⚡️ Speed up method ArrowParserWrapper._get_pyarrow_options by 11%

pandas/io/parsers/arrow_parser_wrapper.py

@@ -62,7 +62,8 @@ def _get_pyarrow_options(self) -> None:
             "quotechar": "quote_char",
         }
         for pandas_name, pyarrow_name in mapping.items():
-            if pandas_name in self.kwds and self.kwds.get(pandas_name) is not None:
+            option_value = self.kwds.get(pandas_name)
+            if option_value is not None:
                 self.kwds[pyarrow_name] = self.kwds.pop(pandas_name)
 
         # Date format handling
@@ -80,22 +81,28 @@ def _get_pyarrow_options(self) -> None:
             date_format = None
         self.kwds["timestamp_parsers"] = date_format
 
-        self.parse_options = {
-            option_name: option_value
-            for option_name, option_value in self.kwds.items()
-            if option_value is not None
-            and option_name
-            in ("delimiter", "quote_char", "escape_char", "ignore_empty_lines")
-        }
+        # Prefetch kwds used in parse_options to avoid per-iteration dict lookup
+        delimiter = self.kwds.get("delimiter")
+        quote_char = self.kwds.get("quote_char")
+        escape_char = self.kwds.get("escape_char")
+        ignore_empty_lines = self.kwds.get("ignore_empty_lines")
+
+        self.parse_options = {}
+        if delimiter is not None:
+            self.parse_options["delimiter"] = delimiter
+        if quote_char is not None:
+            self.parse_options["quote_char"] = quote_char
+        if escape_char is not None:
+            self.parse_options["escape_char"] = escape_char
+        if ignore_empty_lines is not None:
+            self.parse_options["ignore_empty_lines"] = ignore_empty_lines
 
         on_bad_lines = self.kwds.get("on_bad_lines")
         if on_bad_lines is not None:
             if callable(on_bad_lines):
                 self.parse_options["invalid_row_handler"] = on_bad_lines
             elif on_bad_lines == ParserBase.BadLineHandleMethod.ERROR:
-                self.parse_options["invalid_row_handler"] = (
-                    None  # PyArrow raises an exception by default
-                )
+                self.parse_options["invalid_row_handler"] = None
             elif on_bad_lines == ParserBase.BadLineHandleMethod.WARN:
 
                 def handle_warning(invalid_row) -> str:
@@ -111,27 +118,41 @@ def handle_warning(invalid_row) -> str:
             elif on_bad_lines == ParserBase.BadLineHandleMethod.SKIP:
                 self.parse_options["invalid_row_handler"] = lambda _: "skip"
 
-        self.convert_options = {
-            option_name: option_value
-            for option_name, option_value in self.kwds.items()
-            if option_value is not None
-            and option_name
-            in (
-                "include_columns",
-                "null_values",
-                "true_values",
-                "false_values",
-                "decimal_point",
-                "timestamp_parsers",
-            )
-        }
-        self.convert_options["strings_can_be_null"] = "" in self.kwds["null_values"]
+        # Prefetch and build convert_options dict directly for known option names
+        # This avoids recreating a larger intermediate dictionary
+        convert_options = {}
+        include_columns = self.kwds.get("include_columns")
+        if include_columns is not None:
+            convert_options["include_columns"] = include_columns
+        null_values = self.kwds.get("null_values")
+        if null_values is not None:
+            convert_options["null_values"] = null_values
+        true_values = self.kwds.get("true_values")
+        if true_values is not None:
+            convert_options["true_values"] = true_values
+        false_values = self.kwds.get("false_values")
+        if false_values is not None:
+            convert_options["false_values"] = false_values
+        decimal_point = self.kwds.get("decimal_point")
+        if decimal_point is not None:
+            convert_options["decimal_point"] = decimal_point
+        timestamp_parsers = self.kwds.get("timestamp_parsers")
+        if timestamp_parsers is not None:
+            convert_options["timestamp_parsers"] = timestamp_parsers
+
+        # Efficient membership check for strings_can_be_null
+        convert_options["strings_can_be_null"] = (
+            "" in null_values if null_values is not None else False
+        )
         # autogenerated column names are prefixed with 'f' in pyarrow.csv
-        if self.header is None and "include_columns" in self.convert_options:
-            self.convert_options["include_columns"] = [
-                f"f{n}" for n in self.convert_options["include_columns"]
+        if self.header is None and "include_columns" in convert_options:
+            convert_options["include_columns"] = [
+                f"f{n}" for n in convert_options["include_columns"]
             ]
 
+        self.convert_options = convert_options
+
+        # Compose read_options without extra getitem overhead
         self.read_options = {
             "autogenerate_column_names": self.header is None,
             "skip_rows": self.header