> {
         match target {
             ast::Expr::Name(target_name) => {
                 self.targets
-                    .insert(target_name.scoped_ast_id(self.db, scope), value_ty);
+                    .insert(target_name.scoped_expression_id(self.db, scope), value_ty);
             }
             ast::Expr::Starred(ast::ExprStarred { value, .. }) => {
                 self.unpack(value, value_ty, scope);
@@ -93,11 +93,10 @@ impl<'db> Unpacker<'db> {
                     // further and deconstruct to an array of `StringLiteral` with each
                     // individual character, instead of just an array of `LiteralString`, but
                     // there would be a cost and it's not clear that it's worth it.
-                    let value_ty = Type::Tuple(TupleType::new(
+                    let value_ty = Type::tuple(
                         self.db,
-                        vec![Type::LiteralString; string_literal_ty.len(self.db)]
-                            .into_boxed_slice(),
-                    ));
+                        &vec![Type::LiteralString; string_literal_ty.len(self.db)],
+                    );
                     self.unpack(target, value_ty, scope);
                 }
                 _ => {

crates/red_knot_python_semantic/tests/mdtest.rs

@@ -1,7 +1,8 @@
-use dir_test::{dir_test, Fixture};
-use red_knot_test::run;
+use std::ffi::OsStr;
 use std::path::Path;
 
+use dir_test::{dir_test, Fixture};
+
 /// See `crates/red_knot_test/README.md` for documentation on these tests.
 #[dir_test(
     dir: "$CARGO_MANIFEST_DIR/resources/mdtest",
@@ -9,16 +10,16 @@ use std::path::Path;
 )]
 #[allow(clippy::needless_pass_by_value)]
 fn mdtest(fixture: Fixture<&str>) {
-    let path = fixture.path();
+    let fixture_path = Path::new(fixture.path());
+    let crate_dir = Path::new(env!("CARGO_MANIFEST_DIR"));
+    let workspace_root = crate_dir.parent().and_then(Path::parent).unwrap();
 
-    let crate_dir = Path::new(env!("CARGO_MANIFEST_DIR"))
-        .join("resources/mdtest")
-        .canonicalize()
+    let long_title = fixture_path
+        .strip_prefix(workspace_root)
+        .unwrap()
+        .to_str()
         .unwrap();
+    let short_title = fixture_path.file_name().and_then(OsStr::to_str).unwrap();
 
-    let relative_path = path
-        .strip_prefix(crate_dir.to_str().unwrap())
-        .unwrap_or(path);
-
-    run(Path::new(path), relative_path);
+    red_knot_test::run(fixture_path, long_title, short_title);
 }

crates/red_knot_server/src/session.rs

@@ -68,7 +68,7 @@ impl Session {
             let system = LSPSystem::new(index.clone());
 
             // TODO(dhruvmanila): Get the values from the client settings
-            let metadata = WorkspaceMetadata::from_path(system_path, &system, None)?;
+            let metadata = WorkspaceMetadata::discover(system_path, &system, None)?;
             // TODO(micha): Handle the case where the program settings are incorrect more gracefully.
             workspaces.insert(path, RootDatabase::new(metadata, system)?);
         }

crates/red_knot_server/src/system.rs

@@ -7,8 +7,8 @@ use lsp_types::Url;
 use ruff_db::file_revision::FileRevision;
 use ruff_db::system::walk_directory::WalkDirectoryBuilder;
 use ruff_db::system::{
-    DirectoryEntry, FileType, Metadata, OsSystem, Result, System, SystemPath, SystemPathBuf,
-    SystemVirtualPath, SystemVirtualPathBuf,
+    DirectoryEntry, FileType, GlobError, Metadata, OsSystem, PatternError, Result, System,
+    SystemPath, SystemPathBuf, SystemVirtualPath, SystemVirtualPathBuf,
 };
 use ruff_notebook::{Notebook, NotebookError};
 
@@ -198,6 +198,16 @@ impl System for LSPSystem {
         self.os_system.walk_directory(path)
     }
 
+    fn glob(
+        &self,
+        pattern: &str,
+    ) -> std::result::Result<
+        Box<dyn Iterator<Item = std::result::Result<SystemPathBuf, GlobError>>>,
+        PatternError,
+    > {
+        self.os_system.glob(pattern)
+    }
+
     fn as_any(&self) -> &dyn Any {
         self
     }

crates/red_knot_test/src/lib.rs

@@ -19,9 +19,9 @@ mod parser;
 ///
 /// Panic on test failure, and print failure details.
 #[allow(clippy::print_stdout)]
-pub fn run(path: &Path, title: &str) {
+pub fn run(path: &Path, long_title: &str, short_title: &str) {
     let source = std::fs::read_to_string(path).unwrap();
-    let suite = match test_parser::parse(title, &source) {
+    let suite = match test_parser::parse(short_title, &source) {
         Ok(suite) => suite,
         Err(err) => {
             panic!("Error parsing `{}`: {err}", path.to_str().unwrap())
@@ -49,8 +49,8 @@ pub fn run(path: &Path, title: &str) {
                     for failure in failures {
                         let absolute_line_number =
                             backtick_line.checked_add(relative_line_number).unwrap();
-                        let line_info = format!("{title}:{absolute_line_number}").cyan();
-                        println!("    {line_info} {failure}");
+                        let line_info = format!("{long_title}:{absolute_line_number}").cyan();
+                        println!("  {line_info} {failure}");
                     }
                 }
             }

crates/red_knot_vendored/vendor/typeshed/source_commit.txt

@@ -1 +1 @@
-d262beb07502cda412db2179fb406d45d1a9486f
+5052fa2f18db4493892e0f2775030683c9d06531

crates/red_knot_vendored/vendor/typeshed/stdlib/VERSIONS

@@ -24,18 +24,22 @@ _asyncio: 3.0-
 _bisect: 3.0-
 _blake2: 3.6-
 _bootlocale: 3.4-3.9
+_bz2: 3.3-
 _codecs: 3.0-
 _collections_abc: 3.3-
 _compat_pickle: 3.1-
 _compression: 3.5-
+_contextvars: 3.7-
 _csv: 3.0-
 _ctypes: 3.0-
 _curses: 3.0-
+_dbm: 3.0-
 _decimal: 3.3-
 _dummy_thread: 3.0-3.8
 _dummy_threading: 3.0-3.8
 _frozen_importlib: 3.0-
 _frozen_importlib_external: 3.5-
+_gdbm: 3.0-
 _heapq: 3.0-
 _imp: 3.0-
 _interpchannels: 3.13-
@@ -45,6 +49,7 @@ _io: 3.0-
 _json: 3.0-
 _locale: 3.0-
 _lsprof: 3.0-
+_lzma: 3.3-
 _markupbase: 3.0-
 _msi: 3.0-3.12
 _operator: 3.4-
@@ -52,12 +57,14 @@ _osx_support: 3.0-
 _posixsubprocess: 3.2-
 _py_abc: 3.7-
 _pydecimal: 3.5-
+_queue: 3.7-
 _random: 3.0-
 _sitebuiltins: 3.4-
 _socket: 3.0-  # present in 3.0 at runtime, but not in typeshed
 _sqlite3: 3.0-
 _ssl: 3.0-
 _stat: 3.4-
+_struct: 3.0-
 _thread: 3.0-
 _threading_local: 3.0-
 _tkinter: 3.0-

crates/red_knot_vendored/vendor/typeshed/stdlib/_bz2.pyi

@@ -0,0 +1,18 @@
+from _typeshed import ReadableBuffer
+from typing import final
+
+@final
+class BZ2Compressor:
+    def __init__(self, compresslevel: int = 9) -> None: ...
+    def compress(self, data: ReadableBuffer, /) -> bytes: ...
+    def flush(self) -> bytes: ...
+
+@final
+class BZ2Decompressor:
+    def decompress(self, data: ReadableBuffer, max_length: int = -1) -> bytes: ...
+    @property
+    def eof(self) -> bool: ...
+    @property
+    def needs_input(self) -> bool: ...
+    @property
+    def unused_data(self) -> bytes: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/_contextvars.pyi

@@ -0,0 +1,61 @@
+import sys
+from collections.abc import Callable, Iterator, Mapping
+from typing import Any, ClassVar, Generic, TypeVar, final, overload
+from typing_extensions import ParamSpec
+
+if sys.version_info >= (3, 9):
+    from types import GenericAlias
+
+_T = TypeVar("_T")
+_D = TypeVar("_D")
+_P = ParamSpec("_P")
+
+@final
+class ContextVar(Generic[_T]):
+    @overload
+    def __init__(self, name: str) -> None: ...
+    @overload
+    def __init__(self, name: str, *, default: _T) -> None: ...
+    def __hash__(self) -> int: ...
+    @property
+    def name(self) -> str: ...
+    @overload
+    def get(self) -> _T: ...
+    @overload
+    def get(self, default: _T, /) -> _T: ...
+    @overload
+    def get(self, default: _D, /) -> _D | _T: ...
+    def set(self, value: _T, /) -> Token[_T]: ...
+    def reset(self, token: Token[_T], /) -> None: ...
+    if sys.version_info >= (3, 9):
+        def __class_getitem__(cls, item: Any, /) -> GenericAlias: ...
+
+@final
+class Token(Generic[_T]):
+    @property
+    def var(self) -> ContextVar[_T]: ...
+    @property
+    def old_value(self) -> Any: ...  # returns either _T or MISSING, but that's hard to express
+    MISSING: ClassVar[object]
+    if sys.version_info >= (3, 9):
+        def __class_getitem__(cls, item: Any, /) -> GenericAlias: ...
+
+def copy_context() -> Context: ...
+
+# It doesn't make sense to make this generic, because for most Contexts each ContextVar will have
+# a different value.
+@final
+class Context(Mapping[ContextVar[Any], Any]):
+    def __init__(self) -> None: ...
+    @overload
+    def get(self, key: ContextVar[_T], default: None = None, /) -> _T | None: ...
+    @overload
+    def get(self, key: ContextVar[_T], default: _T, /) -> _T: ...
+    @overload
+    def get(self, key: ContextVar[_T], default: _D, /) -> _T | _D: ...
+    def run(self, callable: Callable[_P, _T], *args: _P.args, **kwargs: _P.kwargs) -> _T: ...
+    def copy(self) -> Context: ...
+    def __getitem__(self, key: ContextVar[_T], /) -> _T: ...
+    def __iter__(self) -> Iterator[ContextVar[Any]]: ...
+    def __len__(self) -> int: ...
+    def __eq__(self, value: object, /) -> bool: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/_dbm.pyi

@@ -0,0 +1,43 @@
+import sys
+from _typeshed import ReadOnlyBuffer, StrOrBytesPath
+from types import TracebackType
+from typing import TypeVar, overload
+from typing_extensions import Self, TypeAlias
+
+if sys.platform != "win32":
+    _T = TypeVar("_T")
+    _KeyType: TypeAlias = str | ReadOnlyBuffer
+    _ValueType: TypeAlias = str | ReadOnlyBuffer
+
+    class error(OSError): ...
+    library: str
+
+    # Actual typename dbm, not exposed by the implementation
+    class _dbm:
+        def close(self) -> None: ...
+        if sys.version_info >= (3, 13):
+            def clear(self) -> None: ...
+
+        def __getitem__(self, item: _KeyType) -> bytes: ...
+        def __setitem__(self, key: _KeyType, value: _ValueType) -> None: ...
+        def __delitem__(self, key: _KeyType) -> None: ...
+        def __len__(self) -> int: ...
+        def __del__(self) -> None: ...
+        def __enter__(self) -> Self: ...
+        def __exit__(
+            self, exc_type: type[BaseException] | None, exc_val: BaseException | None, exc_tb: TracebackType | None
+        ) -> None: ...
+        @overload
+        def get(self, k: _KeyType) -> bytes | None: ...
+        @overload
+        def get(self, k: _KeyType, default: _T) -> bytes | _T: ...
+        def keys(self) -> list[bytes]: ...
+        def setdefault(self, k: _KeyType, default: _ValueType = ...) -> bytes: ...
+        # Don't exist at runtime
+        __new__: None  # type: ignore[assignment]
+        __init__: None  # type: ignore[assignment]
+
+    if sys.version_info >= (3, 11):
+        def open(filename: StrOrBytesPath, flags: str = "r", mode: int = 0o666, /) -> _dbm: ...
+    else:
+        def open(filename: str, flags: str = "r", mode: int = 0o666, /) -> _dbm: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/_frozen_importlib_external.pyi

@@ -107,9 +107,9 @@ class FileLoader:
     def get_filename(self, name: str | None = None) -> str: ...
     def load_module(self, name: str | None = None) -> types.ModuleType: ...
     if sys.version_info >= (3, 10):
-        def get_resource_reader(self, module: types.ModuleType) -> importlib.readers.FileReader: ...
+        def get_resource_reader(self, name: str | None = None) -> importlib.readers.FileReader: ...
     else:
-        def get_resource_reader(self, module: types.ModuleType) -> Self | None: ...
+        def get_resource_reader(self, name: str | None = None) -> Self | None: ...
         def open_resource(self, resource: str) -> _io.FileIO: ...
         def resource_path(self, resource: str) -> str: ...
         def is_resource(self, name: str) -> bool: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/_gdbm.pyi

@@ -0,0 +1,47 @@
+import sys
+from _typeshed import ReadOnlyBuffer, StrOrBytesPath
+from types import TracebackType
+from typing import TypeVar, overload
+from typing_extensions import Self, TypeAlias
+
+if sys.platform != "win32":
+    _T = TypeVar("_T")
+    _KeyType: TypeAlias = str | ReadOnlyBuffer
+    _ValueType: TypeAlias = str | ReadOnlyBuffer
+
+    open_flags: str
+
+    class error(OSError): ...
+    # Actual typename gdbm, not exposed by the implementation
+    class _gdbm:
+        def firstkey(self) -> bytes | None: ...
+        def nextkey(self, key: _KeyType) -> bytes | None: ...
+        def reorganize(self) -> None: ...
+        def sync(self) -> None: ...
+        def close(self) -> None: ...
+        if sys.version_info >= (3, 13):
+            def clear(self) -> None: ...
+
+        def __getitem__(self, item: _KeyType) -> bytes: ...
+        def __setitem__(self, key: _KeyType, value: _ValueType) -> None: ...
+        def __delitem__(self, key: _KeyType) -> None: ...
+        def __contains__(self, key: _KeyType) -> bool: ...
+        def __len__(self) -> int: ...
+        def __enter__(self) -> Self: ...
+        def __exit__(
+            self, exc_type: type[BaseException] | None, exc_val: BaseException | None, exc_tb: TracebackType | None
+        ) -> None: ...
+        @overload
+        def get(self, k: _KeyType) -> bytes | None: ...
+        @overload
+        def get(self, k: _KeyType, default: _T) -> bytes | _T: ...
+        def keys(self) -> list[bytes]: ...
+        def setdefault(self, k: _KeyType, default: _ValueType = ...) -> bytes: ...
+        # Don't exist at runtime
+        __new__: None  # type: ignore[assignment]
+        __init__: None  # type: ignore[assignment]
+
+    if sys.version_info >= (3, 11):
+        def open(filename: StrOrBytesPath, flags: str = "r", mode: int = 0o666, /) -> _gdbm: ...
+    else:
+        def open(filename: str, flags: str = "r", mode: int = 0o666, /) -> _gdbm: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/_io.pyi

@@ -33,10 +33,10 @@ class _IOBase:
     def readable(self) -> bool: ...
     read: Callable[..., Any]
     def readlines(self, hint: int = -1, /) -> list[bytes]: ...
-    def seek(self, offset: int, whence: int = ..., /) -> int: ...
+    def seek(self, offset: int, whence: int = 0, /) -> int: ...
     def seekable(self) -> bool: ...
     def tell(self) -> int: ...
-    def truncate(self, size: int | None = ..., /) -> int: ...
+    def truncate(self, size: int | None = None, /) -> int: ...
     def writable(self) -> bool: ...
     write: Callable[..., Any]
     def writelines(self, lines: Iterable[ReadableBuffer], /) -> None: ...
@@ -59,8 +59,8 @@ class _BufferedIOBase(_IOBase):
     def readinto(self, buffer: WriteableBuffer, /) -> int: ...
     def write(self, buffer: ReadableBuffer, /) -> int: ...
     def readinto1(self, buffer: WriteableBuffer, /) -> int: ...
-    def read(self, size: int | None = ..., /) -> bytes: ...
-    def read1(self, size: int = ..., /) -> bytes: ...
+    def read(self, size: int | None = -1, /) -> bytes: ...
+    def read1(self, size: int = -1, /) -> bytes: ...
 
 class FileIO(RawIOBase, _RawIOBase, BinaryIO):  # type: ignore[misc]  # incompatible definitions of writelines in the base classes
     mode: str
@@ -69,13 +69,15 @@ class FileIO(RawIOBase, _RawIOBase, BinaryIO):  # type: ignore[misc]  # incompat
     # "name" is a str. In the future, making FileIO generic might help.
     name: Any
     def __init__(
-        self, file: FileDescriptorOrPath, mode: str = ..., closefd: bool = ..., opener: _Opener | None = ...
+        self, file: FileDescriptorOrPath, mode: str = "r", closefd: bool = True, opener: _Opener | None = None
     ) -> None: ...
     @property
     def closefd(self) -> bool: ...
+    def seek(self, pos: int, whence: int = 0, /) -> int: ...
+    def read(self, size: int | None = -1, /) -> bytes | MaybeNone: ...
 
 class BytesIO(BufferedIOBase, _BufferedIOBase, BinaryIO):  # type: ignore[misc]  # incompatible definitions of methods in the base classes
-    def __init__(self, initial_bytes: ReadableBuffer = ...) -> None: ...
+    def __init__(self, initial_bytes: ReadableBuffer = b"") -> None: ...
     # BytesIO does not contain a "name" field. This workaround is necessary
     # to allow BytesIO sub-classes to add this field, as it is defined
     # as a read-only property on IO[].
@@ -83,16 +85,22 @@ class BytesIO(BufferedIOBase, _BufferedIOBase, BinaryIO):  # type: ignore[misc]
     def getvalue(self) -> bytes: ...
     def getbuffer(self) -> memoryview: ...
     def read1(self, size: int | None = -1, /) -> bytes: ...
+    def readlines(self, size: int | None = None, /) -> list[bytes]: ...
+    def seek(self, pos: int, whence: int = 0, /) -> int: ...
 
 class BufferedReader(BufferedIOBase, _BufferedIOBase, BinaryIO):  # type: ignore[misc]  # incompatible definitions of methods in the base classes
     raw: RawIOBase
     def __init__(self, raw: RawIOBase, buffer_size: int = 8192) -> None: ...
     def peek(self, size: int = 0, /) -> bytes: ...
+    def seek(self, target: int, whence: int = 0, /) -> int: ...
+    def truncate(self, pos: int | None = None, /) -> int: ...
 
 class BufferedWriter(BufferedIOBase, _BufferedIOBase, BinaryIO):  # type: ignore[misc]  # incompatible definitions of writelines in the base classes
     raw: RawIOBase
     def __init__(self, raw: RawIOBase, buffer_size: int = 8192) -> None: ...
     def write(self, buffer: ReadableBuffer, /) -> int: ...
+    def seek(self, target: int, whence: int = 0, /) -> int: ...
+    def truncate(self, pos: int | None = None, /) -> int: ...
 
 class BufferedRandom(BufferedIOBase, _BufferedIOBase, BinaryIO):  # type: ignore[misc]  # incompatible definitions of methods in the base classes
     mode: str
@@ -101,10 +109,11 @@ class BufferedRandom(BufferedIOBase, _BufferedIOBase, BinaryIO):  # type: ignore
     def __init__(self, raw: RawIOBase, buffer_size: int = 8192) -> None: ...
     def seek(self, target: int, whence: int = 0, /) -> int: ...  # stubtest needs this
     def peek(self, size: int = 0, /) -> bytes: ...
+    def truncate(self, pos: int | None = None, /) -> int: ...
 
 class BufferedRWPair(BufferedIOBase, _BufferedIOBase):
     def __init__(self, reader: RawIOBase, writer: RawIOBase, buffer_size: int = 8192) -> None: ...
-    def peek(self, size: int = ..., /) -> bytes: ...
+    def peek(self, size: int = 0, /) -> bytes: ...
 
 class _TextIOBase(_IOBase):
     encoding: str
@@ -115,9 +124,9 @@ class _TextIOBase(_IOBase):
     def detach(self) -> BinaryIO: ...
     def write(self, s: str, /) -> int: ...
     def writelines(self, lines: Iterable[str], /) -> None: ...  # type: ignore[override]
-    def readline(self, size: int = ..., /) -> str: ...  # type: ignore[override]
+    def readline(self, size: int = -1, /) -> str: ...  # type: ignore[override]
     def readlines(self, hint: int = -1, /) -> list[str]: ...  # type: ignore[override]
-    def read(self, size: int | None = ..., /) -> str: ...
+    def read(self, size: int | None = -1, /) -> str: ...
 
 @type_check_only
 class _WrappedBuffer(Protocol):
@@ -177,9 +186,10 @@ class TextIOWrapper(TextIOBase, _TextIOBase, TextIO, Generic[_BufferT_co]):  # t
     # TextIOWrapper's version of seek only supports a limited subset of
     # operations.
     def seek(self, cookie: int, whence: int = 0, /) -> int: ...
+    def truncate(self, pos: int | None = None, /) -> int: ...
 
 class StringIO(TextIOBase, _TextIOBase, TextIO):  # type: ignore[misc]  # incompatible definitions of write in the base classes
-    def __init__(self, initial_value: str | None = ..., newline: str | None = ...) -> None: ...
+    def __init__(self, initial_value: str | None = "", newline: str | None = "\n") -> None: ...
     # StringIO does not contain a "name" field. This workaround is necessary
     # to allow StringIO sub-classes to add this field, as it is defined
     # as a read-only property on IO[].
@@ -187,9 +197,11 @@ class StringIO(TextIOBase, _TextIOBase, TextIO):  # type: ignore[misc]  # incomp
     def getvalue(self) -> str: ...
     @property
     def line_buffering(self) -> bool: ...
+    def seek(self, pos: int, whence: int = 0, /) -> int: ...
+    def truncate(self, pos: int | None = None, /) -> int: ...
 
 class IncrementalNewlineDecoder:
-    def __init__(self, decoder: codecs.IncrementalDecoder | None, translate: bool, errors: str = ...) -> None: ...
+    def __init__(self, decoder: codecs.IncrementalDecoder | None, translate: bool, errors: str = "strict") -> None: ...
     def decode(self, input: ReadableBuffer | str, final: bool = False) -> str: ...
     @property
     def newlines(self) -> str | tuple[str, ...] | None: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/_lzma.pyi

@@ -0,0 +1,60 @@
+from _typeshed import ReadableBuffer
+from collections.abc import Mapping, Sequence
+from typing import Any, Final, final
+from typing_extensions import TypeAlias
+
+_FilterChain: TypeAlias = Sequence[Mapping[str, Any]]
+
+FORMAT_AUTO: Final = 0
+FORMAT_XZ: Final = 1
+FORMAT_ALONE: Final = 2
+FORMAT_RAW: Final = 3
+CHECK_NONE: Final = 0
+CHECK_CRC32: Final = 1
+CHECK_CRC64: Final = 4
+CHECK_SHA256: Final = 10
+CHECK_ID_MAX: Final = 15
+CHECK_UNKNOWN: Final = 16
+FILTER_LZMA1: int  # v big number
+FILTER_LZMA2: Final = 33
+FILTER_DELTA: Final = 3
+FILTER_X86: Final = 4
+FILTER_IA64: Final = 6
+FILTER_ARM: Final = 7
+FILTER_ARMTHUMB: Final = 8
+FILTER_SPARC: Final = 9
+FILTER_POWERPC: Final = 5
+MF_HC3: Final = 3
+MF_HC4: Final = 4
+MF_BT2: Final = 18
+MF_BT3: Final = 19
+MF_BT4: Final = 20
+MODE_FAST: Final = 1
+MODE_NORMAL: Final = 2
+PRESET_DEFAULT: Final = 6
+PRESET_EXTREME: int  # v big number
+
+@final
+class LZMADecompressor:
+    def __init__(self, format: int | None = ..., memlimit: int | None = ..., filters: _FilterChain | None = ...) -> None: ...
+    def decompress(self, data: ReadableBuffer, max_length: int = -1) -> bytes: ...
+    @property
+    def check(self) -> int: ...
+    @property
+    def eof(self) -> bool: ...
+    @property
+    def unused_data(self) -> bytes: ...
+    @property
+    def needs_input(self) -> bool: ...
+
+@final
+class LZMACompressor:
+    def __init__(
+        self, format: int | None = ..., check: int = ..., preset: int | None = ..., filters: _FilterChain | None = ...
+    ) -> None: ...
+    def compress(self, data: ReadableBuffer, /) -> bytes: ...
+    def flush(self) -> bytes: ...
+
+class LZMAError(Exception): ...
+
+def is_check_supported(check_id: int, /) -> bool: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/_queue.pyi

@@ -0,0 +1,20 @@
+import sys
+from typing import Any, Generic, TypeVar
+
+if sys.version_info >= (3, 9):
+    from types import GenericAlias
+
+_T = TypeVar("_T")
+
+class Empty(Exception): ...
+
+class SimpleQueue(Generic[_T]):
+    def __init__(self) -> None: ...
+    def empty(self) -> bool: ...
+    def get(self, block: bool = True, timeout: float | None = None) -> _T: ...
+    def get_nowait(self) -> _T: ...
+    def put(self, item: _T, block: bool = True, timeout: float | None = None) -> None: ...
+    def put_nowait(self, item: _T) -> None: ...
+    def qsize(self) -> int: ...
+    if sys.version_info >= (3, 9):
+        def __class_getitem__(cls, item: Any, /) -> GenericAlias: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/_struct.pyi

@@ -0,0 +1,22 @@
+from _typeshed import ReadableBuffer, WriteableBuffer
+from collections.abc import Iterator
+from typing import Any
+
+def pack(fmt: str | bytes, /, *v: Any) -> bytes: ...
+def pack_into(fmt: str | bytes, buffer: WriteableBuffer, offset: int, /, *v: Any) -> None: ...
+def unpack(format: str | bytes, buffer: ReadableBuffer, /) -> tuple[Any, ...]: ...
+def unpack_from(format: str | bytes, /, buffer: ReadableBuffer, offset: int = 0) -> tuple[Any, ...]: ...
+def iter_unpack(format: str | bytes, buffer: ReadableBuffer, /) -> Iterator[tuple[Any, ...]]: ...
+def calcsize(format: str | bytes, /) -> int: ...
+
+class Struct:
+    @property
+    def format(self) -> str: ...
+    @property
+    def size(self) -> int: ...
+    def __init__(self, format: str | bytes) -> None: ...
+    def pack(self, *v: Any) -> bytes: ...
+    def pack_into(self, buffer: WriteableBuffer, offset: int, *v: Any) -> None: ...
+    def unpack(self, buffer: ReadableBuffer, /) -> tuple[Any, ...]: ...
+    def unpack_from(self, buffer: ReadableBuffer, offset: int = 0) -> tuple[Any, ...]: ...
+    def iter_unpack(self, buffer: ReadableBuffer, /) -> Iterator[tuple[Any, ...]]: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/_threading_local.pyi

@@ -1,5 +1,5 @@
 from typing import Any
-from typing_extensions import TypeAlias
+from typing_extensions import Self, TypeAlias
 from weakref import ReferenceType
 
 __all__ = ["local"]
@@ -12,6 +12,7 @@ class _localimpl:
     def create_dict(self) -> _LocalDict: ...
 
 class local:
+    def __new__(cls, /, *args: Any, **kw: Any) -> Self: ...
     def __getattribute__(self, name: str) -> Any: ...
     def __setattr__(self, name: str, value: Any) -> None: ...
     def __delattr__(self, name: str) -> None: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/builtins.pyi

@@ -1284,9 +1284,7 @@ class property:
 
 @final
 class _NotImplementedType(Any):
-    # A little weird, but typing the __call__ as NotImplemented makes the error message
-    # for NotImplemented() much better
-    __call__: NotImplemented  # type: ignore[valid-type]  # pyright: ignore[reportInvalidTypeForm]
+    __call__: None
 
 NotImplemented: _NotImplementedType
 
@@ -1917,7 +1915,7 @@ class StopIteration(Exception):
 
 class OSError(Exception):
     errno: int | None
-    strerror: str
+    strerror: str | None
     # filename, filename2 are actually str | bytes | None
     filename: Any
     filename2: Any

crates/red_knot_vendored/vendor/typeshed/stdlib/bz2.pyi

@@ -1,9 +1,10 @@
 import _compression
 import sys
+from _bz2 import BZ2Compressor as BZ2Compressor, BZ2Decompressor as BZ2Decompressor
 from _compression import BaseStream
 from _typeshed import ReadableBuffer, StrOrBytesPath, WriteableBuffer
 from collections.abc import Iterable
-from typing import IO, Any, Literal, Protocol, SupportsIndex, TextIO, final, overload
+from typing import IO, Any, Literal, Protocol, SupportsIndex, TextIO, overload
 from typing_extensions import Self, TypeAlias
 
 __all__ = ["BZ2File", "BZ2Compressor", "BZ2Decompressor", "open", "compress", "decompress"]
@@ -128,19 +129,3 @@ class BZ2File(BaseStream, IO[bytes]):
     def seek(self, offset: int, whence: int = 0) -> int: ...
     def write(self, data: ReadableBuffer) -> int: ...
     def writelines(self, seq: Iterable[ReadableBuffer]) -> None: ...
-
-@final
-class BZ2Compressor:
-    def __init__(self, compresslevel: int = 9) -> None: ...
-    def compress(self, data: ReadableBuffer, /) -> bytes: ...
-    def flush(self) -> bytes: ...
-
-@final
-class BZ2Decompressor:
-    def decompress(self, data: ReadableBuffer, max_length: int = -1) -> bytes: ...
-    @property
-    def eof(self) -> bool: ...
-    @property
-    def needs_input(self) -> bool: ...
-    @property
-    def unused_data(self) -> bytes: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/contextvars.pyi

@@ -1,63 +1,3 @@
-import sys
-from collections.abc import Callable, Iterator, Mapping
-from typing import Any, ClassVar, Generic, TypeVar, final, overload
-from typing_extensions import ParamSpec
-
-if sys.version_info >= (3, 9):
-    from types import GenericAlias
+from _contextvars import Context as Context, ContextVar as ContextVar, Token as Token, copy_context as copy_context
 
 __all__ = ("Context", "ContextVar", "Token", "copy_context")
-
-_T = TypeVar("_T")
-_D = TypeVar("_D")
-_P = ParamSpec("_P")
-
-@final
-class ContextVar(Generic[_T]):
-    @overload
-    def __init__(self, name: str) -> None: ...
-    @overload
-    def __init__(self, name: str, *, default: _T) -> None: ...
-    def __hash__(self) -> int: ...
-    @property
-    def name(self) -> str: ...
-    @overload
-    def get(self) -> _T: ...
-    @overload
-    def get(self, default: _T, /) -> _T: ...
-    @overload
-    def get(self, default: _D, /) -> _D | _T: ...
-    def set(self, value: _T, /) -> Token[_T]: ...
-    def reset(self, token: Token[_T], /) -> None: ...
-    if sys.version_info >= (3, 9):
-        def __class_getitem__(cls, item: Any, /) -> GenericAlias: ...
-
-@final
-class Token(Generic[_T]):
-    @property
-    def var(self) -> ContextVar[_T]: ...
-    @property
-    def old_value(self) -> Any: ...  # returns either _T or MISSING, but that's hard to express
-    MISSING: ClassVar[object]
-    if sys.version_info >= (3, 9):
-        def __class_getitem__(cls, item: Any, /) -> GenericAlias: ...
-
-def copy_context() -> Context: ...
-
-# It doesn't make sense to make this generic, because for most Contexts each ContextVar will have
-# a different value.
-@final
-class Context(Mapping[ContextVar[Any], Any]):
-    def __init__(self) -> None: ...
-    @overload
-    def get(self, key: ContextVar[_T], default: None = None, /) -> _T | None: ...
-    @overload
-    def get(self, key: ContextVar[_T], default: _T, /) -> _T: ...
-    @overload
-    def get(self, key: ContextVar[_T], default: _D, /) -> _T | _D: ...
-    def run(self, callable: Callable[_P, _T], *args: _P.args, **kwargs: _P.kwargs) -> _T: ...
-    def copy(self) -> Context: ...
-    def __getitem__(self, key: ContextVar[_T], /) -> _T: ...
-    def __iter__(self) -> Iterator[ContextVar[Any]]: ...
-    def __len__(self) -> int: ...
-    def __eq__(self, value: object, /) -> bool: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/dbm/gnu.pyi

@@ -1,47 +1 @@
-import sys
-from _typeshed import ReadOnlyBuffer, StrOrBytesPath
-from types import TracebackType
-from typing import TypeVar, overload
-from typing_extensions import Self, TypeAlias
-
-if sys.platform != "win32":
-    _T = TypeVar("_T")
-    _KeyType: TypeAlias = str | ReadOnlyBuffer
-    _ValueType: TypeAlias = str | ReadOnlyBuffer
-
-    open_flags: str
-
-    class error(OSError): ...
-    # Actual typename gdbm, not exposed by the implementation
-    class _gdbm:
-        def firstkey(self) -> bytes | None: ...
-        def nextkey(self, key: _KeyType) -> bytes | None: ...
-        def reorganize(self) -> None: ...
-        def sync(self) -> None: ...
-        def close(self) -> None: ...
-        if sys.version_info >= (3, 13):
-            def clear(self) -> None: ...
-
-        def __getitem__(self, item: _KeyType) -> bytes: ...
-        def __setitem__(self, key: _KeyType, value: _ValueType) -> None: ...
-        def __delitem__(self, key: _KeyType) -> None: ...
-        def __contains__(self, key: _KeyType) -> bool: ...
-        def __len__(self) -> int: ...
-        def __enter__(self) -> Self: ...
-        def __exit__(
-            self, exc_type: type[BaseException] | None, exc_val: BaseException | None, exc_tb: TracebackType | None
-        ) -> None: ...
-        @overload
-        def get(self, k: _KeyType) -> bytes | None: ...
-        @overload
-        def get(self, k: _KeyType, default: _T) -> bytes | _T: ...
-        def keys(self) -> list[bytes]: ...
-        def setdefault(self, k: _KeyType, default: _ValueType = ...) -> bytes: ...
-        # Don't exist at runtime
-        __new__: None  # type: ignore[assignment]
-        __init__: None  # type: ignore[assignment]
-
-    if sys.version_info >= (3, 11):
-        def open(filename: StrOrBytesPath, flags: str = "r", mode: int = 0o666, /) -> _gdbm: ...
-    else:
-        def open(filename: str, flags: str = "r", mode: int = 0o666, /) -> _gdbm: ...
+from _gdbm import *

crates/red_knot_vendored/vendor/typeshed/stdlib/dbm/ndbm.pyi

@@ -1,43 +1 @@
-import sys
-from _typeshed import ReadOnlyBuffer, StrOrBytesPath
-from types import TracebackType
-from typing import TypeVar, overload
-from typing_extensions import Self, TypeAlias
-
-if sys.platform != "win32":
-    _T = TypeVar("_T")
-    _KeyType: TypeAlias = str | ReadOnlyBuffer
-    _ValueType: TypeAlias = str | ReadOnlyBuffer
-
-    class error(OSError): ...
-    library: str
-
-    # Actual typename dbm, not exposed by the implementation
-    class _dbm:
-        def close(self) -> None: ...
-        if sys.version_info >= (3, 13):
-            def clear(self) -> None: ...
-
-        def __getitem__(self, item: _KeyType) -> bytes: ...
-        def __setitem__(self, key: _KeyType, value: _ValueType) -> None: ...
-        def __delitem__(self, key: _KeyType) -> None: ...
-        def __len__(self) -> int: ...
-        def __del__(self) -> None: ...
-        def __enter__(self) -> Self: ...
-        def __exit__(
-            self, exc_type: type[BaseException] | None, exc_val: BaseException | None, exc_tb: TracebackType | None
-        ) -> None: ...
-        @overload
-        def get(self, k: _KeyType) -> bytes | None: ...
-        @overload
-        def get(self, k: _KeyType, default: _T) -> bytes | _T: ...
-        def keys(self) -> list[bytes]: ...
-        def setdefault(self, k: _KeyType, default: _ValueType = ...) -> bytes: ...
-        # Don't exist at runtime
-        __new__: None  # type: ignore[assignment]
-        __init__: None  # type: ignore[assignment]
-
-    if sys.version_info >= (3, 11):
-        def open(filename: StrOrBytesPath, flags: str = "r", mode: int = 0o666, /) -> _dbm: ...
-    else:
-        def open(filename: str, flags: str = "r", mode: int = 0o666, /) -> _dbm: ...
+from _dbm import *

crates/red_knot_vendored/vendor/typeshed/stdlib/distutils/ccompiler.pyi

@@ -1,5 +1,5 @@
 from _typeshed import BytesPath, StrPath, Unused
-from collections.abc import Callable, Iterable
+from collections.abc import Callable, Iterable, Sequence
 from distutils.file_util import _BytesPathT, _StrPathT
 from typing import Literal, overload
 from typing_extensions import TypeAlias, TypeVarTuple, Unpack
@@ -63,7 +63,7 @@ class CCompiler:
     def set_executables(self, **args: str) -> None: ...
     def compile(
         self,
-        sources: list[str],
+        sources: Sequence[StrPath],
         output_dir: str | None = None,
         macros: list[_Macro] | None = None,
         include_dirs: list[str] | None = None,

crates/red_knot_vendored/vendor/typeshed/stdlib/fractions.pyi

@@ -2,7 +2,7 @@ import sys
 from collections.abc import Callable
 from decimal import Decimal
 from numbers import Integral, Rational, Real
-from typing import Any, Literal, SupportsIndex, overload
+from typing import Any, Literal, Protocol, SupportsIndex, overload
 from typing_extensions import Self, TypeAlias
 
 _ComparableNum: TypeAlias = int | float | Decimal | Real
@@ -20,11 +20,19 @@ else:
     @overload
     def gcd(a: Integral, b: Integral) -> Integral: ...
 
+class _ConvertibleToIntegerRatio(Protocol):
+    def as_integer_ratio(self) -> tuple[int | Rational, int | Rational]: ...
+
 class Fraction(Rational):
     @overload
     def __new__(cls, numerator: int | Rational = 0, denominator: int | Rational | None = None) -> Self: ...
     @overload
     def __new__(cls, value: float | Decimal | str, /) -> Self: ...
+
+    if sys.version_info >= (3, 14):
+        @overload
+        def __new__(cls, value: _ConvertibleToIntegerRatio) -> Self: ...
+
     @classmethod
     def from_float(cls, f: float) -> Self: ...
     @classmethod

crates/red_knot_vendored/vendor/typeshed/stdlib/getopt.pyi

@@ -1,7 +1,11 @@
+from collections.abc import Iterable
+
 __all__ = ["GetoptError", "error", "getopt", "gnu_getopt"]
 
-def getopt(args: list[str], shortopts: str, longopts: list[str] = []) -> tuple[list[tuple[str, str]], list[str]]: ...
-def gnu_getopt(args: list[str], shortopts: str, longopts: list[str] = []) -> tuple[list[tuple[str, str]], list[str]]: ...
+def getopt(args: list[str], shortopts: str, longopts: Iterable[str] | str = []) -> tuple[list[tuple[str, str]], list[str]]: ...
+def gnu_getopt(
+    args: list[str], shortopts: str, longopts: Iterable[str] | str = []
+) -> tuple[list[tuple[str, str]], list[str]]: ...
 
 class GetoptError(Exception):
     msg: str

crates/red_knot_vendored/vendor/typeshed/stdlib/ipaddress.pyi

@@ -128,19 +128,6 @@ class _BaseNetwork(_IPAddressBase, Generic[_A]):
     @property
     def hostmask(self) -> _A: ...
 
-class _BaseInterface(_BaseAddress, Generic[_A, _N]):
-    hostmask: _A
-    netmask: _A
-    network: _N
-    @property
-    def ip(self) -> _A: ...
-    @property
-    def with_hostmask(self) -> str: ...
-    @property
-    def with_netmask(self) -> str: ...
-    @property
-    def with_prefixlen(self) -> str: ...
-
 class _BaseV4:
     @property
     def version(self) -> Literal[4]: ...
@@ -154,9 +141,21 @@ class IPv4Address(_BaseV4, _BaseAddress):
 
 class IPv4Network(_BaseV4, _BaseNetwork[IPv4Address]): ...
 
-class IPv4Interface(IPv4Address, _BaseInterface[IPv4Address, IPv4Network]):
+class IPv4Interface(IPv4Address):
+    netmask: IPv4Address
+    network: IPv4Network
     def __eq__(self, other: object) -> bool: ...
     def __hash__(self) -> int: ...
+    @property
+    def hostmask(self) -> IPv4Address: ...
+    @property
+    def ip(self) -> IPv4Address: ...
+    @property
+    def with_hostmask(self) -> str: ...
+    @property
+    def with_netmask(self) -> str: ...
+    @property
+    def with_prefixlen(self) -> str: ...
 
 class _BaseV6:
     @property
@@ -184,9 +183,21 @@ class IPv6Network(_BaseV6, _BaseNetwork[IPv6Address]):
     @property
     def is_site_local(self) -> bool: ...
 
-class IPv6Interface(IPv6Address, _BaseInterface[IPv6Address, IPv6Network]):
+class IPv6Interface(IPv6Address):
+    netmask: IPv6Address
+    network: IPv6Network
     def __eq__(self, other: object) -> bool: ...
     def __hash__(self) -> int: ...
+    @property
+    def hostmask(self) -> IPv6Address: ...
+    @property
+    def ip(self) -> IPv6Address: ...
+    @property
+    def with_hostmask(self) -> str: ...
+    @property
+    def with_netmask(self) -> str: ...
+    @property
+    def with_prefixlen(self) -> str: ...
 
 def v4_int_to_packed(address: int) -> bytes: ...
 def v6_int_to_packed(address: int) -> bytes: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/logging/config.pyi

@@ -1,6 +1,6 @@
 import sys
 from _typeshed import StrOrBytesPath
-from collections.abc import Callable, Hashable, Iterable, Sequence
+from collections.abc import Callable, Hashable, Iterable, Mapping, Sequence
 from configparser import RawConfigParser
 from re import Pattern
 from threading import Thread
@@ -63,7 +63,7 @@ def dictConfig(config: _DictConfigArgs | dict[str, Any]) -> None: ...
 if sys.version_info >= (3, 10):
     def fileConfig(
         fname: StrOrBytesPath | IO[str] | RawConfigParser,
-        defaults: dict[str, str] | None = None,
+        defaults: Mapping[str, str] | None = None,
         disable_existing_loggers: bool = True,
         encoding: str | None = None,
     ) -> None: ...
@@ -71,7 +71,7 @@ if sys.version_info >= (3, 10):
 else:
     def fileConfig(
         fname: StrOrBytesPath | IO[str] | RawConfigParser,
-        defaults: dict[str, str] | None = None,
+        defaults: Mapping[str, str] | None = None,
         disable_existing_loggers: bool = True,
     ) -> None: ...
 

crates/red_knot_vendored/vendor/typeshed/stdlib/logging/handlers.pyi

@@ -260,6 +260,8 @@ class QueueHandler(Handler):
     def __init__(self, queue: _QueueLike[Any]) -> None: ...
     def prepare(self, record: LogRecord) -> Any: ...
     def enqueue(self, record: LogRecord) -> None: ...
+    if sys.version_info >= (3, 12):
+        listener: QueueListener | None
 
 class QueueListener:
     handlers: tuple[Handler, ...]  # undocumented

crates/red_knot_vendored/vendor/typeshed/stdlib/lzma.pyi

@@ -1,7 +1,41 @@
 from _compression import BaseStream
+from _lzma import (
+    CHECK_CRC32 as CHECK_CRC32,
+    CHECK_CRC64 as CHECK_CRC64,
+    CHECK_ID_MAX as CHECK_ID_MAX,
+    CHECK_NONE as CHECK_NONE,
+    CHECK_SHA256 as CHECK_SHA256,
+    CHECK_UNKNOWN as CHECK_UNKNOWN,
+    FILTER_ARM as FILTER_ARM,
+    FILTER_ARMTHUMB as FILTER_ARMTHUMB,
+    FILTER_DELTA as FILTER_DELTA,
+    FILTER_IA64 as FILTER_IA64,
+    FILTER_LZMA1 as FILTER_LZMA1,
+    FILTER_LZMA2 as FILTER_LZMA2,
+    FILTER_POWERPC as FILTER_POWERPC,
+    FILTER_SPARC as FILTER_SPARC,
+    FILTER_X86 as FILTER_X86,
+    FORMAT_ALONE as FORMAT_ALONE,
+    FORMAT_AUTO as FORMAT_AUTO,
+    FORMAT_RAW as FORMAT_RAW,
+    FORMAT_XZ as FORMAT_XZ,
+    MF_BT2 as MF_BT2,
+    MF_BT3 as MF_BT3,
+    MF_BT4 as MF_BT4,
+    MF_HC3 as MF_HC3,
+    MF_HC4 as MF_HC4,
+    MODE_FAST as MODE_FAST,
+    MODE_NORMAL as MODE_NORMAL,
+    PRESET_DEFAULT as PRESET_DEFAULT,
+    PRESET_EXTREME as PRESET_EXTREME,
+    LZMACompressor as LZMACompressor,
+    LZMADecompressor as LZMADecompressor,
+    LZMAError as LZMAError,
+    _FilterChain,
+    is_check_supported as is_check_supported,
+)
 from _typeshed import ReadableBuffer, StrOrBytesPath
-from collections.abc import Mapping, Sequence
-from typing import IO, Any, Final, Literal, TextIO, final, overload
+from typing import IO, Literal, TextIO, overload
 from typing_extensions import Self, TypeAlias
 
 __all__ = [
@@ -48,62 +82,6 @@ _OpenTextWritingMode: TypeAlias = Literal["wt", "xt", "at"]
 
 _PathOrFile: TypeAlias = StrOrBytesPath | IO[bytes]
 
-_FilterChain: TypeAlias = Sequence[Mapping[str, Any]]
-
-FORMAT_AUTO: Final = 0
-FORMAT_XZ: Final = 1
-FORMAT_ALONE: Final = 2
-FORMAT_RAW: Final = 3
-CHECK_NONE: Final = 0
-CHECK_CRC32: Final = 1
-CHECK_CRC64: Final = 4
-CHECK_SHA256: Final = 10
-CHECK_ID_MAX: Final = 15
-CHECK_UNKNOWN: Final = 16
-FILTER_LZMA1: int  # v big number
-FILTER_LZMA2: Final = 33
-FILTER_DELTA: Final = 3
-FILTER_X86: Final = 4
-FILTER_IA64: Final = 6
-FILTER_ARM: Final = 7
-FILTER_ARMTHUMB: Final = 8
-FILTER_SPARC: Final = 9
-FILTER_POWERPC: Final = 5
-MF_HC3: Final = 3
-MF_HC4: Final = 4
-MF_BT2: Final = 18
-MF_BT3: Final = 19
-MF_BT4: Final = 20
-MODE_FAST: Final = 1
-MODE_NORMAL: Final = 2
-PRESET_DEFAULT: Final = 6
-PRESET_EXTREME: int  # v big number
-
-# from _lzma.c
-@final
-class LZMADecompressor:
-    def __init__(self, format: int | None = ..., memlimit: int | None = ..., filters: _FilterChain | None = ...) -> None: ...
-    def decompress(self, data: ReadableBuffer, max_length: int = -1) -> bytes: ...
-    @property
-    def check(self) -> int: ...
-    @property
-    def eof(self) -> bool: ...
-    @property
-    def unused_data(self) -> bytes: ...
-    @property
-    def needs_input(self) -> bool: ...
-
-# from _lzma.c
-@final
-class LZMACompressor:
-    def __init__(
-        self, format: int | None = ..., check: int = ..., preset: int | None = ..., filters: _FilterChain | None = ...
-    ) -> None: ...
-    def compress(self, data: ReadableBuffer, /) -> bytes: ...
-    def flush(self) -> bytes: ...
-
-class LZMAError(Exception): ...
-
 class LZMAFile(BaseStream, IO[bytes]):  # type: ignore[misc]  # incompatible definitions of writelines in the base classes
     def __init__(
         self,
@@ -194,4 +172,3 @@ def compress(
 def decompress(
     data: ReadableBuffer, format: int = 0, memlimit: int | None = None, filters: _FilterChain | None = None
 ) -> bytes: ...
-def is_check_supported(check_id: int, /) -> bool: ...

crates/red_knot_vendored/vendor/typeshed/stdlib/mmap.pyi

@@ -34,9 +34,22 @@ class mmap:
     if sys.platform == "win32":
         def __init__(self, fileno: int, length: int, tagname: str | None = ..., access: int = ..., offset: int = ...) -> None: ...
     else:
-        def __init__(
-            self, fileno: int, length: int, flags: int = ..., prot: int = ..., access: int = ..., offset: int = ...
-        ) -> None: ...
+        if sys.version_info >= (3, 13):
+            def __init__(
+                self,
+                fileno: int,
+                length: int,
+                flags: int = ...,
+                prot: int = ...,
+                access: int = ...,
+                offset: int = ...,
+                *,
+                trackfd: bool = True,
+            ) -> None: ...
+        else:
+            def __init__(
+                self, fileno: int, length: int, flags: int = ..., prot: int = ..., access: int = ..., offset: int = ...
+            ) -> None: ...
 
     def close(self) -> None: ...
     def flush(self, offset: int = ..., size: int = ...) -> None: ...