ENH: Improve to_csv and from_{fwf,csv,table}

pandas-stubs/_libs/lib.pyi

@@ -1,8 +1,3 @@
-from enum import Enum
-
 no_default = None
 
-class NoDefault(Enum):
-    no_default: int
-
 def infer_dtype(value: object, skipna: bool = ...) -> str: ...

pandas-stubs/_typing.pyi

@@ -242,6 +242,7 @@ JsonSeriesOrient = Literal["split", "records", "index"]
 TimestampConvention = Literal["start", "end", "s", "e"]
 
 CSVEngine = Literal["c", "python", "pyarrow", "python-fwf"]
+CSVQuoting = Literal[0, 1, 2, 3]
 
 HDFCompLib = Literal["zlib", "lzo", "bzip2", "blosc"]
 ParquetEngine = Literal["auto", "pyarrow", "fastparquet"]

pandas-stubs/io/clipboards.pyi

@@ -15,6 +15,7 @@ from pandas.core.series import Series
 from pandas._typing import (
     CompressionOptions,
     CSVEngine,
+    CSVQuoting,
     DtypeArg,
     StorageOptions,
     npt,
@@ -71,7 +72,7 @@ def read_clipboard(
     decimal: str = ...,
     lineterminator: str | None = ...,
     quotechar: str = ...,
-    quoting: int = ...,
+    quoting: CSVQuoting | None = ...,
     doublequote: bool = ...,
     escapechar: str | None = ...,
     comment: str | None = ...,
@@ -137,7 +138,7 @@ def read_clipboard(
     decimal: str = ...,
     lineterminator: str | None = ...,
     quotechar: str = ...,
-    quoting: int = ...,
+    quoting: CSVQuoting | None = ...,
     doublequote: bool = ...,
     escapechar: str | None = ...,
     comment: str | None = ...,
@@ -203,7 +204,7 @@ def read_clipboard(
     decimal: str = ...,
     lineterminator: str | None = ...,
     quotechar: str = ...,
-    quoting: int = ...,
+    quoting: CSVQuoting | None = ...,
     doublequote: bool = ...,
     escapechar: str | None = ...,
     comment: str | None = ...,

pandas-stubs/io/parsers/readers.pyi

@@ -17,6 +17,7 @@ import pandas._libs.lib as lib
 from pandas._typing import (
     CompressionOptions,
     CSVEngine,
+    CSVQuoting,
     DtypeArg,
     FilePath,
     ReadCsvBuffer,
@@ -30,10 +31,10 @@ from pandas.io.common import IOHandles
 def read_csv(
     filepath_or_buffer: FilePath | ReadCsvBuffer[bytes] | ReadCsvBuffer[str],
     *,
-    sep: str | None | lib.NoDefault = ...,
-    delimiter: str | None | lib.NoDefault = ...,
+    sep: str | None = ...,
+    delimiter: str | None = ...,
     header: int | Sequence[int] | Literal["infer"] | None = ...,
-    names: list[str] | None | lib.NoDefault = ...,
+    names: list[str] | None = ...,
     index_col: int | str | Sequence[str | int] | Literal[False] | None = ...,
     usecols: list[str]
     | tuple[str, ...]
@@ -74,7 +75,7 @@ def read_csv(
     decimal: str = ...,
     lineterminator: str | None = ...,
     quotechar: str = ...,
-    quoting: int = ...,
+    quoting: CSVQuoting | None = ...,
     doublequote: bool = ...,
     escapechar: str | None = ...,
     comment: str | None = ...,
@@ -93,10 +94,10 @@ def read_csv(
 def read_csv(
     filepath_or_buffer: FilePath | ReadCsvBuffer[bytes] | ReadCsvBuffer[str],
     *,
-    sep: str | None | lib.NoDefault = ...,
-    delimiter: str | None | lib.NoDefault = ...,
+    sep: str | None = ...,
+    delimiter: str | None = ...,
     header: int | Sequence[int] | Literal["infer"] | None = ...,
-    names: list[str] | None | lib.NoDefault = ...,
+    names: list[str] | None = ...,
     index_col: int | str | Sequence[str | int] | Literal[False] | None = ...,
     usecols: list[str]
     | tuple[str, ...]
@@ -137,7 +138,7 @@ def read_csv(
     decimal: str = ...,
     lineterminator: str | None = ...,
     quotechar: str = ...,
-    quoting: int = ...,
+    quoting: CSVQuoting | None = ...,
     doublequote: bool = ...,
     escapechar: str | None = ...,
     comment: str | None = ...,
@@ -156,10 +157,10 @@ def read_csv(
 def read_csv(
     filepath_or_buffer: FilePath | ReadCsvBuffer[bytes] | ReadCsvBuffer[str],
     *,
-    sep: str | None | lib.NoDefault = ...,
-    delimiter: str | None | lib.NoDefault = ...,
+    sep: str | None = ...,
+    delimiter: str | None = ...,
     header: int | Sequence[int] | Literal["infer"] | None = ...,
-    names: list[str] | None | lib.NoDefault = ...,
+    names: list[str] | None = ...,
     index_col: int | str | Sequence[str | int] | Literal[False] | None = ...,
     usecols: list[str]
     | tuple[str, ...]
@@ -200,7 +201,7 @@ def read_csv(
     decimal: str = ...,
     lineterminator: str | None = ...,
     quotechar: str = ...,
-    quoting: int = ...,
+    quoting: CSVQuoting | None = ...,
     doublequote: bool = ...,
     escapechar: str | None = ...,
     comment: str | None = ...,
@@ -219,10 +220,10 @@ def read_csv(
 def read_table(
     filepath_or_buffer: FilePath | ReadCsvBuffer[bytes] | ReadCsvBuffer[str],
     *,
-    sep: str | None | lib.NoDefault = ...,
-    delimiter: str | None | lib.NoDefault = ...,
+    sep: str | None = ...,
+    delimiter: str | None = ...,
     header: int | Sequence[int] | Literal["infer"] | None = ...,
-    names: list[str] | None | lib.NoDefault = ...,
+    names: list[str] | None = ...,
     index_col: int | str | Sequence[str | int] | Literal[False] | None = ...,
     usecols: list[str]
     | tuple[str, ...]
@@ -263,7 +264,7 @@ def read_table(
     decimal: str = ...,
     lineterminator: str | None = ...,
     quotechar: str = ...,
-    quoting: int = ...,
+    quoting: CSVQuoting | None = ...,
     doublequote: bool = ...,
     escapechar: str | None = ...,
     comment: str | None = ...,
@@ -282,10 +283,10 @@ def read_table(
 def read_table(
     filepath_or_buffer: FilePath | ReadCsvBuffer[bytes] | ReadCsvBuffer[str],
     *,
-    sep: str | None | lib.NoDefault = ...,
-    delimiter: str | None | lib.NoDefault = ...,
+    sep: str | None = ...,
+    delimiter: str | None = ...,
     header: int | Sequence[int] | Literal["infer"] | None = ...,
-    names: list[str] | None | lib.NoDefault = ...,
+    names: list[str] | None = ...,
     index_col: int | str | Sequence[str | int] | Literal[False] | None = ...,
     usecols: list[str]
     | tuple[str, ...]
@@ -326,7 +327,7 @@ def read_table(
     decimal: str = ...,
     lineterminator: str | None = ...,
     quotechar: str = ...,
-    quoting: int = ...,
+    quoting: CSVQuoting | None = ...,
     doublequote: bool = ...,
     escapechar: str | None = ...,
     comment: str | None = ...,
@@ -345,10 +346,10 @@ def read_table(
 def read_table(
     filepath_or_buffer: FilePath | ReadCsvBuffer[bytes] | ReadCsvBuffer[str],
     *,
-    sep: str | None | lib.NoDefault = ...,
-    delimiter: str | None | lib.NoDefault = ...,
+    sep: str | None = ...,
+    delimiter: str | None = ...,
     header: int | Sequence[int] | Literal["infer"] | None = ...,
-    names: list[str] | None | lib.NoDefault = ...,
+    names: list[str] | None = ...,
     index_col: int | str | Sequence[str | int] | Literal[False] | None = ...,
     usecols: list[str]
     | tuple[str, ...]
@@ -389,7 +390,7 @@ def read_table(
     decimal: str = ...,
     lineterminator: str | None = ...,
     quotechar: str = ...,
-    quoting: int = ...,
+    quoting: CSVQuoting | None = ...,
     doublequote: bool = ...,
     escapechar: str | None = ...,
     comment: str | None = ...,

tests/test_io.py

@@ -10,6 +10,7 @@
     Union,
 )
 
+import numpy as np
 from packaging.version import parse
 import pandas as pd
 from pandas import (
@@ -381,6 +382,48 @@ def test_read_csv_iterator():
         )
 
 
+def test_types_read_csv() -> None:
+    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
+    csv_df: str = df.to_csv()
+
+    with ensure_clean() as path:
+        df.to_csv(path)
+        df2: pd.DataFrame = pd.read_csv(path)
+        df3: pd.DataFrame = pd.read_csv(path, sep="a")
+        df4: pd.DataFrame = pd.read_csv(
+            path,
+            header=None,
+        )
+        df5: pd.DataFrame = pd.read_csv(
+            path, engine="python", true_values=["no", "No", "NO"], na_filter=False
+        )
+        df6: pd.DataFrame = pd.read_csv(
+            path,
+            skiprows=lambda x: x in [0, 2],
+            skip_blank_lines=True,
+            dayfirst=False,
+        )
+        df7: pd.DataFrame = pd.read_csv(path, nrows=2)
+        df8: pd.DataFrame = pd.read_csv(path, dtype={"a": float, "b": int})
+        df9: pd.DataFrame = pd.read_csv(path, usecols=["col1"])
+        df10: pd.DataFrame = pd.read_csv(path, usecols=[0])
+        df11: pd.DataFrame = pd.read_csv(path, usecols=np.array([0]))
+        df12: pd.DataFrame = pd.read_csv(path, usecols=("col1",))
+        df13: pd.DataFrame = pd.read_csv(path, usecols=pd.Series(data=["col1"]))
+
+        tfr1: TextFileReader = pd.read_csv(path, nrows=2, iterator=True, chunksize=3)
+        tfr1.close()
+
+        tfr2: TextFileReader = pd.read_csv(path, nrows=2, chunksize=1)
+        tfr2.close()
+
+        tfr3: TextFileReader = pd.read_csv(path, nrows=2, iterator=False, chunksize=1)
+        tfr3.close()
+
+        tfr4: TextFileReader = pd.read_csv(path, nrows=2, iterator=True)
+        tfr4.close()
+
+
 def test_read_table():
     with ensure_clean() as path:
         check(assert_type(DF.to_csv(path, sep="\t"), None), type(None))
@@ -455,6 +498,75 @@ def test_to_csv_series():
         check(assert_type(s.to_csv(path), None), type(None))
 
 
+def test_read_sql():
+    with ensure_clean() as path:
+        con = sqlite3.connect(path)
+        check(assert_type(DF.to_sql("test", con=con), Union[int, None]), int)
+        check(
+            assert_type(read_sql("select * from test", con=con), DataFrame), DataFrame
+        )
+        con.close()
+
+
+def test_read_sql_generator():
+    with ensure_clean() as path:
+        con = sqlite3.connect(path)
+        check(assert_type(DF.to_sql("test", con=con), Union[int, None]), int)
+
+        check(
+            assert_type(
+                read_sql("select * from test", con=con, chunksize=1),
+                Generator[DataFrame, None, None],
+            ),
+            Generator,
+        )
+        con.close()
+
+
+def test_read_sql_table():
+    if TYPE_CHECKING:
+        # sqlite3 doesn't support read_table, which is required for this function
+        # Could only run in pytest if SQLAlchemy was installed
+        with ensure_clean() as path:
+            con = sqlite3.connect(path)
+            assert_type(DF.to_sql("test", con=con), Union[int, None])
+            assert_type(read_sql_table("test", con=con), DataFrame)
+            assert_type(
+                read_sql_table("test", con=con, chunksize=1),
+                Generator[DataFrame, None, None],
+            )
+            con.close()
+
+
+def test_read_sql_query():
+    with ensure_clean() as path:
+        con = sqlite3.connect(path)
+        check(assert_type(DF.to_sql("test", con=con), Union[int, None]), int)
+        check(
+            assert_type(
+                read_sql_query("select * from test", con=con, index_col="index"),
+                DataFrame,
+            ),
+            DataFrame,
+        )
+        con.close()
+
+
+def test_read_sql_query_generator():
+    with ensure_clean() as path:
+        con = sqlite3.connect(path)
+        check(assert_type(DF.to_sql("test", con=con), Union[int, None]), int)
+
+        check(
+            assert_type(
+                read_sql_query("select * from test", con=con, chunksize=1),
+                Generator[DataFrame, None, None],
+            ),
+            Generator,
+        )
+        con.close()
+
+
 def test_read_html():
     check(assert_type(DF.to_html(), str), str)
     with ensure_clean() as path: