Update identifier escaping logic #1055

Signed-off-by: Marcel Coetzee <[email protected]>
dlt-hub · Mar 14, 2024 · c1f106b · c1f106b
1 parent 0df32f5
commit c1f106b
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 47 deletions.
diff --git a/dlt/common/data_writers/escape.py b/dlt/common/data_writers/escape.py
@@ -137,42 +137,5 @@ def escape_databricks_literal(v: Any) -> Any:
     return "NULL" if v is None else str(v)
 
 
-# https://github.com/ClickHouse/ClickHouse/blob/master/docs/en/sql-reference/syntax.md#string
-CLICKHOUSE_ESCAPE_DICT = {
-    "'": "''",
-    "\\": "\\\\",
-    "\n": "\\n",
-    "\t": "\\t",
-    "\b": "\\b",
-    "\f": "\\f",
-    "\r": "\\r",
-    "\0": "\\0",
-    "\a": "\\a",
-    "\v": "\\v",
-}
-
-CLICKHOUSE_ESCAPE_RE = _make_sql_escape_re(CLICKHOUSE_ESCAPE_DICT)
-
-
-def escape_clickhouse_literal(v: Any) -> Any:
-    if isinstance(v, str):
-        return _escape_extended(
-            v, prefix="'", escape_dict=CLICKHOUSE_ESCAPE_DICT, escape_re=CLICKHOUSE_ESCAPE_RE
-        )
-    if isinstance(v, (datetime, date, time)):
-        return f"'{v.isoformat()}'"
-    if isinstance(v, (list, dict)):
-        return _escape_extended(
-            json.dumps(v),
-            prefix="'",
-            escape_dict=CLICKHOUSE_ESCAPE_DICT,
-            escape_re=CLICKHOUSE_ESCAPE_RE,
-        )
-    if isinstance(v, bytes):
-        return f"'{v.hex()}'"
-    return "NULL" if v is None else str(v)
-
-
-def escape_clickhouse_identifier(v: str, quote_char: str = "`") -> str:
-    quote_char = quote_char if quote_char in {'"', "`"} else "`"
-    return quote_char + v.replace(quote_char, quote_char * 2).replace("\\", "\\\\") + quote_char
+def escape_clickhouse_identifier(v: str) -> str:
+    return '`' + v.replace('`', '``').replace("\\", "\\\\") + '"'
diff --git a/dlt/destinations/impl/clickhouse/__init__.py b/dlt/destinations/impl/clickhouse/__init__.py
@@ -1,18 +1,17 @@
 from dlt.common.arithmetics import DEFAULT_NUMERIC_PRECISION, DEFAULT_NUMERIC_SCALE
-from dlt.common.data_writers.escape import escape_clickhouse_identifier, escape_clickhouse_literal
+from dlt.common.data_writers.escape import escape_clickhouse_identifier
 from dlt.common.destination import DestinationCapabilitiesContext
 
 
 def capabilities() -> DestinationCapabilitiesContext:
     caps = DestinationCapabilitiesContext()
 
-    caps.preferred_loader_file_format = "parquet"
-    caps.supported_loader_file_formats = ["jsonl", "parquet", "insert_values"]
-    caps.preferred_staging_file_format = "parquet"
-    caps.supported_staging_file_formats = ["jsonl", "parquet"]
+    caps.preferred_loader_file_format = "jsonl"
+    caps.supported_loader_file_formats = ["jsonl", "parquet", "arrow"]
+    caps.preferred_staging_file_format = "jsonl"
+    caps.supported_staging_file_formats = ["jsonl", "parquet", "arrow"]
 
     caps.escape_identifier = escape_clickhouse_identifier
-    caps.escape_literal = escape_clickhouse_literal
 
     caps.schema_supports_numeric_precision = True
     # Use 'Decimal128' with these defaults.
@@ -25,8 +24,8 @@ def capabilities() -> DestinationCapabilitiesContext:
     caps.is_max_query_length_in_bytes = True
     caps.max_query_length = 262144
 
-    # Clickhouse has limited support for transactional semantics, especially for `ReplicatedMergeTree`, the default ClickHouse Cloud engine.
-    # It does, however, provide atomicity for individual DDL operations like `ALTER TABLE`.
+    # Clickhouse has limited support for transactional semantics, especially for `ReplicatedMergeTree`,
+    # the default ClickHouse Cloud engine. It does, however, provide atomicity for individual DDL operations like `ALTER TABLE`.
     # https://clickhouse-driver.readthedocs.io/en/latest/dbapi.html#clickhouse_driver.dbapi.connection.Connection.commit
     # https://clickhouse.com/docs/en/guides/developer/transactional#transactions-commit-and-rollback
     caps.supports_transactions = False