Artifact

Classes:

`Artifact`([name, cacher])
`ArtifactFilter`([starting_artifacts, ...])
`ArtifactList`([name, artifacts])
`DBArtifact`([name, connection_str])	Artifact that represents a duckdb connection
`StageReportables`()

Functions:

`pointer_based_property`(name)
`pointer_based_property_getter`(self, name)
`pointer_based_property_setter`(self, value, name)

class curifactory.experimental.artifact.Artifact(name=None, cacher=None)

Methods:

`artifact_list`([building_list])	Recursively builds a list of _all_ artifacts prior to this one.
`artifact_list_debug`()
`artifact_tree`()
`check_shared_artifact`(other_artifact)	Two artifacts are considered equivalent (can be shared) if their hash and name is the same
`compute_hash`()
`context_names_minus`(minus)
`copy`()
`dependencies`()	Gets any input artifacts from the compute stage.
`determine_overwrite`()	If any artifacts upstream from this one have overwrite specified, this artifact needs to be overwritten as well.
`filter`(search_str)
`from_cacher`(cacher)
`from_list`(name, artifacts)
`from_metadata`([metadata, path])
`get`()
`get_from_db`()
`load_from_uuid`(uuid[, building_stages, ...])	Recurisvely load the full DAG prior to this artifact and then return this artifact.
`map`([mapped, need, source])
`replace`(artifact)
`reset_map`()
`verify`()
`visualize`([g])

Attributes:

`artifacts`
`cache_status`	Integer (unofficial enum) referring to whether this artifact is in cache or doesn't even have a cacher.
`cacher`	Cacher instance/strategy used to save/load the computed object.
`compute`	The `Stage` object that outputs this artifact.
`computed`	A boolean representing whether the compute stage that outputs this artifact has run or not.
`context`	The `Pipeline` that "owns" this artifact.
`context_name`
`contextualized_name`
`db_id`	UUID of this artifact within the store database.
`generated_time`	Timestamp of when the compute stage outputting this artifact ran.
`hash_debug`	A dictionary of parameters/parameter values that results in the hash for this artifact.
`hash_str`	The string of the numerical hash encoding the parameters for the stage that produces this artifact.
`internal_id`	Refers to the `id()` of the underlying Artifact object.
`map_status`	Integer (unofficial enum) referring to what will be done with this artifact in a run (e.g. compute, skip, use cache, or overwrite, etc.
`name`	Name of the artifact
`obj`	The computed object itself - this is populated either by running the compute stage or by loading it with the cacher if previously computed.
`overwrite`	Boolean flag for whether to force this artifact's compute stage to run (ignoring cached values).
`previous_context_names`	Any previous pipeline names through which this artifact has been passed.

artifact_list(building_list=None)

Recursively builds a list of _all_ artifacts prior to this one.

Parameters:: building_list (list)

artifact_list_debug()

artifact_tree()

property artifacts

property cache_status: Integer (unofficial enum) referring to whether this artifact is in cache or doesn’t even have a cacher. See __init__.py for possible statuses.

property cacher: Cacher instance/strategy used to save/load the computed object.

check_shared_artifact(other_artifact): Two artifacts are considered equivalent (can be shared) if their hash and name is the same

property compute: The Stage object that outputs this artifact.

compute_hash()

property computed: A boolean representing whether the compute stage that outputs this artifact has run or not.

property context: The Pipeline that “owns” this artifact.

property context_name

context_names_minus(minus)

Parameters:: minus (str)

property contextualized_name

copy()

property db_id: UUID of this artifact within the store database.

dependencies()

Gets any input artifacts from the compute stage.

Return type:: list[Artifact]

determine_overwrite()

If any artifacts upstream from this one have overwrite specified, this artifact needs to be overwritten as well.

Return type:: bool

filter(search_str)

Parameters:: search_str (str)
Return type:: ArtifactFilter

static from_cacher(cacher)

static from_list(name, artifacts)

static from_metadata(metadata=None, path=None)

property generated_time: Timestamp of when the compute stage outputting this artifact ran.

get()

get_from_db()

property hash_debug: A dictionary of parameters/parameter values that results in the hash for this artifact.

property hash_str: The string of the numerical hash encoding the parameters for the stage that produces this artifact.

property internal_id: Refers to the id() of the underlying Artifact object. Since Artifacts are also loosely pointers, can use this to determine which underlying object is being pointed to.

static load_from_uuid(uuid, building_stages=None, building_artifacts=None)

Recurisvely load the full DAG prior to this artifact and then return this artifact.

Parameters:

building_stages (dict)
building_artifacts (dict)

map(mapped=None, need=True, source=None)

Parameters:

mapped (dict)
need (bool)

property map_status: Integer (unofficial enum) referring to what will be done with this artifact in a run (e.g. compute, skip, use cache, or overwrite, etc. See __init__.py for possible statuses.

property name: Name of the artifact

property obj: The computed object itself - this is populated either by running the compute stage or by loading it with the cacher if previously computed.

property overwrite: Boolean flag for whether to force this artifact’s compute stage to run (ignoring cached values). Setting this to True impacts all downstream artifacts.

property previous_context_names: Any previous pipeline names through which this artifact has been passed. When a pipeline is passed into a new pipeline, or a previous pipeline’s artifact is copied into a new one, that previous pipeline name is retained here.

replace(artifact)

reset_map()

verify()

visualize(g=None, **kwargs)

class curifactory.experimental.artifact.ArtifactFilter(starting_artifacts=None, filter_string='')

Methods:

`copy`()
`filter`(search_str)
`list`()	TODO: maybe this should return ArtifactList instead?
`replace`(new_artifact)
`resolve`()

copy()

filter(search_str)

Parameters:: search_str (str)
Return type:: ArtifactFilter

list(): TODO: maybe this should return ArtifactList instead?

replace(new_artifact)

resolve()

Return type:: Artifact

class curifactory.experimental.artifact.ArtifactList(name=None, artifacts=None)

Methods:

append(value)

Parameters:: name (str)

append(value)

class curifactory.experimental.artifact.DBArtifact(name=None, connection_str=None, **kwargs)

Artifact that represents a duckdb connection

Methods:

compute_hash()

Parameters:

name (str)
connection_str (str)

compute_hash()

class curifactory.experimental.artifact.StageReportables

curifactory.experimental.artifact.pointer_based_property(name)

curifactory.experimental.artifact.pointer_based_property_getter(self, name)

curifactory.experimental.artifact.pointer_based_property_setter(self, value, name)