Added changes to regex pattern to include hyphen

databrickslabs · Dec 22, 2023 · 7437f7b · 7437f7b
1 parent e8d5d28
commit 7437f7b
Show file tree

Hide file tree

Showing 3 changed files with 4 additions and 3 deletions.
diff --git a/discoverx/explorer.py b/discoverx/explorer.py
@@ -17,7 +17,7 @@
 
 
 class DataExplorer:
-    FROM_COMPONENTS_EXPR = r"^(([0-9a-zA-Z_\*]+)\.([0-9a-zA-Z_\*]+)\.([0-9a-zA-Z_\*]+))$"
+    FROM_COMPONENTS_EXPR = r"^(([0-9a-zA-Z_\*-]+)\.([0-9a-zA-Z_\*-]+)\.([0-9a-zA-Z_\*-]+))$"
 
     def __init__(self, from_tables, spark: SparkSession, info_fetcher: InfoFetcher) -> None:
         self._from_tables = from_tables

diff --git a/discoverx/msql.py b/discoverx/msql.py
@@ -22,8 +22,8 @@ class SQLRow:
 class Msql:
     """This class compiles M-SQL expressions into regular SQL"""
 
-    from_statement_expr = r"(FROM\s+)(([0-9a-zA-Z_\*]+).([0-9a-zA-Z_\*]+).([0-9a-zA-Z_\*]+))"
-    from_components_expr = r"^(([0-9a-zA-Z_\*]+).([0-9a-zA-Z_\*]+).([0-9a-zA-Z_\*]+))$"
+    from_statement_expr = r"(FROM\s+)(([0-9a-zA-Z_\*-]+).([0-9a-zA-Z_\*-]+).([0-9a-zA-Z_\*-]+))"
+    from_components_expr = r"^(([0-9a-zA-Z_\*-]+).([0-9a-zA-Z_\*-]+).([0-9a-zA-Z_\*-]+))$"
     command_expr = r"^\s*(\w+)\s"
     class_regex = r"\[([\w_-]+)\]"
     valid_commands = ["SELECT", "DELETE"]

diff --git a/tests/unit/msql_test.py b/tests/unit/msql_test.py
@@ -296,6 +296,7 @@ def test_execute_sql_should_fail_for_no_successful_queries(spark):
 def test_validate_from_components():
     assert Msql.validate_from_components("c.d.t") == ("c", "d", "t")
     assert Msql.validate_from_components("*.*.*") == ("*", "*", "*")
+    assert Msql.validate_from_components("c-1.d.t-1") == ("c-1", "d", "t-1")
 
     with pytest.raises(ValueError):
         Msql.validate_from_components("c.d")