Merge pull request #261 from 23andMe/feature/add-exclude

cblakkan · web-flow · commit c989fabfb281 · 2025-01-02T15:35:16.000-05:00
Add --exclude argument to CLI
diff --git a/README.md b/README.md
@@ -1,5 +1,10 @@
 Yamale (ya·ma·lē)
 =================
+[![Build Status](https://github.com/23andMe/Yamale/actions/workflows/run-tests.yml/badge.svg)](https://github.com/23andMe/Yamale/actions/workflows/run-tests.yml)
+[![PyPI](https://img.shields.io/pypi/v/yamale.svg)](https://pypi.python.org/pypi/yamale)
+[![downloads](https://static.pepy.tech/badge/yamale/month)](https://pepy.tech/project/yamale)
+[![versions](https://img.shields.io/pypi/pyversions/yamale.svg)](https://github.com/yamale/yamale)
+[![license](https://img.shields.io/github/license/23andMe/yamale.svg)](https://github.com/23andMe/Yamale/blob/master/LICENSE)
 
 | :warning: Ensure that your schema definitions come from internal or trusted sources. Yamale does not protect against intentionally malicious schemas. |
 |:------------|
@@ -11,8 +16,6 @@ A schema and validator for YAML.
 What's YAML? See the current spec [here](http://www.yaml.org/spec/1.2/spec.html) and an introduction
 to the syntax [here](https://github.com/Animosity/CraftIRC/wiki/Complete-idiot's-introduction-to-yaml).
 
-[![Build Status](https://github.com/23andMe/Yamale/actions/workflows/run-tests.yml/badge.svg)](https://github.com/23andMe/Yamale/actions/workflows/run-tests.yml)
-[![PyPI](https://img.shields.io/pypi/v/yamale.svg)](https://pypi.python.org/pypi/yamale)
 
 Requirements
 ------------
@@ -23,8 +26,10 @@ Requirements
 Install
 -------
 ### pip
-```bash
+```
 $ pip install yamale
+# or to include ruamel.yaml as a dependency
+$ pip install yamale[ruamel]
 ```
 
 NOTE: Some platforms, e.g., Mac OS, may ship with only Python 2 and may not have pip installed.
@@ -49,25 +54,28 @@ looking up the directory tree until it finds one. If Yamale can not find a schem
 
 Usage:
 
-```bash
-usage: yamale [-h] [-s SCHEMA] [-n CPU_NUM] [-p PARSER] [--no-strict] [PATH]
+```
+usage: yamale [-h] [-s SCHEMA] [-e PATTERN] [-p PARSER] [-n CPU_NUM] [-x] [-v] [-V] [PATH ...]
 
 Validate yaml files.
 
 positional arguments:
-  PATH                  folder to validate. Default is current directory.
+  PATH                  Paths to validate, either directories or files. Default is the current directory.
 
-optional arguments:
+options:
   -h, --help            show this help message and exit
   -s SCHEMA, --schema SCHEMA
                         filename of schema. Default is schema.yaml.
-  -n CPU_NUM, --cpu-num CPU_NUM
-                        number of CPUs to use. Default is 4.
+  -e PATTERN, --exclude PATTERN
+                        Python regex used to exclude files from validation. Any substring match of a file's absolute path will be excluded. Uses
+                        default Python3 regex. Option can be supplied multiple times.
   -p PARSER, --parser PARSER
-                        YAML library to load files. Choices are "ruamel" or
-                        "pyyaml" (default).
-  --no-strict           Disable strict mode, unexpected elements in the data
-                        will be accepted.
+                        YAML library to load files. Choices are "ruamel" or "pyyaml" (default).
+  -n CPU_NUM, --cpu-num CPU_NUM
+                        Number of child processes to spawn for validation. Default is 4. 'auto' to use CPU count.
+  -x, --no-strict       Disable strict mode, unexpected elements in the data will be accepted.
+  -v, --verbose         show verbose information
+  -V, --version         show program's version number and exit
 ```
 
 ### API
diff --git a/setup.py b/setup.py
@@ -18,7 +18,7 @@
     packages=find_packages(),
     include_package_data=True,
     install_requires=["pyyaml"],
-    extras_requires={"ruamel": ["ruamel.yaml"]},
+    extras_require={"ruamel": ["ruamel.yaml"]},
     python_requires=">=3.8",
     entry_points={
         "console_scripts": ["yamale=yamale.command_line:main"],
diff --git a/yamale/VERSION b/yamale/VERSION
@@ -1 +1 @@
-5.3.0
+6.0.0
diff --git a/yamale/command_line.py b/yamale/command_line.py
@@ -11,7 +11,8 @@
 import argparse
 import glob
 import os
-from multiprocessing import Pool
+import re
+import multiprocessing
 from .yamale_error import YamaleError
 from .schema.validationresults import Result
 from .version import __version__
@@ -64,31 +65,32 @@ def _find_schema(data_path, schema_name):
     return _find_data_path_schema(data_path, schema_name)
 
 
-def _validate_single(yaml_path, schema_name, parser, strict):
-    print("Validating %s..." % yaml_path)
+def _validate_file(yaml_path, schema_name, parser, strict, should_exclude):
+    if should_exclude(yaml_path):
+        return
     s = _find_schema(yaml_path, schema_name)
     if not s:
         raise ValueError("Invalid schema name for '{}' or schema not found.".format(schema_name))
     _validate(s, yaml_path, parser, strict, True)
 
 
-def _validate_dir(root, schema_name, cpus, parser, strict):
-    pool = Pool(processes=cpus)
+def _validate_dir(root, schema_name, cpus, parser, strict, should_exclude):
+    pool = multiprocessing.Pool(processes=cpus)
     res = []
     error_messages = []
-    print("Finding yaml files...")
-    for root, dirs, files in os.walk(root):
+    for root, _, files in os.walk(root):
         for f in files:
             if (f.endswith(".yaml") or f.endswith(".yml")) and f != schema_name:
-                d = os.path.join(root, f)
-                s = _find_schema(d, schema_name)
-                if s:
-                    res.append(pool.apply_async(_validate, (s, d, parser, strict, False)))
+                yaml_path = os.path.join(root, f)
+                if should_exclude(yaml_path):
+                    continue
+                schema_path = _find_schema(yaml_path, schema_name)
+                if schema_path:
+                    res.append(pool.apply_async(_validate, (schema_path, yaml_path, parser, strict, False)))
                 else:
-                    print("No schema found for: %s" % d)
+                    print(f"No schema found for: {yaml_path}")
 
-    print("Found %s yaml files." % len(res))
-    print("Validating...")
+    print(f"Found {len(res)} yaml files to validate...")
     for r in res:
         sub_results = r.get(timeout=300)
         error_messages.extend([str(sub_result) for sub_result in sub_results if not sub_result.isValid()])
@@ -98,16 +100,34 @@ def _validate_dir(root, schema_name, cpus, parser, strict):
         raise ValueError("\n----\n".join(set(error_messages)))
 
 
-def _router(paths, schema_name, cpus, parser, strict=True):
+def _router(paths, schema_name, cpus, parser, excludes=None, strict=True, verbose=False):
+    EXCLUDE_REGEXES = tuple(re.compile(e) for e in excludes) if excludes else tuple()
+
+    def should_exclude(yaml_path):
+        has_match = any(pattern.search(yaml_path) for pattern in EXCLUDE_REGEXES)
+        if has_match and verbose:
+            print("Skipping validation for %s due to exclude pattern" % yaml_path)
+        return has_match
+
     for path in paths:
-        path = os.path.abspath(path)
-        if os.path.isdir(path):
-            _validate_dir(path, schema_name, cpus, parser, strict)
+        abs_path = os.path.abspath(path)
+        if os.path.exists(abs_path):
+            print(f"Validating {path}...")
+        else:
+            raise ValueError(f"Path does not exist: {path}")
+
+        if os.path.isdir(abs_path):
+            _validate_dir(abs_path, schema_name, cpus, parser, strict, should_exclude)
         else:
-            _validate_single(path, schema_name, parser, strict)
+            _validate_file(abs_path, schema_name, parser, strict, should_exclude)
 
 
 def main():
+    def int_or_auto(num_cpu):
+        if num_cpu == "auto":
+            return multiprocessing.cpu_count()
+        return int(num_cpu)
+
     parser = argparse.ArgumentParser(description="Validate yaml files.")
     parser.add_argument(
         "paths",
@@ -116,21 +136,46 @@ def main():
         nargs="*",
         help="Paths to validate, either directories or files. Default is the current directory.",
     )
-    parser.add_argument("-V", "--version", action="version", version=__version__)
     parser.add_argument("-s", "--schema", default="schema.yaml", help="filename of schema. Default is schema.yaml.")
-    parser.add_argument("-n", "--cpu-num", default=4, type=int, help="number of CPUs to use. Default is 4.")
+    parser.add_argument(
+        "-e",
+        "--exclude",
+        metavar="PATTERN",
+        action="append",
+        help="Python regex used to exclude files from validation. Any substring match of a file's absolute path will be excluded. Uses default Python3 regex. Option can be supplied multiple times.",
+    )
     parser.add_argument(
         "-p",
         "--parser",
         default="pyyaml",
         help='YAML library to load files. Choices are "ruamel" or "pyyaml" (default).',
     )
     parser.add_argument(
-        "--no-strict", action="store_true", help="Disable strict mode, unexpected elements in the data will be accepted."
+        "-n",
+        "--cpu-num",
+        default=4,
+        type=int_or_auto,
+        help="Number of child processes to spawn for validation. Default is 4. 'auto' to use CPU count.",
     )
+    parser.add_argument(
+        "-x",
+        "--no-strict",
+        action="store_true",
+        help="Disable strict mode, unexpected elements in the data will be accepted.",
+    )
+    parser.add_argument("-v", "--verbose", action="store_true", help="show verbose information")
+    parser.add_argument("-V", "--version", action="version", version=__version__)
     args = parser.parse_args()
     try:
-        _router(args.paths, args.schema, args.cpu_num, args.parser, not args.no_strict)
+        _router(
+            paths=args.paths,
+            schema_name=args.schema,
+            cpus=args.cpu_num,
+            parser=args.parser,
+            excludes=args.exclude,
+            strict=not args.no_strict,
+            verbose=args.verbose,
+        )
     except (SyntaxError, NameError, TypeError, ValueError) as e:
         print("Validation failed!\n%s" % str(e))
         exit(1)
diff --git a/yamale/tests/test_command_line.py b/yamale/tests/test_command_line.py
@@ -76,6 +76,19 @@ def test_multiple_paths_bad_yaml():
     assert "map.bad: '12.5' is not a int." in e.value.message
 
 
+def test_excludes():
+    command_line._router(
+        paths=[
+            "yamale/tests/command_line_fixtures/yamls/good.yaml",
+            "yamale/tests/command_line_fixtures/yamls/bad.yaml",
+        ],
+        schema_name="schema.yaml",
+        excludes="bad.yaml",
+        cpus=1,
+        parser="PyYAML",
+    )
+
+
 @pytest.mark.parametrize("parser", parsers)
 def test_good_relative_yaml(parser):
     command_line._router(
@@ -126,6 +139,17 @@ def test_bad_dir():
         command_line._router("yamale/tests/command_line_fixtures/yamls", "schema.yaml", 4, "PyYAML")
 
 
+def test_bad_path_raises():
+    with pytest.raises(ValueError) as e:
+        command_line._router(
+            paths=["yamale/tests/command_line_fixtures/yamls/a path that does not exist.yaml"],
+            schema_name="schema.yaml",
+            cpus=1,
+            parser="PyYAML",
+        )
+    assert "Path does not exist" in str(e)
+
+
 def test_bad_strict():
     with pytest.raises(ValueError) as e:
         command_line._router(
