hypertabular 0.7.0__cp311-abi3-win_arm64.whl

hypertabular/__init__.py

@@ -0,0 +1,489 @@
+"""Delimited text — CSV, TSV, any single-byte ASCII separator — and workbooks — XLSX and
+ODS — read a batch at a time into typed columns, with a
+`HyperCast <https://github.com/SkunkWerkx/HyperCast>`_ verdict for every cell.
+
+The native core owns no memory and reads no files. It is linked straight into a CPython
+extension module (``hypertabular._native``, PyO3), and the extension is the binding: it
+allocates a value array and a verdict array per column, once, and the core fills them in
+one call per batch, with the interpreter released. A column then reaches Python whole — a
+read-only ``memoryview`` of the door's own item type over the batch's copy of that array,
+which anything that takes a buffer takes — with no Python call per cell::
+
+    from hypertabular import Column, DelimitedReader, Dialect, Fault, Success
+
+    plan = [Column.i32(0), Column.text(1), Column.f64(2)]
+    with DelimitedReader.open("orders.csv", Dialect.CSV, plan) as reader:
+        for batch in reader:
+            ids, names, scores = batch.columns
+
+            # A column at a time, as the core wrote it…
+            total = sum(scores.values)            # memoryview of C doubles
+            if scores.fault_count:
+                for row, fault in scores.faults():
+                    print(row, fault.reason.name, scores.raw(row))
+
+            # …or a cell at a time, as HyperCast's union.
+            for row in range(batch.rows):
+                match scores[row]:
+                    case Success(value):
+                        print(names.values[row], value)
+                    case Fault(reason, offset, length):
+                        print(reason.name, "on line", batch.line(row), "in", scores.raw(row))
+
+    # A workbook reads into the same batch.
+    book = Workbook.open("orders.xlsx")
+    for batch in book.sheet("Orders", SheetOptions(), plan):
+        ...
+
+- **Nothing is sniffed.** The :class:`Dialect` states the separator, the quoting and the
+  header; :class:`SheetOptions` states a sheet's header and whether empty rows are skipped;
+  the plan states each column's door and, for numbers, its ``NumFormat``.
+- **HyperCast is the judge.** :class:`Success`, :class:`Fault`, :class:`CastFailure`,
+  :class:`NumFormat`, :class:`UnixPrecision`, :class:`DateOrder` and :class:`ExcelEpoch`
+  are the ``hypercast`` package's own objects, re-exported here unchanged — not copies. A
+  text cell means exactly what ``hypercast.cast_*`` says of the same text, a typed workbook
+  cell is converted by the door directly, and its value is the
+  Python type that door gives: ``int``, ``float``, ``bool``, ``decimal.Decimal``,
+  ``uuid.UUID``, ``datetime``, ``date``, ``time``, ``timedelta`` (and ``str`` for text).
+- **A bad value is a verdict; a broken file is an exception.** A cell that does not cast
+  is a ``Fault`` in its column and the read goes on. A record of the wrong width, input
+  that ends inside a quoted cell, a workbook whose container or parts cannot be read,
+  raises :class:`TabularError` — after every intact row before it has been delivered.
+- **A batch owns what it shows.** It stays valid after the reader has moved on.
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+from enum import IntEnum
+from typing import ClassVar
+
+from hypercast import (
+    CastFailure,
+    DateOrder,
+    ExcelEpoch,
+    Fault,
+    NumFormat,
+    Success,
+    UnixPrecision,
+    Verdict,
+)
+
+from . import _native
+
+#: Which backend this process loaded: always ``"native"``, the PyO3 extension that links the
+#: Rust core straight into CPython — the only backend, and the one every wheel ships.
+BACKEND: str = "native"
+
+__all__ = [
+    "BACKEND",
+    "native_version",
+    "DelimitedReader",
+    "Workbook",
+    "WorkbookFormat",
+    "Sheet",
+    "SheetInfo",
+    "SheetOptions",
+    "Batch",
+    "ColumnData",
+    "Dialect",
+    "Column",
+    "Door",
+    "TabularError",
+    "TabularFailure",
+    # HyperCast's own, re-exported unchanged.
+    "CastFailure",
+    "Success",
+    "Fault",
+    "Verdict",
+    "NumFormat",
+    "UnixPrecision",
+    "DateOrder",
+    "ExcelEpoch",
+]
+
+
+class Door(IntEnum):
+    """The door a column is cast through: HyperCast's, plus :data:`TEXT` for the bytes
+    themselves. The values are the core's own door codes."""
+
+    BOOL = 1
+    I8 = 2
+    I16 = 3
+    I32 = 4
+    I64 = 5
+    U8 = 6
+    U16 = 7
+    U32 = 8
+    U64 = 9
+    F32 = 10
+    F64 = 11
+    UUID = 12
+    TIMESTAMP = 13
+    UNIX = 14
+    DATE = 15
+    TIME = 16
+    DURATION = 17
+    TEXT = 18
+    DECIMAL = 19
+    DATE_ORDERED = 20
+    DATETIME = 21
+    EXCEL_SERIAL = 22
+
+
+class TabularFailure(IntEnum):
+    """Why an input could not be read as rows at all. The values are the core's codes."""
+
+    UNCLOSED_QUOTE = 1
+    """The input ended inside a quoted cell."""
+    COLUMN_COUNT = 2
+    """A record's cell count disagrees with the first record's."""
+    ROW_TOO_LONG = 3
+    """A single record is larger than :data:`DelimitedReader.MAX_ROW_BYTES`."""
+    NOT_A_ZIP = 16
+    """The workbook's container is not a zip file."""
+    CONTAINER = 17
+    """The zip's own structure is broken."""
+    ENCRYPTED = 18
+    """The workbook is encrypted."""
+    METHOD = 19
+    """A part is compressed by a method other than stored or deflate."""
+    MISSING_PART = 20
+    """A part the workbook cannot be read without is missing."""
+    XML = 21
+    """A part's XML ends inside a construct."""
+    DEFLATE = 22
+    """A part's bytes are not a deflate stream, or stop before the stream does."""
+    NOT_A_WORKBOOK = 23
+    """The zip is neither an XLSX nor an ODS workbook."""
+    SHARED_STRING = 24
+    """A cell names a shared string the table does not have."""
+    TOO_LARGE = 25
+    """More text than can be addressed: over 4 GiB in a batch or in the shared strings, or
+    2 GiB in a cell."""
+
+
+class WorkbookFormat(IntEnum):
+    """Which kind of workbook a :class:`Workbook` is."""
+
+    XLSX = 1
+    """Office Open XML: ``.xlsx``, ``.xlsm``."""
+    ODS = 2
+    """OpenDocument: ``.ods``."""
+
+
+class TabularError(Exception):
+    """A structural failure: the input is not rows of cells — a record of the wrong width,
+    input that ends inside a quoted cell, a workbook whose container or parts cannot be
+    read. Never a cell's verdict: a value that does not cast is a ``Fault`` in its column,
+    and the read goes on. A structural failure ends the input, after every intact row
+    before it has been delivered.
+
+    For a workbook, :attr:`record` is the part the failure is in, :attr:`line` the sheet
+    row and :attr:`byte` the offset within the part's inflated bytes.
+    """
+
+    kind: TabularFailure
+    """What is wrong."""
+    record: int
+    """Zero-based index of the offending record — the header and skipped blank lines included."""
+    line: int
+    """One-based line the offending record starts on."""
+    byte: int
+    """Absolute byte offset of the offending record's start."""
+    expected: int
+    """Cells in the first record, for :data:`TabularFailure.COLUMN_COUNT`."""
+    found: int
+    """Cells in this record, for :data:`TabularFailure.COLUMN_COUNT`."""
+
+    def __init__(
+        self,
+        kind: TabularFailure,
+        record: int,
+        line: int,
+        byte: int,
+        expected: int = 0,
+        found: int = 0,
+    ) -> None:
+        """Builds the failure the reader raises; the arguments are its attributes."""
+        # The arguments themselves, not the message, so the exception copies and pickles.
+        super().__init__(kind, record, line, byte, expected, found)
+        self.kind = TabularFailure(kind)
+        self.record = record
+        self.line = line
+        self.byte = byte
+        self.expected = expected
+        self.found = found
+
+    def __str__(self) -> str:
+        """The failure in words, with its position."""
+        where = f"record {self.record} (line {self.line}, byte {self.byte})"
+        part = f"part {self.record} of the workbook"
+        match self.kind:
+            case TabularFailure.COLUMN_COUNT:
+                return f"{where} has {self.found} cells; the first record had {self.expected}"
+            case TabularFailure.UNCLOSED_QUOTE:
+                return f"the input ended inside a quoted cell in {where}"
+            case TabularFailure.ROW_TOO_LONG:
+                limit = _native.DelimitedReader.MAX_ROW_BYTES
+                return f"{where} exceeds the {limit}-byte row ceiling"
+            case TabularFailure.NOT_A_ZIP:
+                return "the workbook is not a zip file"
+            case TabularFailure.ENCRYPTED:
+                return "the workbook is encrypted"
+            case TabularFailure.METHOD:
+                return f"{part} is compressed by method {self.found}, neither stored nor deflate"
+            case TabularFailure.MISSING_PART:
+                return f"part {self.record}, which the workbook cannot be read without, is missing"
+            case TabularFailure.XML:
+                return f"{part} ends inside an XML construct (byte {self.byte})"
+            case TabularFailure.DEFLATE:
+                return f"{part} is not a whole deflate stream (byte {self.byte})"
+            case TabularFailure.NOT_A_WORKBOOK:
+                return "the zip is neither an XLSX nor an ODS workbook"
+            case TabularFailure.SHARED_STRING:
+                return f"row {self.line} names shared string {self.found}; the table has {self.expected}"
+            case TabularFailure.TOO_LARGE:
+                return "the workbook holds more text than a batch can address"
+            case _:
+                return "the workbook's zip structure is broken"
+
+
+@dataclass(frozen=True, slots=True)
+class Dialect:
+    """How the text is delimited, declared by the caller. Nothing is sniffed: the separator
+    is stated, quoting is stated, the header is stated — the same stance HyperCast's
+    ``NumFormat`` takes for numeric notation.
+    """
+
+    separator: str
+    """The single-byte separator: tab, or any printable ASCII character except ``"``."""
+    quoting: bool = True
+    """Whether ``"`` quotes cells (RFC 4180, ``""`` for a literal quote). Off, a quote is
+    an ordinary byte."""
+    has_header: bool = True
+    """Whether the first record is a header, exposed through
+    :attr:`DelimitedReader.header` and never delivered as a row."""
+    skip_blank_lines: bool = True
+    """Whether a completely empty line is skipped rather than read as a one-cell row."""
+
+    CSV: ClassVar[Dialect]
+    """Comma-separated, quoted, with a header, blank lines skipped."""
+    TSV: ClassVar[Dialect]
+    """Tab-separated, otherwise as :data:`CSV`."""
+    PSV: ClassVar[Dialect]
+    """Pipe-separated, otherwise as :data:`CSV`."""
+
+    def __post_init__(self) -> None:
+        """Refuses a separator the scanner cannot honour — a caller bug, not a data verdict."""
+        if not isinstance(self.separator, str):
+            raise TypeError(f"separator must be a str, not {type(self.separator).__name__}")
+        valid = len(self.separator) == 1 and (
+            self.separator == "\t" or (" " <= self.separator <= "~" and self.separator != '"')
+        )
+        if not valid:
+            raise ValueError(
+                f"separator {self.separator!r} is not tab or a printable ASCII character other than '\"'"
+            )
+
+
+Dialect.CSV = Dialect(",")
+Dialect.TSV = Dialect("\t")
+Dialect.PSV = Dialect("|")
+
+
+@dataclass(frozen=True, slots=True, repr=False)
+class Column:
+    """One output column of a plan: which source column it reads, the door it casts
+    through, and — for the numeric doors — the notation. Build one with the factory named
+    for its door (the names are HyperCast's: ``Column.i32`` is what ``hypercast.cast_i32``
+    would make of each cell).
+
+    A plan is a projection: a forty-column file can be read into five typed columns, in
+    any order, and a source column can be read through more than one door.
+    """
+
+    ordinal: int
+    """Zero-based ordinal of the source column. Past a record's last cell reads as empty."""
+    door: Door
+    """The door."""
+    format: NumFormat = NumFormat.INVARIANT
+    """The numeric notation, read by the numeric doors."""
+    declared: UnixPrecision | DateOrder | ExcelEpoch | None = None
+    """What the door declares beside itself: the Unix precision, the date order, or the
+    Excel date system — never guessed."""
+
+    def __post_init__(self) -> None:
+        """Refuses an ordinal no source column can have."""
+        if not isinstance(self.ordinal, int) or isinstance(self.ordinal, bool):
+            raise TypeError(f"ordinal must be an int, not {type(self.ordinal).__name__}")
+        if not 0 <= self.ordinal < 2**31:
+            raise ValueError(f"ordinal must be between 0 and 2**31 - 1, not {self.ordinal}")
+
+    def __repr__(self) -> str:
+        """The column as the factory call that makes it."""
+        said = [str(self.ordinal)]
+        if self.declared is not None:
+            said.append(f"{type(self.declared).__name__}.{self.declared.name}")
+        if self.format is not NumFormat.INVARIANT:
+            said.append(repr(self.format))
+        return f"Column.{self.door.name.lower()}({', '.join(said)})"
+
+    @classmethod
+    def _numeric(cls, ordinal: int, door: Door, fmt: NumFormat) -> Column:
+        if not isinstance(fmt, NumFormat):
+            raise TypeError(f"fmt must be a hypercast.NumFormat, not {type(fmt).__name__}")
+        return cls(ordinal, door, fmt)
+
+    @classmethod
+    def bool(cls, ordinal: int) -> Column:
+        """A ``bool`` column: HyperCast's boolean lexicon."""
+        return cls(ordinal, Door.BOOL)
+
+    @classmethod
+    def i8(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """A signed 8-bit integer column under the declared notation."""
+        return cls._numeric(ordinal, Door.I8, fmt)
+
+    @classmethod
+    def i16(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """A signed 16-bit integer column under the declared notation."""
+        return cls._numeric(ordinal, Door.I16, fmt)
+
+    @classmethod
+    def i32(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """A signed 32-bit integer column under the declared notation."""
+        return cls._numeric(ordinal, Door.I32, fmt)
+
+    @classmethod
+    def i64(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """A signed 64-bit integer column under the declared notation."""
+        return cls._numeric(ordinal, Door.I64, fmt)
+
+    @classmethod
+    def u8(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """An unsigned 8-bit integer column under the declared notation."""
+        return cls._numeric(ordinal, Door.U8, fmt)
+
+    @classmethod
+    def u16(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """An unsigned 16-bit integer column under the declared notation."""
+        return cls._numeric(ordinal, Door.U16, fmt)
+
+    @classmethod
+    def u32(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """An unsigned 32-bit integer column under the declared notation."""
+        return cls._numeric(ordinal, Door.U32, fmt)
+
+    @classmethod
+    def u64(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """An unsigned 64-bit integer column under the declared notation."""
+        return cls._numeric(ordinal, Door.U64, fmt)
+
+    @classmethod
+    def f32(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """An IEEE single column under the declared notation."""
+        return cls._numeric(ordinal, Door.F32, fmt)
+
+    @classmethod
+    def f64(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """An IEEE double column under the declared notation."""
+        return cls._numeric(ordinal, Door.F64, fmt)
+
+    @classmethod
+    def decimal(cls, ordinal: int, fmt: NumFormat = NumFormat.INVARIANT) -> Column:
+        """An exact ``decimal.Decimal`` column under the declared notation; no float is
+        ever formed."""
+        return cls._numeric(ordinal, Door.DECIMAL, fmt)
+
+    @classmethod
+    def uuid(cls, ordinal: int) -> Column:
+        """A ``uuid.UUID`` column."""
+        return cls(ordinal, Door.UUID)
+
+    @classmethod
+    def timestamp(cls, ordinal: int) -> Column:
+        """An RFC 3339 instant column, as an aware UTC ``datetime``."""
+        return cls(ordinal, Door.TIMESTAMP)
+
+    @classmethod
+    def unix(cls, ordinal: int, precision: UnixPrecision) -> Column:
+        """A Unix-epoch column at the declared precision — never guessed from magnitude —
+        as an aware UTC ``datetime``."""
+        return cls(ordinal, Door.UNIX, declared=UnixPrecision(precision))
+
+    @classmethod
+    def excel_serial(cls, ordinal: int, epoch: ExcelEpoch) -> Column:
+        """An Excel date-serial column under the declared date system, as an aware UTC
+        ``datetime``."""
+        return cls(ordinal, Door.EXCEL_SERIAL, declared=ExcelEpoch(epoch))
+
+    @classmethod
+    def date(cls, ordinal: int, order: DateOrder | None = None) -> Column:
+        """A ``date`` column: strict ISO ``yyyy-MM-dd`` with no order, the separated forms
+        under a declared one (which is :meth:`date_ordered`) — as ``hypercast.cast_date``."""
+        if order is not None:
+            return cls.date_ordered(ordinal, order)
+        return cls(ordinal, Door.DATE)
+
+    @classmethod
+    def date_ordered(cls, ordinal: int, order: DateOrder) -> Column:
+        """A separated-date column under the declared field order, as a ``date``."""
+        return cls(ordinal, Door.DATE_ORDERED, declared=DateOrder(order))
+
+    @classmethod
+    def datetime(cls, ordinal: int, order: DateOrder) -> Column:
+        """A zone-less civil date-time column under the declared field order, as a naive
+        ``datetime`` — the text named no zone and none is invented."""
+        return cls(ordinal, Door.DATETIME, declared=DateOrder(order))
+
+    @classmethod
+    def time(cls, ordinal: int) -> Column:
+        """A 24-hour time-of-day column, as a ``time``."""
+        return cls(ordinal, Door.TIME)
+
+    @classmethod
+    def duration(cls, ordinal: int) -> Column:
+        """A duration column, as a ``timedelta``."""
+        return cls(ordinal, Door.DURATION)
+
+    @classmethod
+    def text(cls, ordinal: int) -> Column:
+        """A text column: the cell's bytes themselves, untrimmed, as a ``str``. A cell
+        with no bytes at all is the one way text fails (``CastFailure.EMPTY``)."""
+        return cls(ordinal, Door.TEXT)
+
+
+@dataclass(frozen=True, slots=True)
+class SheetOptions:
+    """How a sheet of a :class:`Workbook` is read."""
+
+    has_header: bool = True
+    """Whether the sheet's first row is a header, exposed through :attr:`Sheet.header` and
+    never delivered as a row."""
+    skip_empty_rows: bool = True
+    """Whether a row with no cells is skipped rather than delivered as a row of empty
+    cells."""
+    batch_rows: int = 4096
+    """Rows per batch, at most."""
+
+    def __post_init__(self) -> None:
+        """Refuses a batch size no batch can have."""
+        if not isinstance(self.batch_rows, int) or isinstance(self.batch_rows, bool):
+            raise TypeError(f"batch_rows must be an int, not {type(self.batch_rows).__name__}")
+        if self.batch_rows < 1:
+            raise ValueError(f"batch_rows must be at least 1, not {self.batch_rows}")
+
+
+# The extension's own classes are the package surface. _bind hands it this package's plan
+# column, dialect, structural failure and workbook format; it imports HyperCast's verdict
+# types itself.
+_native._bind(Column, Dialect, TabularError, TabularFailure, WorkbookFormat)
+
+DelimitedReader = _native.DelimitedReader
+Workbook = _native.Workbook
+Sheet = _native.Sheet
+SheetInfo = _native.SheetInfo
+Batch = _native.Batch
+ColumnData = _native.ColumnData
+native_version = _native.native_version

hypertabular/_native.pyi

@@ -0,0 +1,126 @@
+"""The typed surface of ``hypertabular._native``, the PyO3 extension (``tests/test_typing.py``
+holds the loaded module to this file). ``hypertabular`` re-exports all of it directly, so
+this is what a type checker sees behind a reader, a batch and a column — and behind a cell,
+which is HyperCast's own ``Success[...] | Fault``, so a ``match`` over the two cases can be
+checked for exhaustiveness with ``typing.assert_never``.
+"""
+
+from collections.abc import Iterable, Iterator, Sequence
+from os import PathLike
+from typing import Any, ClassVar, Protocol, final
+
+from hypercast import ExcelEpoch, Fault, Success
+
+from . import Column, Dialect, SheetOptions, TabularError, TabularFailure, WorkbookFormat
+
+class _Readable(Protocol):
+    def read(self, size: int, /) -> bytes: ...
+
+@final
+class ColumnData:
+    @property
+    def column(self) -> Column: ...
+    @property
+    def values(self) -> Sequence[Any]: ...
+    @property
+    def verdicts(self) -> memoryview: ...
+    @property
+    def fault_count(self) -> int: ...
+    def faults(self) -> list[tuple[int, Fault]]: ...
+    def raw(self, row: int) -> bytes: ...
+    def __len__(self) -> int: ...
+    def __getitem__(self, row: int) -> Success[Any] | Fault: ...
+    def __iter__(self) -> Iterator[Success[Any] | Fault]: ...
+
+@final
+class Batch:
+    @property
+    def rows(self) -> int: ...
+    @property
+    def columns(self) -> tuple[ColumnData, ...]: ...
+    def column(self, index: int) -> ColumnData: ...
+    def line(self, row: int) -> int: ...
+    def raw(self, column: int, row: int) -> bytes: ...
+    def __len__(self) -> int: ...
+
+@final
+class DelimitedReader:
+    DEFAULT_BATCH_ROWS: ClassVar[int]
+    DEFAULT_BUFFER_BYTES: ClassVar[int]
+    MAX_ROW_BYTES: ClassVar[int]
+    def __new__(
+        cls,
+        source: bytes | _Readable,
+        dialect: Dialect,
+        plan: Iterable[Column],
+        *,
+        batch_rows: int = 4096,
+        buffer_bytes: int = 262144,
+    ) -> DelimitedReader: ...
+    @staticmethod
+    def open(
+        path: str | PathLike[str],
+        dialect: Dialect,
+        plan: Iterable[Column],
+        *,
+        batch_rows: int = 4096,
+        buffer_bytes: int = 262144,
+    ) -> DelimitedReader: ...
+    @property
+    def header(self) -> tuple[str, ...] | None: ...
+    @property
+    def dialect(self) -> Dialect: ...
+    @property
+    def plan(self) -> tuple[Column, ...]: ...
+    @property
+    def batch_rows(self) -> int: ...
+    @property
+    def records(self) -> int: ...
+    def read(self) -> Batch | None: ...
+    def close(self) -> None: ...
+    def __iter__(self) -> DelimitedReader: ...
+    def __next__(self) -> Batch: ...
+    def __enter__(self) -> DelimitedReader: ...
+    def __exit__(self, *_exc: object) -> None: ...
+
+@final
+class SheetInfo:
+    @property
+    def name(self) -> str: ...
+    @property
+    def hidden(self) -> bool: ...
+
+@final
+class Workbook:
+    def __new__(cls, data: bytes) -> Workbook: ...
+    @staticmethod
+    def open(path: str | PathLike[str]) -> Workbook: ...
+    @property
+    def format(self) -> WorkbookFormat: ...
+    @property
+    def date_system(self) -> ExcelEpoch: ...
+    @property
+    def sheets(self) -> tuple[SheetInfo, ...]: ...
+    def sheet(self, which: int | str, options: SheetOptions, plan: Iterable[Column]) -> Sheet: ...
+
+@final
+class Sheet:
+    @property
+    def options(self) -> SheetOptions: ...
+    @property
+    def plan(self) -> tuple[Column, ...]: ...
+    @property
+    def header(self) -> tuple[str, ...] | None: ...
+    def read(self) -> Batch | None: ...
+    def __iter__(self) -> Sheet: ...
+    def __next__(self) -> Batch: ...
+
+def native_version() -> str: ...
+def _bind(
+    column: type[Column],
+    dialect: type[Dialect],
+    error: type[TabularError],
+    failure: type[TabularFailure],
+    workbook_format: type[WorkbookFormat],
+) -> None: ...
+def _limits(reader: DelimitedReader, max_row_bytes: int, window_bytes: int) -> None: ...