add check-yaml-sorted script and tests

.pre-commit-hooks.yaml

@@ -95,6 +95,12 @@
     entry: check-yaml
     language: python
     types: [yaml]
+-   id: check-yaml-sorted
+    name: check yaml sorted
+    description: checks top-level items of yaml files are sorted.
+    entry: check-yaml-sorted
+    language: python
+    types: [yaml]
 -   id: debug-statements
     name: debug statements (python)
     description: checks for debugger imports and py37+ `breakpoint()` calls in python source.

README.md

@@ -87,6 +87,11 @@ Attempts to load all yaml files to verify syntax.
     portability to other yaml implementations.
     Implies `--allow-multiple-documents`.
 
+#### `check-yaml-sorted`
+Checks that top-level items in yaml files are sorted. This is done on the
+stringified representation, which means it can handle lists of scalars, lists
+of dicts by first key or val, etc.
+
 #### `debug-statements`
 Check for debugger imports and py37+ `breakpoint()` calls in python source.
 

pre_commit_hooks/check_yaml_sorted.py

@@ -0,0 +1,51 @@
+"""Pre-commit hook to check that yaml files are sorted at the top-level.
+
+Does not modify files. Simply parses and compares the stringified values of
+each element of the top-level object.
+
+This allows it to handle all kinds of cases
+- lists of scalars
+- dicts by top-level key
+- lists of dicts by first key name
+- lists of dicts with same keys by first value...
+"""
+from __future__ import annotations
+
+import argparse
+from itertools import tee
+from typing import Any
+from typing import Iterable
+from typing import Sequence
+
+import yaml
+
+
+def is_sorted(iterable: Iterable[Any]) -> bool:
+    a_iter, b_iter = tee(str(e) for e in iterable)
+    next(b_iter, None)
+    for a, b in zip(a_iter, b_iter):
+        if a > b:
+            print(f'Items ({a[:32]}..., {b[:32]}...)')
+            return False
+    return True
+
+
+def main(argv: Sequence[str] | None = None) -> int:
+    parser = argparse.ArgumentParser()
+    parser.add_argument('filenames', nargs='*', help='Filenames to check')
+    args = parser.parse_args(argv)
+
+    retval = 0
+
+    for filename in args.filenames:
+        with open(filename) as fp:
+            data = yaml.safe_load(fp)
+        if not is_sorted(data):
+            print(f'In file {filename}, items are out of order.')
+            retval += 1
+
+    return retval
+
+
+if __name__ == '__main__':
+    raise SystemExit(main())

setup.cfg

@@ -45,6 +45,7 @@ console_scripts =
     check-vcs-permalinks = pre_commit_hooks.check_vcs_permalinks:main
     check-xml = pre_commit_hooks.check_xml:main
     check-yaml = pre_commit_hooks.check_yaml:main
+    check-yaml-sorted = pre_commit_hooks.check_yaml_sorted:main
     debug-statement-hook = pre_commit_hooks.debug_statement_hook:main
     destroyed-symlinks = pre_commit_hooks.destroyed_symlinks:main
     detect-aws-credentials = pre_commit_hooks.detect_aws_credentials:main

tests/check_yaml_sorted_test.py

@@ -0,0 +1,102 @@
+from __future__ import annotations
+
+import yaml
+
+from pre_commit_hooks.check_yaml_sorted import is_sorted
+
+
+def test_sort_list_by_items():
+    assert is_sorted(['a'])
+    assert is_sorted(['a', 'b'])
+    assert not is_sorted(['b', 'a'])
+    assert is_sorted(['a', 'b', 'c', 'd'])
+    assert not is_sorted(['a', 'b', 'd', 'c'])
+
+
+def test_sort_dicts_by_keys():
+    assert is_sorted({'a': 1, 'b': ['nested', 'list'], 'c': 3})
+    assert not is_sorted({'a': 1, 'c': ['nested', 'list'], 'b': 3})
+
+
+_list_dicts_first_key_sorted = """
+-   first_dict:
+        some: stuff
+-   second_dict:
+        other: stuff
+-   third_dict:
+        even: more stuff
+"""
+
+_list_dicts_first_key_unsorted = """
+-   second_dict:
+        some: stuff
+-   first_dict:
+        other: stuff
+-   third_dict:
+        even: more stuff
+"""
+
+
+def test_sort_list_of_dicts_by_first_key():
+    assert is_sorted(yaml.safe_load(_list_dicts_first_key_sorted))
+    assert not is_sorted(yaml.safe_load(_list_dicts_first_key_unsorted))
+
+
+_sorted_yaml = """
+-   id: check-builtin-literals
+-   id: check-case-conflict
+-   id: check-docstring-first
+"""
+
+_sorted_yaml_long = """
+-   id: check-builtin-literals
+    name: check builtin type constructor use
+    description: requires literal syntax when initializing empty or zero....
+    entry: check-builtin-literals
+    language: python
+    types: [python]
+-   id: check-case-conflict
+    name: check for case conflicts
+    description: checks for files that would conflict in case-insensitive...
+    entry: check-case-conflict
+    language: python
+-   id: check-docstring-first
+    name: check docstring is first
+    description: checks a common error of defining a docstring after code.
+    entry: check-docstring-first
+    language: python
+    types: [python]
+"""
+
+_unsorted_yaml = """
+-   id: check-builtin-literals
+-   id: check-docstring-first
+-   id: check-case-conflict
+"""
+
+_unsorted_yaml_long = """
+-   id: check-builtin-literals
+    name: check builtin type constructor use
+    description: requires literal syntax when initializing empty or zero....
+    entry: check-builtin-literals
+    language: python
+    types: [python]
+-   id: check-docstring-first
+    name: check docstring is first
+    description: checks a common error of defining a docstring after code.
+    entry: check-docstring-first
+    language: python
+    types: [python]
+-   id: check-case-conflict
+    name: check for case conflicts
+    description: checks for files that would conflict in case-insensitive...
+    entry: check-case-conflict
+    language: python
+"""
+
+
+def test_sort_list_of_dicts_same_first_key_by_val():
+    assert is_sorted(yaml.safe_load(_sorted_yaml))
+    assert is_sorted(yaml.safe_load(_sorted_yaml_long))
+    assert not is_sorted(yaml.safe_load(_unsorted_yaml))
+    assert not is_sorted(yaml.safe_load(_unsorted_yaml_long))