blob: 61155de96b4f64270c51fdd675f33344bb706924 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
|
# -*- coding: utf-8 -*-
"""Utilities for task assignment
"""
from flask import current_app
from app.models import Dataset, Task
def generate_user_task(user):
"""
Generate new task for a given user.
This function assigns tasks to a given user and ensures that:
1) datasets that are nearly annotated with the desired number of
datasets get priority
2) users never are given more tasks than max_per_user
3) users never get the same dataset twice
"""
max_per_user = current_app.config["TASKS_MAX_PER_USER"]
num_per_dataset = current_app.config["TASKS_NUM_PER_DATASET"]
user_tasks = Task.query.filter_by(annotator_id=user.id).all()
user_tasks = [t for t in user_tasks if not t.dataset.is_demo]
# don't assign a new task if the user has assigned tasks
not_done = [t for t in user_tasks if not t.done]
if len(not_done) > 0:
return None
# don't assign a new task if the user has reached maximum
n_user_tasks = len(user_tasks)
if n_user_tasks >= max_per_user:
return None
potential_datasets = []
for dataset in Dataset.query.filter_by(is_demo=False).all():
dataset_tasks = Task.query.filter_by(dataset_id=dataset.id).all()
# check that this dataset needs more annotations
n_needed = num_per_dataset - len(dataset_tasks)
if n_needed <= 0:
continue
# check that this dataset is not already assigned to the user
task = Task.query.filter_by(
dataset_id=dataset.id, annotator_id=user.id
).first()
if not task is None:
continue
potential_datasets.append((n_needed, dataset))
# don't assign a dataset if there are no more datasets to annotate
if len(potential_datasets) == 0:
return None
# sort datasets so that the ones who need the least are at the front.
potential_datasets.sort(key=lambda x: x[0])
_, dataset = potential_datasets[0]
task = Task(annotator_id=user.id, dataset_id=dataset.id)
return task
|
