✨ feat(discovery): add iter_interpreters for enumeration

README.md

@@ -36,7 +36,8 @@ The `get_interpreter()` function accepts various specification formats:
 - Absolute path: `/usr/bin/python3.12`
 - Version: `3.12`
 - Implementation prefix: `cpython3.12`
-- PEP 440 specifier: `>=3.10`, `>=3.11,<3.13`
+- [Version specifier](https://packaging.python.org/en/latest/specifications/version-specifiers/): `>=3.10`,
+  `>=3.11,<3.13`
 
 ## Documentation
 

docs/changelog/65.bugfix.rst

@@ -0,0 +1,3 @@
+discover uv-managed Pythons on Windows. Previously the glob assumed Unix layout (``<root>/<key>/bin/python``) and
+silently found nothing on Windows, where uv places ``python.exe`` directly under the install root - by
+:user:`gaborbernat`.

docs/changelog/65.feature.rst

@@ -0,0 +1,3 @@
+add :func:`~python_discovery.iter_interpreters` for enumerating every discovered interpreter, with PATH and
+UV-install support for non-CPython implementations listed in :data:`~python_discovery.KNOWN_IMPLEMENTATIONS`
+- by :user:`gaborbernat`.

docs/explanation.rst

@@ -65,6 +65,106 @@ detects these shims and resolves them to the actual binary.
 `mise <https://mise.jdx.dev/>`_ and `asdf <https://asdf-vm.com/>`_ work similarly, using the
 ``MISE_DATA_DIR`` and ``ASDF_DATA_DIR`` environment variables to locate their installations.
 
+How uv-managed Pythons are discovered
+---------------------------------------
+
+`uv <https://docs.astral.sh/uv/>`_ installs Python interpreters under a single root directory (configurable via
+``UV_PYTHON_INSTALL_DIR``, otherwise defaulting under ``XDG_DATA_HOME`` or the platform user-data path). Each
+install lives in its own subdirectory, but the actual binary location varies by OS and implementation:
+
+.. list-table::
+   :header-rows: 1
+   :widths: 25 35 40
+
+   * - Implementation
+     - Unix layout
+     - Windows layout
+   * - CPython
+     - ``<root>/<key>/bin/python``
+     - ``<root>/<key>/python.exe``
+   * - PyPy
+     - ``<root>/<key>/bin/pypy*``
+     - ``<root>/<key>/pypy*.exe``
+   * - GraalPy
+     - ``<root>/<key>/bin/graalpy``
+     - ``<root>/<key>/bin/graalpy.exe``
+
+.. mermaid::
+
+    flowchart LR
+        Call(["iter_interpreters(key)"]) --> Mode{"key is None?"}
+        Mode -->|"narrow"| N1["*/bin/python"]
+        Mode -->|"narrow"| N2["*/python.exe"]
+        Mode -->|"wide"| W1["*/bin/pypy*"]
+        Mode -->|"wide"| W2["*/bin/graalpy"]
+        Mode -->|"wide"| W3["*/pypy*.exe"]
+        Mode -->|"wide"| W4["*/bin/graalpy.exe"]
+
+        N1 --> Dedup[/"realpath dedup"/]
+        N2 --> Dedup
+        W1 --> Dedup
+        W2 --> Dedup
+        W3 --> Dedup
+        W4 --> Dedup
+
+        Dedup --> Interrogate(["subprocess interrogation"])
+
+        style Call fill:#4a90d9,stroke:#2a5f8f,color:#fff
+        style Mode fill:#d9904a,stroke:#8f5f2a,color:#fff
+        style N1 fill:#3a7fc2,stroke:#1f4d7a,color:#fff
+        style N2 fill:#3a7fc2,stroke:#1f4d7a,color:#fff
+        style W1 fill:#9f4ad9,stroke:#5f2a8f,color:#fff
+        style W2 fill:#9f4ad9,stroke:#5f2a8f,color:#fff
+        style W3 fill:#9f4ad9,stroke:#5f2a8f,color:#fff
+        style W4 fill:#9f4ad9,stroke:#5f2a8f,color:#fff
+        style Dedup fill:#c2873a,stroke:#7a4c1f,color:#fff
+        style Interrogate fill:#4a9f4a,stroke:#2a6f2a,color:#fff
+
+GraalPy keeps its ``bin/`` segment on Windows (an upstream choice in uv); PyPy and CPython do not. python-discovery
+globs all of these patterns regardless of the host OS, because globs that do not match anything are essentially
+free, and the cross-platform list is short. Symlinked aliases inside an install (``bin/python``,
+``bin/python3``, ``bin/python3.14`` all pointing at the same real file) are deduplicated by resolved path before
+the subprocess interrogation, so each install is interrogated once.
+
+Selecting one interpreter vs. enumerating all of them
+-------------------------------------------------------
+
+:func:`~python_discovery.get_interpreter` and :func:`~python_discovery.iter_interpreters` walk the same candidate
+sources, but they answer different questions and behave differently in three ways.
+
+.. mermaid::
+
+    flowchart LR
+        Sources["candidate sources<br>(try_first_with → current →<br>PEP 514 → PATH → uv)"]
+        Sources --> Get["get_interpreter()<br>first match wins, returns one"]
+        Sources --> Iter["iter_interpreters()<br>yields every match"]
+
+        style Get fill:#4a9f4a,stroke:#2a6f2a,color:#fff
+        style Iter fill:#4a90d9,stroke:#2a5f8f,color:#fff
+
+**Implementation coverage on PATH.** :func:`~python_discovery.get_interpreter` matches only ``python*`` filenames on
+PATH unless the spec names another implementation explicitly (``pypy3.12``, ``graalpy3.11``). This keeps backwards
+compatibility with tools that have always read "no implementation in the spec" as "give me CPython."
+:func:`~python_discovery.iter_interpreters` with no spec broadens the search to every name in
+:data:`~python_discovery.KNOWN_IMPLEMENTATIONS` -- otherwise an "all interpreters" call would silently miss every
+PyPy and GraalPy on the system. When you pass a spec to :func:`~python_discovery.iter_interpreters`, it falls back
+to the same narrow regex as :func:`~python_discovery.get_interpreter`, so behavior is consistent across the two
+APIs whenever a spec is given.
+
+**Deduplication.** :func:`~python_discovery.get_interpreter` deduplicates per call so it does not interrogate the
+same binary twice while searching, and stops as soon as a match is found. :func:`~python_discovery.iter_interpreters`
+deduplicates by the resolved real path of each candidate's ``system_executable`` (falling back to ``executable``).
+That means symlinked aliases like ``/bin/python3`` and ``/usr/bin/python3``, or a virtualenv whose ``python``
+symlinks to its base interpreter, collapse to a single yield. The semantic is "one entry per distinct install,"
+which is what callers building choosers or version-range pickers usually want.
+
+**Iteration order.** Yields come back in *priority order*: ``try_first_with`` first, then the running interpreter,
+then :pep:`514` entries on Windows, then PATH left-to-right, then UV-managed installs. This matches what
+:func:`~python_discovery.get_interpreter` would have returned at each step. If your ordering differs (newest
+version first, smallest install root, etc.), wrap the call in :func:`sorted` -- the API deliberately does not
+include a ``sort_by`` parameter because keeping discovery order preserves the priority signal for callers who
+want it.
+
 How caching works
 -------------------
 
@@ -165,9 +265,12 @@ A spec string follows the pattern ``[impl][version][t][-arch][-machine]``. Every
    * - ``/usr/bin/python3``
      - Absolute path, used directly (no search)
    * - ``>=3.11,<3.13``
-     - :pep:`440` version specifier (any Python in range)
+     - `Version specifier <https://packaging.python.org/en/latest/specifications/version-specifiers/>`_
+       (any Python in range)
    * - ``cpython>=3.11``
-     - :pep:`440` specifier restricted to CPython
+     - `Version specifier <https://packaging.python.org/en/latest/specifications/version-specifiers/>`_
+       restricted to CPython
 
-:pep:`440` specifiers (``>=``, ``<=``, ``~=``, ``!=``, ``==``, ``===``) are supported. Multiple
-specifiers can be comma-separated, for example ``>=3.11,<3.13``.
+`Version specifiers <https://packaging.python.org/en/latest/specifications/version-specifiers/>`_
+(``>=``, ``<=``, ``~=``, ``!=``, ``==``, ``===``) are supported. Multiple specifiers can be comma-separated,
+for example ``>=3.11,<3.13``.

docs/how-to/standalone-usage.rst

@@ -62,6 +62,60 @@ the ``PY_DISCOVERY_TIMEOUT`` environment variable.
 The timeout value should be a number in seconds. Each interpreter candidate is given this much time
 to respond. If a timeout occurs, the candidate is skipped and the search continues with the next one.
 
+List every interpreter on the system
+--------------------------------------
+
+Use :func:`~python_discovery.iter_interpreters` to enumerate every Python python-discovery can find. With no spec
+it yields all known implementations (CPython, PyPy, GraalPy -- see
+:data:`~python_discovery.KNOWN_IMPLEMENTATIONS`). Pass a spec to filter, exactly like
+:func:`~python_discovery.get_interpreter`.
+
+.. code-block:: python
+
+   from pathlib import Path
+
+   from python_discovery import DiskCache, iter_interpreters
+
+   cache = DiskCache(root=Path("~/.cache/python-discovery").expanduser())
+
+   # Every interpreter, no filter
+   for info in iter_interpreters(cache=cache):
+       print(info.executable, info.version_str, info.implementation)
+
+   # Every CPython 3.10 or newer, newest first
+   newest_first = sorted(
+       iter_interpreters("cpython>=3.10", cache=cache),
+       key=lambda info: info.version_info,
+       reverse=True,
+   )
+
+Enumeration interrogates every candidate as a subprocess on a cold cache. Always pass a
+:class:`~python_discovery.DiskCache` if you call this more than once.
+
+Pick an interpreter from a range, preferring newer
+----------------------------------------------------
+
+A common need: search a version range and prefer newer interpreters when more than one matches. Sort the result of
+:func:`~python_discovery.iter_interpreters` and take the first.
+
+.. code-block:: python
+
+   from pathlib import Path
+
+   from python_discovery import DiskCache, iter_interpreters
+
+   cache = DiskCache(root=Path("~/.cache/python-discovery").expanduser())
+
+   matches = sorted(
+       iter_interpreters("cpython>=3.10,<3.15", cache=cache),
+       key=lambda info: info.version_info,
+       reverse=True,
+   )
+   info = matches[0] if matches else None
+
+Use :func:`get_interpreter` instead when you only need the first PATH-priority hit; use the sort-and-take pattern
+when *your* ordering differs from PATH order (newest version, smallest install size, preferred install root, etc.).
+
 Read interpreter metadata
 ---------------------------
 

docs/tutorial/getting-started.rst

@@ -87,6 +87,49 @@ You can pass multiple specs as a list -- the library tries each one in order and
 
    result = get_interpreter(["python3.12", "python3.11"], cache=cache)
 
+Listing every interpreter
+---------------------------
+
+When you need *every* interpreter rather than just the first match -- for example, to show the user a chooser, or
+to apply your own ranking -- use :func:`~python_discovery.iter_interpreters`. Pass no arguments to enumerate every
+implementation python-discovery knows about, or pass a spec to filter.
+
+.. mermaid::
+
+    flowchart TD
+        Call["iter_interpreters(spec, cache)"] --> Yield["yields PythonInfo"]
+        Yield --> A["1. try_first_with paths"]
+        Yield --> B["2. running interpreter"]
+        Yield --> C["3. PATH (left to right)"]
+        Yield --> D["4. uv-managed installs"]
+
+        style Call fill:#4a90d9,stroke:#2a5f8f,color:#fff
+        style Yield fill:#4a9f4a,stroke:#2a6f2a,color:#fff
+
+.. code-block:: python
+
+   from pathlib import Path
+
+   from python_discovery import DiskCache, iter_interpreters
+
+   cache = DiskCache(root=Path("~/.cache/python-discovery").expanduser())
+   for info in iter_interpreters(cache=cache):
+       print(info.executable, info.version_str, info.implementation)
+
+The result is an iterator, so :func:`list`, :func:`sorted`, generator expressions and early ``break`` all work as you
+would expect. Symlinked aliases (``/bin/python3`` and ``/usr/bin/python3``, or a virtualenv and the base it points at)
+collapse to a single entry, so you do not see the same install twice.
+
+To prefer newer interpreters in a range, sort the result by ``version_info`` after filtering:
+
+.. code-block:: python
+
+   newest_first = sorted(
+       iter_interpreters(">=3.10,<3.15", cache=cache),
+       key=lambda info: info.version_info,
+       reverse=True,
+   )
+
 Writing specs
 -------------
 
@@ -131,7 +174,8 @@ Common examples:
    * - ``/usr/bin/python3``
      - An absolute path, used directly without searching
    * - ``>=3.11,<3.13``
-     - Any Python in the 3.11--3.12 range (:pep:`440` syntax)
+     - Any Python in the 3.11--3.12 range
+       (`version specifier <https://packaging.python.org/en/latest/specifications/version-specifiers/>`_ syntax)
 
 See the :doc:`full spec reference </explanation>` for all options.
 

pyproject.toml

@@ -105,6 +105,7 @@ lint.per-file-ignores."tests/**/*.py" = [
   "PLC0415", # imports inside test functions (conditional on mocking)
   "PLC2701", # private imports needed to test internal APIs
   "PLR0913", # too many arguments (pytest fixtures)
+  "PLR0917", # too many positional arguments (pytest fixtures)
   "PLR2004", # Magic value used in comparison
   "S101",    # asserts allowed in tests
   "S404",    # subprocess import

src/python_discovery/__init__.py

@@ -5,15 +5,16 @@
 from importlib.metadata import version
 
 from ._cache import ContentStore, DiskCache, PyInfoCache
-from ._discovery import get_interpreter
+from ._discovery import get_interpreter, iter_interpreters
 from ._py_info import KNOWN_ARCHITECTURES, PythonInfo, normalize_isa
-from ._py_spec import PythonSpec
+from ._py_spec import KNOWN_IMPLEMENTATIONS, PythonSpec
 from ._specifier import SimpleSpecifier, SimpleSpecifierSet, SimpleVersion
 
 __version__ = version("python-discovery")
 
 __all__ = [
     "KNOWN_ARCHITECTURES",
+    "KNOWN_IMPLEMENTATIONS",
     "ContentStore",
     "DiskCache",
     "PyInfoCache",
@@ -24,5 +25,6 @@
     "SimpleVersion",
     "__version__",
     "get_interpreter",
+    "iter_interpreters",
     "normalize_isa",
 ]