From dd7b6e23dd57551b5005984ec9465675dc5bc341 Mon Sep 17 00:00:00 2001
From: Johan Herland <johan.herland@tweag.io>
Date: Mon, 30 Oct 2023 17:07:08 +0100
Subject: [PATCH] Add --exclude-from option to CLI

Include a few tests demonstrating the successful use of --exclude-from.
Rely on unit tests in previous commits for verifying the underlying
exclude-from functionality.
---
 README.md                | 23 +++++++++-----
 fawltydeps/cli_parser.py | 13 ++++++++
 tests/test_cmdline.py    | 66 ++++++++++++++++++++++++++++++++++++++++
 3 files changed, 94 insertions(+), 8 deletions(-)

diff --git a/README.md b/README.md
index 1642c88d..33480c46 100644
--- a/README.md
+++ b/README.md
@@ -311,9 +311,6 @@ fawltydeps --exclude tests/
 
 The format of the exclude patterns is the same as used by `.gitignore` files,
 [see here for a full description](https://git-scm.com/docs/gitignore#_pattern_format).
-(Note that some patterns that must be interpreted relative to the location of a
-`.gitignore` file are not (yet) supported by FawltyDeps. Search for "relative"
-in the linked page to see which pattern types are not yet supported.)
 
 When the `--exclude` option is not specified, its default value is `".*"`, which
 matches all paths that start with a dot (`.`), aka. "hidden" paths. In the above
@@ -324,8 +321,21 @@ example, if you want to exclude both hidden paths, and everything under
 fawltydeps --exclude tests/ ".*"
 ```
 
-The `--exclude` patterns have lower priority than any paths you pass directly
-on the command line, e.g. in this command:
+(The extra quotes here are needed to prevent the shell from interpreting and
+replacing the `*` wildcard.)
+
+You can also point to exclude patterns stored in a file, with the
+`--exclude-from` option. E.g. to read exclude patterns from `./my_excludes.txt`:
+
+```sh
+fawltydeps --exclude my_excludes.txt
+```
+
+When not given, the `--exclude-from` option will default to reading exclude
+patterns from `./.ignore`, `./.gitignore`, and `./.git/info/exclude`.
+
+Exclude patterns have lower priority than any paths you pass directly on the
+command line, e.g. in this command:
 
 ```sh
 fawltydeps --code my_file.py --exclude my_file.py
@@ -340,9 +350,6 @@ notebooks:
 fawltydeps --code my_dir --exclude "*.ipynb"
 ```
 
-(The extra quotes here are needed to prevent the shell from interpreting and
-replacing the `*` wildcard.)
-
 ### Ignoring irrelevant results
 
 There may be `import` statements in your code that should not be considered an
diff --git a/fawltydeps/cli_parser.py b/fawltydeps/cli_parser.py
index 8236685f..46dd3722 100644
--- a/fawltydeps/cli_parser.py
+++ b/fawltydeps/cli_parser.py
@@ -191,6 +191,19 @@ def populate_parser_paths_options(parser: argparse._ActionsContainer) -> None:
             " Defaults to '.*', meaning that hidden/dot paths are excluded."
         ),
     )
+    parser.add_argument(
+        "--exclude-from",
+        nargs="+",
+        action="union",
+        type=Path,
+        metavar="PATH",
+        help=(
+            "Files containing exclude patterns to apply when looking for code"
+            " (imports), dependency declarations and/or Python environments."
+            " Defaults to '.ignore', `.gitignore`, and `.git/info/exclude`"
+            " under the current directory."
+        ),
+    )
     parser.add_argument(
         "--install-deps",
         dest="install_deps",
diff --git a/tests/test_cmdline.py b/tests/test_cmdline.py
index dec93f7c..057d28b2 100644
--- a/tests/test_cmdline.py
+++ b/tests/test_cmdline.py
@@ -588,6 +588,46 @@ def test_list_sources_detailed__from_both_python_file_and_stdin(fake_project):
     assert returncode == 0
 
 
+def test_list_sources__with_exclude_from(fake_project):
+    tmp_path = fake_project(
+        files_with_imports={
+            "code.py": ["foo"],
+            str(Path("subdir", "notebook.ipynb")): ["foo"],
+            str(Path("subdir", "other.py")): ["foo"],
+        },
+        files_with_declared_deps={
+            "pyproject.toml": ["foo"],
+            "requirements.txt": ["foo"],
+            "setup.cfg": ["foo"],
+            "setup.py": ["foo"],
+        },
+        fake_venvs={"venvs/my_venv": {}},
+        extra_file_contents={
+            "my_ignore": dedent(
+                """
+                subdir/*.py
+                /setup.*
+                *envs
+                """
+            )
+        },
+    )
+    output, returncode = run_fawltydeps_function(
+        "--list-sources", f"{tmp_path}", "--exclude-from", f"{tmp_path / 'my_ignore'}"
+    )
+    expect = [
+        str(tmp_path / filename)
+        for filename in [
+            "code.py",
+            str(Path("subdir", "notebook.ipynb")),
+            "requirements.txt",
+            "pyproject.toml",
+        ]
+    ]
+    assert_unordered_equivalence(output.splitlines()[:-2], expect)
+    assert returncode == 0
+
+
 @dataclass
 class ProjectTestVector:
     """Test vectors for FawltyDeps Settings configuration."""
@@ -1279,6 +1319,32 @@ def test_cmdline_on_ignored_undeclared_option(
             ).splitlines(),
             id="generate_toml_config_with_install_deps",
         ),
+        pytest.param(
+            {"exclude": ["/foo/bar", "baz/*"]},
+            ["--list-sources", "--exclude-from", "my_ignore", "--generate-toml-config"],
+            dedent(
+                f"""\
+                # Copy this TOML section into your pyproject.toml to configure FawltyDeps
+                # (default values are commented)
+                [tool.fawltydeps]
+                actions = ['list_sources']
+                # output_format = 'human_summary'
+                # code = ['.']
+                # deps = ['.']
+                # pyenvs = ['.']
+                # ignore_undeclared = []
+                # ignore_unused = {sorted(DEFAULT_IGNORE_UNUSED)}
+                # deps_parser_choice = ...
+                # install_deps = false
+                exclude = ['/foo/bar', 'baz/*']
+                exclude_from = ['my_ignore']
+                # verbosity = 0
+                # custom_mapping_file = []
+                # [tool.fawltydeps.custom_mapping]
+                """
+            ).splitlines(),
+            id="generate_toml_config_with_list_sources_exclude_and_exclude_from",
+        ),
     ],
 )
 def test_cmdline_args_in_combination_with_config_file(