AddressReader#

class lsst.pipe.base.quantum_graph.AddressReader(stream: ~typing.IO[bytes], int_size: int, n_rows: int, n_addresses: int, rows_per_page: int, rows: dict[~uuid.UUID, ~lsst.pipe.base.quantum_graph._multiblock.AddressRow] = <factory>, pages: list[~lsst.pipe.base.quantum_graph._multiblock.AddressPage] = <factory>, page_bounds: dict[int, ~lsst.pipe.base.quantum_graph._multiblock.PageBounds] = <factory>)#

Bases: object

A helper object for reading address files for multi-block files.

Attributes Summary

`int_size`	Size of each integer in bytes.
`n_addresses`	Number of addresses in each row.
`n_rows`	Number of rows in the file.
`page_bounds`	Mapping from page index to page boundary information.
`pages`	Descriptions of the file offsets and integer row indexes of pages and flags for whether they have been read already.
`row_size`	The size (in bytes) of each row of this address file.
`rows`	Rows that have already been read.
`rows_per_page`	Number of addresses in each page.
`stream`	Stream to read from.

Methods Summary

`compute_header_size`(int_size)	Return the size (in bytes) of the header of an address file.
`compute_row_size`(int_size, n_addresses)	Return the size (in bytes) of each row of an address file.
`find`(key)	Read the row for the given UUID or integer index.
`from_stream`(stream, *, page_size, ...)	Construct from a stream by reading the header.
`open_in_zip`(zf, name, *, page_size, ...)	Make a reader for an address file in a zip archive.
`read_all`()	Read all addresses in the file.

Attributes Documentation

int_size: int = <dataclasses._MISSING_TYPE object>#: Size of each integer in bytes.

n_addresses: int = <dataclasses._MISSING_TYPE object>#: Number of addresses in each row.

n_rows: int = <dataclasses._MISSING_TYPE object>#: Number of rows in the file.

page_bounds: dict[int, PageBounds] = <dataclasses._MISSING_TYPE object>#: Mapping from page index to page boundary information.

pages: list[AddressPage] = <dataclasses._MISSING_TYPE object>#: Descriptions of the file offsets and integer row indexes of pages and flags for whether they have been read already.

row_size#: The size (in bytes) of each row of this address file.

rows: dict[UUID, AddressRow] = <dataclasses._MISSING_TYPE object>#: Rows that have already been read.

rows_per_page: int = <dataclasses._MISSING_TYPE object>#: Number of addresses in each page.

stream: IO[bytes] = <dataclasses._MISSING_TYPE object>#: Stream to read from.

Methods Documentation

static compute_header_size(int_size: int) → int#

Return the size (in bytes) of the header of an address file.

Parameters#

int_sizeint: Size of each integer in bytes.

Returns#

sizeint: Size of the header in bytes.

static compute_row_size(int_size: int, n_addresses: int) → int#

Return the size (in bytes) of each row of an address file.

Parameters#

int_sizeint: Size of each integer in bytes.
n_addressesint: Number of addresses in each row.

Returns#

sizeint: Size of each row in bytes.

find(key: UUID) → AddressRow#

Read the row for the given UUID or integer index.

Parameters#

keyuuid.UUID: UUID to find.

Returns#

rowAddressRow: Addresses for the given UUID.

classmethod from_stream(stream: IO[bytes], *, page_size: int, n_addresses: int, int_size: int) → AddressReader#

Construct from a stream by reading the header.

Parameters#

streamtyping.IO [ bytes ]: File-like object to read from.
page_sizeint: Approximate number of bytes to read at a time when searching for an address.
n_addressesint: Number of addresses to expect per row. This is checked against the size embedded in the file.
int_sizeint: Number of bytes to use for all integers. This is checked against the size embedded in the file.

classmethod open_in_zip(zf: zipfile.ZipFile, name: str, *, page_size: int, n_addresses: int, int_size: int) → Iterator[AddressReader]#

Make a reader for an address file in a zip archive.

Parameters#

zfzipfile.ZipFile: Zip archive to read the file from.
namestr: Base name for the address file; an extension will be added.
page_sizeint: Approximate number of bytes to read at a time when searching for an address.
n_addressesint: Number of addresses to expect per row. This is checked against the size embedded in the file.
int_sizeint: Number of bytes to use for all integers. This is checked against the size embedded in the file.

Returns#

readercontextlib.AbstractContextManager [ AddressReader ]: Context manager that returns a reader when entered.

read_all() → dict[UUID, AddressRow]#

Read all addresses in the file.

Returns#

rowsdict [ uuid.UUID, AddressRow ]: Mapping of loaded address rows, keyed by UUID.