Add find_matching_modules and find_matching_direct_imports

Author: Peter554

CHANGELOG.rst

@@ -5,6 +5,7 @@ Changelog
 Unreleased
 ----------
 
+* Added `find_matching_modules` and `find_matching_direct_imports` methods.
 * Added `as_packages` option to the `find_shortest_chain` method.
 
 3.6 (2025-02-07)

docs/usage.rst

@@ -121,6 +121,16 @@ Methods for analysing the module tree
     :raises: ``ValueError`` if the module is a squashed module, as by definition it represents both itself and all
       of its descendants.
 
+.. py:function:: ImportGraph.find_matching_modules(expression)
+
+    Find all modules matching the passed expression (see :ref:`module_expressions`).
+
+    :param str expression: A module expression used for matching.
+    :return: A set of module names matching the expression.
+    :rtype: A set of strings.
+    :raises: ``grimp.exceptions.InvalidModuleExpression`` if the module expression is invalid.
+
+
 Methods for analysing direct imports
 ------------------------------------
 
@@ -184,6 +194,28 @@ Methods for analysing direct imports
         So if a module is imported twice from the same module, it will only be counted once.
     :rtype: Integer.
 
+.. py:function:: ImportGraph.find_matching_direct_imports(import_expression)
+
+    Find all direct imports matching the passed import expression.
+
+    The imports are returned are in the following form::
+
+        [
+            {
+                'importer': 'mypackage.importer',
+                'imported': 'mypackage.imported',
+            },
+            # (additional imports here)
+        ]
+
+    :param str import_expression: An expression in the form ``"importer_expression -> imported_expression"``,
+        where each expression is a module expression (see :ref:`module_expressions`).
+        Example: ``"mypackage.*.blue -> mypackage.*.green"``.
+    :return: An ordered list of direct imports matching the expressions (ordered alphabetically).
+    :rtype: List of dictionaries with the structure shown above. If you want to use type annotations, you may use the
+        ``grimp.Import`` TypedDict for each dictionary.
+    :raises: ``grimp.exceptions.InvalidImportExpression`` if the expression is not well-formed.
+
 Methods for analysing import chains
 -----------------------------------
 
@@ -517,5 +549,25 @@ Methods for manipulating the graph
     :param str module: The name of a module, for example ``'mypackage.foo'``.
     :return: bool
 
+.. _module_expressions:
+
+Module expressions
+------------------
+
+  A module expression is used to refer to sets of modules.
+
+  - ``*`` stands in for a module name, without including subpackages.
+  - ``**`` includes subpackages too.
+
+  Examples:
+
+  - ``mypackage.foo``:  matches ``mypackage.foo`` exactly.
+  - ``mypackage.*``:  matches ``mypackage.foo`` but not ``mypackage.foo.bar``.
+  - ``mypackage.*.baz``: matches ``mypackage.foo.baz`` but not ``mypackage.foo.bar.baz``.
+  - ``mypackage.*.*``: matches ``mypackage.foo.bar`` and ``mypackage.foobar.baz``.
+  - ``mypackage.**``: matches ``mypackage.foo.bar`` and ``mypackage.foo.bar.baz``.
+  - ``mypackage.**.qux``: matches ``mypackage.foo.bar.qux`` and ``mypackage.foo.bar.baz.qux``.
+  - ``mypackage.foo*``: is not a valid expression. (The wildcard must replace a whole module name.)
+
 .. _namespace packages: https://docs.python.org/3/glossary.html#term-namespace-package
 .. _namespace portion: https://docs.python.org/3/glossary.html#term-portion

rust/Cargo.lock

@@ -20,6 +20,8 @@ itertools = "0.14.0"
 tap = "1.0.1"
 rustc-hash = "2.1.0"
 indexmap = "2.7.1"
+regex = "1.11.1"
+const_format = "0.2.34"
 
 [dependencies.pyo3]
 version = "0.23.4"
@@ -29,4 +31,5 @@ extension-module = ["pyo3/extension-module"]
 default = ["extension-module"]
 
 [dev-dependencies]
+parameterized = "2.0.0"
 serde_json = "1.0.137"

rust/Cargo.toml

@@ -1,4 +1,4 @@
-use crate::exceptions::{ModuleNotPresent, NoSuchContainer};
+use crate::exceptions::{InvalidModuleExpression, ModuleNotPresent, NoSuchContainer};
 use pyo3::exceptions::PyValueError;
 use pyo3::PyErr;
 use thiserror::Error;
@@ -13,6 +13,9 @@ pub enum GrimpError {
 
     #[error("Modules have shared descendants.")]
     SharedDescendants,
+
+    #[error("{0} is not a valid module expression.")]
+    InvalidModuleExpression(String),
 }
 
 pub type GrimpResult<T> = Result<T, GrimpError>;
@@ -24,6 +27,9 @@ impl From<GrimpError> for PyErr {
             GrimpError::ModuleNotPresent(_) => ModuleNotPresent::new_err(value.to_string()),
             GrimpError::NoSuchContainer(_) => NoSuchContainer::new_err(value.to_string()),
             GrimpError::SharedDescendants => PyValueError::new_err(value.to_string()),
+            GrimpError::InvalidModuleExpression(_) => {
+                InvalidModuleExpression::new_err(value.to_string())
+            }
         }
     }
 }

rust/src/errors.rs

@@ -2,3 +2,8 @@ use pyo3::create_exception;
 
 create_exception!(_rustgrimp, ModuleNotPresent, pyo3::exceptions::PyException);
 create_exception!(_rustgrimp, NoSuchContainer, pyo3::exceptions::PyException);
+create_exception!(
+    _rustgrimp,
+    InvalidModuleExpression,
+    pyo3::exceptions::PyException
+);