From ee5bc6eda76cfdb652e97949a9550f686a15fb93 Mon Sep 17 00:00:00 2001
From: Gregor Sturm <gregor.sturm@boehringer-ingelheim.com>
Date: Fri, 29 Nov 2024 13:14:12 +0100
Subject: [PATCH] Close #62

---
 CHANGELOG.md       |  8 ++++++--
 src/dso/lint.py    |  9 +++++++++
 tests/test_lint.py | 15 +++++++++++++++
 3 files changed, 30 insertions(+), 2 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index beb5439..de64276 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -8,11 +8,15 @@ and this project adheres to [Semantic Versioning][].
 [keep a changelog]: https://keepachangelog.com/en/1.0.0/
 [semantic versioning]: https://semver.org/spec/v2.0.0.html
 
-## [Unreleased]
+## v0.10.1
+
+### Fixes
+
+-   Take comments into account when linting for `DSO001`
 
 ### Template updates
 
--   Do not ignore gitignore files in output/report directories of template ([#63](https://github.com/Boehringer-Ingelheim/dso/pull/63))
+-   Do not ignore the `.gitignore` files in output/report directories of template ([#63](https://github.com/Boehringer-Ingelheim/dso/pull/63))
 
 ## v0.10.0
 
diff --git a/src/dso/lint.py b/src/dso/lint.py
index 304e211..3dad81c 100644
--- a/src/dso/lint.py
+++ b/src/dso/lint.py
@@ -87,6 +87,15 @@ def check(cls, file):
         # .parent to remove the dvc.yaml filename
         stage_path_expected = str(stage_path_expected.parent.relative_to(root_path))
         content = file.read_text()
+
+        # remove comments
+        # TODO there are still edge cases, e.g. a `#` within a string doesn't initiate a comment
+        # However this is hard/impossible (?) to solve with a regular expression alone, we'd need a proper
+        # R parse to address all edge cases. See also https://github.com/Boehringer-Ingelheim/dso/issues/66
+        pattern_is_comment = re.compile(r"#.*$")
+        content = "\n".join([re.sub(pattern_is_comment, "", line) for line in content.split("\n")])
+
+        # detect pattern
         pattern = r"[\s\S]*?(dso::)?read_params\s*\(([\s\S]*?)(\s*,.*)?\)"
         res = re.findall(pattern, content, flags=re.MULTILINE)
         if len(res) == 0:
diff --git a/tests/test_lint.py b/tests/test_lint.py
index bb3ea91..bfc6718 100644
--- a/tests/test_lint.py
+++ b/tests/test_lint.py
@@ -116,6 +116,21 @@ class MockQuartoRule(QuartoRule):
             """,
             LintError,
         ),
+        (
+            """\
+            params = read_params("quarto_stage")
+            # params = read_params("quarto_stage")
+            """,
+            None,
+        ),
+        # TODO no good way to cover that with regex alone, see https://github.com/Boehringer-Ingelheim/dso/issues/66
+        # (
+        #     """\
+        #     params = read_params("quarto_stage")
+        #     print(" foo # no comment"); params = read_params("quarto_stage")
+        #     """,
+        #     LintError,
+        # ),
         (
             """\
             params = read_params("wrong_path")