Handle dash in the regex2glob function

mlcommons · Jan 24, 2025 · f02c756 · f02c756
1 parent e96852f
commit f02c756
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 0 deletions.
diff --git a/python/mlcroissant/mlcroissant/_src/core/regex.py b/python/mlcroissant/mlcroissant/_src/core/regex.py
@@ -65,6 +65,8 @@ def _regex_to_glob_for_str(regex: str) -> Iterable[str]:
     regex = re.sub(r"\.\*", "*", regex)
     # Interpret .+ as *
     regex = re.sub(r"\.\+", "*", regex)
+    # Interpret \\- as -
+    regex = re.sub(r"\\-", "-", regex)
     return [regex]
 
 

diff --git a/python/mlcroissant/mlcroissant/_src/core/regex_test.py b/python/mlcroissant/mlcroissant/_src/core/regex_test.py
@@ -27,6 +27,12 @@
                 "*/train/*.parquet",  # ...to a valid glob pattern.
             ],
         ],
+        [
+            "^.+/my\\-train/.*\.parquet$",  # From a valid regex...
+            [
+                "*/my-train/*.parquet",  # ...to a valid glob pattern.
+            ],
+        ],
     ],
 )
 def test_regex_to_glob(regex: str, output: list[str]):