cache

Cache suitable for shards, not allowed to change because they are named after their own sha256 hash.

Classes

AnnotatedRawShard
ShardCache	Handle caching for individual shards (not the index of shards).

Functions

connect([dburi])

Get database connection.

Attributes

SHARD_CACHE_NAME
ZSTD_MAX_SHARD_SIZE

SHARD_CACHE_NAME = 'repodata_shards.db'

ZSTD_MAX_SHARD_SIZE = 16777216

class AnnotatedRawShard(url: str, package: str, compressed_shard: bytes)

url

package

compressed_shard

connect(dburi='cache.db')

Get database connection.

dburi: uri-style sqlite database filename; accepts certain ?= parameters.

class ShardCache(base: pathlib.Path, create=True)

Handle caching for individual shards (not the index of shards).

base: directory and filename prefix for cache.

base

__enter__()

__exit__(exc_type, exc_val, exec_tb)

close()

Clean up connection. ShardCache can no longer be used after close().

copy()

Copy cache with new connection. Useful for threads.

connect(create=True, retry=True)

Parameters:

• create -- if True, create table if not exists.

• retry -- remove cache, log warning, and retry on error.

insert(raw_shard: AnnotatedRawShard)

Parameters:

• url -- of shard

• package -- package name

• raw_shard -- msgpack.zst compressed shard data

retrieve(url) → conda._private.shards.typing.ShardDict | None

retrieve_multiple(urls: list[str]) → dict[str, conda._private.shards.typing.ShardDict | None]

Query database for cached shard urls.

Return a dict of urls in cache mapping to the Shard or None if not present.

clear_cache()

Truncate the database by removing all rows from tables

remove_cache()

Remove the sharded cache database.