Skip to content

External information sources #7

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 8 commits into from
Sep 13, 2022
Merged
Show file tree
Hide file tree
Changes from 5 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 11 additions & 0 deletions business_objects/attribute.py
Original file line number Diff line number Diff line change
Expand Up @@ -39,6 +39,17 @@ def get_by_name(project_id: str, name: str) -> Attribute:
)


def get_all_by_names(project_id: str, attribute_names: List[str]) -> List[Attribute]:
return (
session.query(Attribute)
.filter(
Attribute.project_id == project_id,
Attribute.name.in_(attribute_names),
)
.all()
)


def get_all(project_id: str) -> List[Attribute]:
return session.query(Attribute).filter(Attribute.project_id == project_id).all()

Expand Down
236 changes: 236 additions & 0 deletions business_objects/embedder.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,236 @@
from datetime import datetime
from typing import Dict, List, Any, Optional

from . import general
from ..models import (
Embedder,
EmbedderPayload,
)
from ..session import session


def get(project_id: str, source_id: str) -> Embedder:
return (
session.query(Embedder)
.filter(
Embedder.project_id == project_id,
Embedder.id == source_id,
)
.first()
)


def get_all(project_id: str) -> List[Embedder]:
return (
session.query(Embedder)
.filter(
Embedder.project_id == project_id,
)
.all()
)


def get_payload(project_id: str, payload_id: str) -> EmbedderPayload:
return (
session.query(EmbedderPayload)
.filter(
EmbedderPayload.project_id == project_id,
EmbedderPayload.id == payload_id,
)
.first()
)


def get_last_payload(project_id: str, source_id: str) -> EmbedderPayload:
return (
session.query(EmbedderPayload)
.filter(
EmbedderPayload.project_id == project_id,
EmbedderPayload.embedder_id == source_id,
)
.order_by(EmbedderPayload.created_at.desc())
.first()
)


def get_payloads_by_project_id(project_id: str) -> List[Any]:
query: str = f"""
SELECT
payload.id,
payload.embedder_id,
payload.created_at,
payload.finished_at,
payload.iteration,
payload.source_code,
payload.logs,
payload.state
FROM
embedder_payload AS payload
INNER JOIN
embedder
ON
payload.embedder_id=embedder.id
WHERE
embedder.project_id='{project_id}'
;
"""
return general.execute_all(query)


def get_overview_data(project_id: str) -> List[Dict[str, Any]]:
query = f"""
SELECT array_agg(row_to_json(data_select))
FROM (
SELECT
_is.id,
_is.name,
_is.type "embedderType",
_is.description,
_is.created_at "createdAt",
_is.created_by "createdBy",
isp.state,
isp.created_at "lastRun",
FROM embedder _is
LEFT JOIN LATERAL(
SELECT isp.id,isp.state,isp.created_at
FROM embedder_payload isp
WHERE _is.id = isp.embedder_id
AND _is.project_id = isp.project_id
ORDER BY isp.iteration DESC
LIMIT 1
) isp ON TRUE
WHERE _is.project_id = '{project_id}'
ORDER BY "createdAt" DESC,name
)data_select """
values = general.execute_first(query)

if values:
return values[0]


def continue_payload(project_id: str, embedder_id: str, payload_id: str) -> bool:
query = f"""
SELECT isp.state
FROM embedder_payload isp
INNER JOIN embedder _is
ON isp.embedder_id = _is.id AND isp.project_id = _is.project_id
WHERE isp.id = '{payload_id}'
AND isp.source_id = '{embedder_id}'
AND isp.project_id = '{project_id}' """

value = general.execute_first(query)
if not value or value[0] != "CREATED":
return False
return True


def create(
project_id: str,
name: str,
type: str,
description: str,
source_code: str,
version: Optional[int] = None,
created_at: Optional[datetime] = None,
created_by: Optional[str] = None,
with_commit: bool = False,
) -> Embedder:
embedder: Embedder = Embedder(
project_id=project_id,
name=name,
type=type,
description=description,
source_code=source_code,
version=version,
created_at=created_at,
created_by=created_by,
)
general.add(embedder, with_commit)
return embedder


def create_payload(
project_id: str,
embedder_id: str,
state: str,
created_by: Optional[str] = None,
created_at: Optional[datetime] = None,
finished_at: Optional[datetime] = None,
iteration: Optional[int] = None,
source_code: Optional[str] = None,
logs: List[str] = None,
with_commit: bool = False,
) -> EmbedderPayload:
payload: EmbedderPayload = EmbedderPayload(
embedder_id=embedder_id, project_id=project_id, state=state
)
if created_by:
payload.created_by = created_by
if iteration:
payload.iteration = iteration
if source_code:
payload.source_code = source_code
if created_at:
payload.created_at = created_at
if finished_at:
payload.finished_at = finished_at
if logs:
payload.logs = logs
general.add(payload, with_commit)
return payload


def delete(project_id: str, embedder_id: str, with_commit: bool = False) -> None:
session.query(Embedder).filter(
Embedder.project_id == project_id,
Embedder.id == embedder_id,
).delete()
general.flush_or_commit(with_commit)


def update(
project_id: str,
embedder_id: str,
name: Optional[str] = None,
type: Optional[str] = None,
description: Optional[str] = None,
source_code: Optional[str] = None,
version: Optional[int] = None,
created_at: Optional[datetime] = None,
created_by: Optional[str] = None,
with_commit: bool = False,
) -> None:
information_source = get(project_id, embedder_id)

if name is not None:
information_source.name = name
if type is not None:
information_source.type = type
if description is not None:
information_source.description = description
if source_code is not None:
information_source.source_code = source_code
if version is not None:
information_source.version = version
if created_at is not None:
information_source.created_at = created_at
if created_by is not None:
information_source.created_at = created_by
general.flush_or_commit(with_commit)


def update_payload(
project_id: str,
payload_id: str,
state: str = None,
progress: float = None,
with_commit: bool = False,
) -> None:
payload_item = get_payload(project_id, payload_id)
if not payload_item:
return
if state:
payload_item.state = state
if progress:
payload_item.progress = progress
general.flush_or_commit(with_commit)
42 changes: 42 additions & 0 deletions business_objects/embedder_payload.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,42 @@
from datetime import datetime
from typing import List, Optional

from . import general
from .. import enums
from ..models import Embedder, EmbedderPayload
from ..session import session


def get(project_id: str, payload_id: str) -> EmbedderPayload:
return (
session.query(EmbedderPayload)
.filter(
Embedder.project_id == project_id,
Embedder.id == EmbedderPayload.embedder_id,
EmbedderPayload.id == payload_id,
)
.first()
)


def create(
project_id: str,
source_code: str,
state: enums.PayloadState,
iteration: int,
embedder_id: str,
created_by: str,
created_at: Optional[datetime] = None,
with_commit: bool = False,
) -> EmbedderPayload:
payload: EmbedderPayload = EmbedderPayload(
source_code=source_code,
state=state.value,
iteration=iteration,
embedder_id=embedder_id,
created_by=created_by,
project_id=project_id,
created_at=created_at,
)
general.add(payload, with_commit)
return payload
28 changes: 23 additions & 5 deletions business_objects/information_source.py
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,17 @@ def get(project_id: str, source_id: str) -> InformationSource:
)


def get_by_name(project_id: str, name: str) -> InformationSource:
return (
session.query(InformationSource)
.filter(
InformationSource.project_id == project_id,
InformationSource.name == name,
)
.first()
)


def get_all(project_id: str) -> List[InformationSource]:
return (
session.query(InformationSource)
Expand Down Expand Up @@ -168,7 +179,7 @@ def get_exclusion_record_ids_for_task(task_id: str) -> List[str]:
return exclusion_ids


def get_overview_data(project_id: str) -> List[Dict[str, Any]]:
def get_overview_data(project_id: str, operator: str) -> List[Dict[str, Any]]:
query = f"""
SELECT array_agg(row_to_json(data_select))
FROM (
Expand Down Expand Up @@ -205,14 +216,16 @@ def get_overview_data(project_id: str) -> List[Dict[str, Any]]:
GROUP BY source_id) stats
ON _is.id = stats.source_id
WHERE _is.project_id = '{project_id}'
AND _is.type {operator} 'MODEL_CALLBACK'
ORDER BY "createdAt" DESC,name
)data_select """
values = general.execute_first(query)

if values:
return values[0]

def continue_payload(project_id:str,source_id:str,payload_id:str)->bool:

def continue_payload(project_id: str, source_id: str, payload_id: str) -> bool:
query = f"""
SELECT isp.state
FROM information_source_payload isp
Expand All @@ -223,7 +236,7 @@ def continue_payload(project_id:str,source_id:str,payload_id:str)->bool:
AND isp.project_id = '{project_id}' """

value = general.execute_first(query)
if not value or value[0]!= "CREATED":
if not value or value[0] != "CREATED":
return False
return True

Expand Down Expand Up @@ -444,7 +457,10 @@ def update_quantity_stats(


def update_is_selected_for_project(
project_id: str, update_value: bool, with_commit: bool = False
project_id: str,
update_value: bool,
with_commit: bool = False,
operator: Optional[str] = "!=",
) -> None:
if update_value:
str_value = "TRUE"
Expand All @@ -454,7 +470,9 @@ def update_is_selected_for_project(
query = f"""
UPDATE information_source
SET is_selected = {str_value}
WHERE project_id = '{project_id}' """
WHERE project_id = '{project_id}'
AND type {operator} 'MODEL_CALLBACK'
"""
general.execute(query)
general.flush_or_commit(with_commit)

Expand Down
8 changes: 8 additions & 0 deletions business_objects/labeling_task.py
Original file line number Diff line number Diff line change
Expand Up @@ -32,6 +32,14 @@ def get_task_name_id_dict(project_id: str) -> Dict[str, str]:
return {labeling_task.name: labeling_task.id for labeling_task in labeling_tasks}


def get_labeling_task_by_name(project_id: str, task_name: str) -> LabelingTask:
return (
session.query(LabelingTask)
.filter(LabelingTask.project_id == project_id, LabelingTask.name == task_name)
.first()
)


def get_labeling_tasks_by_selected_sources(project_id: str) -> List[LabelingTask]:
return (
session.query(LabelingTask)
Expand Down
Loading