File size: 6,893 Bytes
aeb6dbc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4d0b8a7
aeb6dbc
 
 
 
 
 
 
 
 
 
 
 
 
 
4d0a7c7
aeb6dbc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b08b226
 
 
 
 
 
 
 
 
 
aeb6dbc
b08b226
 
 
aeb6dbc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0622917
aeb6dbc
 
 
 
 
 
 
 
4d0b8a7
aeb6dbc
 
 
 
 
0622917
 
 
aeb6dbc
 
 
 
 
 
 
 
 
 
 
4d0b8a7
0622917
aeb6dbc
 
 
 
 
 
 
 
 
 
 
 
 
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
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
#
#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
#  Licensed under the Apache License, Version 2.0 (the "License");
#  you may not use this file except in compliance with the License.
#  You may obtain a copy of the License at
#
#      http://www.apache.org/licenses/LICENSE-2.0
#
#  Unless required by applicable law or agreed to in writing, software
#  distributed under the License is distributed on an "AS IS" BASIS,
#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
#  See the License for the specific language governing permissions and
#  limitations under the License.
#
import os
import random

from api.db.db_utils import bulk_insert_into_db
from deepdoc.parser import PdfParser
from peewee import JOIN
from api.db.db_models import DB, File2Document, File
from api.db import StatusEnum, FileType, TaskStatus
from api.db.db_models import Task, Document, Knowledgebase, Tenant
from api.db.services.common_service import CommonService
from api.db.services.document_service import DocumentService
from api.utils import current_timestamp, get_uuid
from deepdoc.parser.excel_parser import RAGFlowExcelParser
from rag.settings import SVR_QUEUE_NAME
from rag.utils.storage_factory import STORAGE_IMPL
from rag.utils.redis_conn import REDIS_CONN


class TaskService(CommonService):
    model = Task

    @classmethod
    @DB.connection_context()
    def get_tasks(cls, task_id):
        fields = [
            cls.model.id,
            cls.model.doc_id,
            cls.model.from_page,
            cls.model.to_page,
            cls.model.retry_count,
            Document.kb_id,
            Document.parser_id,
            Document.parser_config,
            Document.name,
            Document.type,
            Document.location,
            Document.size,
            Knowledgebase.tenant_id,
            Knowledgebase.language,
            Knowledgebase.embd_id,
            Tenant.img2txt_id,
            Tenant.asr_id,
            Tenant.llm_id,
            cls.model.update_time]
        docs = cls.model.select(*fields) \
            .join(Document, on=(cls.model.doc_id == Document.id)) \
            .join(Knowledgebase, on=(Document.kb_id == Knowledgebase.id)) \
            .join(Tenant, on=(Knowledgebase.tenant_id == Tenant.id)) \
            .where(cls.model.id == task_id)
        docs = list(docs.dicts())
        if not docs: return []

        msg = "\nTask has been received."
        prog = random.random() / 10.
        if docs[0]["retry_count"] >= 3:
            msg = "\nERROR: Task is abandoned after 3 times attempts."
            prog = -1

        cls.model.update(progress_msg=cls.model.progress_msg + msg,
                         progress=prog,
                         retry_count=docs[0]["retry_count"]+1
                         ).where(
            cls.model.id == docs[0]["id"]).execute()

        if docs[0]["retry_count"] >= 3: return []

        return docs

    @classmethod
    @DB.connection_context()
    def get_ongoing_doc_name(cls):
        with DB.lock("get_task", -1):
            docs = cls.model.select(*[Document.id, Document.kb_id, Document.location, File.parent_id]) \
                .join(Document, on=(cls.model.doc_id == Document.id)) \
                .join(File2Document, on=(File2Document.document_id == Document.id), join_type=JOIN.LEFT_OUTER) \
                .join(File, on=(File2Document.file_id == File.id), join_type=JOIN.LEFT_OUTER) \
                .where(
                    Document.status == StatusEnum.VALID.value,
                    Document.run == TaskStatus.RUNNING.value,
                    ~(Document.type == FileType.VIRTUAL.value),
                    cls.model.progress < 1,
                    cls.model.create_time >= current_timestamp() - 1000 * 600
                )
            docs = list(docs.dicts())
            if not docs: return []

            return list(set([(d["parent_id"] if d["parent_id"] else d["kb_id"], d["location"]) for d in docs]))

    @classmethod
    @DB.connection_context()
    def do_cancel(cls, id):
        try:
            task = cls.model.get_by_id(id)
            _, doc = DocumentService.get_by_id(task.doc_id)
            return doc.run == TaskStatus.CANCEL.value or doc.progress < 0
        except Exception as e:
            pass
        return False

    @classmethod
    @DB.connection_context()
    def update_progress(cls, id, info):
        if os.environ.get("MACOS"):
            if info["progress_msg"]:
                cls.model.update(progress_msg=cls.model.progress_msg + "\n" + info["progress_msg"]).where(
                    cls.model.id == id).execute()
            if "progress" in info:
                cls.model.update(progress=info["progress"]).where(
                    cls.model.id == id).execute()
            return

        with DB.lock("update_progress", -1):
            if info["progress_msg"]:
                cls.model.update(progress_msg=cls.model.progress_msg + "\n" + info["progress_msg"]).where(
                    cls.model.id == id).execute()
            if "progress" in info:
                cls.model.update(progress=info["progress"]).where(
                    cls.model.id == id).execute()


def queue_tasks(doc: dict, bucket: str, name: str):
    def new_task():
        return {
            "id": get_uuid(),
            "doc_id": doc["id"]
        }
    tsks = []

    if doc["type"] == FileType.PDF.value:
        file_bin = STORAGE_IMPL.get(bucket, name)
        do_layout = doc["parser_config"].get("layout_recognize", True)
        pages = PdfParser.total_page_number(doc["name"], file_bin)
        page_size = doc["parser_config"].get("task_page_size", 12)
        if doc["parser_id"] == "paper":
            page_size = doc["parser_config"].get("task_page_size", 22)
        if doc["parser_id"] in ["one", "knowledge_graph"] or not do_layout:
            page_size = 10 ** 9
        page_ranges = doc["parser_config"].get("pages") or [(1, 10 ** 5)]
        for s, e in page_ranges:
            s -= 1
            s = max(0, s)
            e = min(e - 1, pages)
            for p in range(s, e, page_size):
                task = new_task()
                task["from_page"] = p
                task["to_page"] = min(p + page_size, e)
                tsks.append(task)

    elif doc["parser_id"] == "table":
        file_bin = STORAGE_IMPL.get(bucket, name)
        rn = RAGFlowExcelParser.row_number(doc["name"], file_bin)
        for i in range(0, rn, 3000):
            task = new_task()
            task["from_page"] = i
            task["to_page"] = min(i + 3000, rn)
            tsks.append(task)
    else:
        tsks.append(new_task())

    bulk_insert_into_db(Task, tsks, True)
    DocumentService.begin2parse(doc["id"])

    for t in tsks:
        assert REDIS_CONN.queue_product(SVR_QUEUE_NAME, message=t), "Can't access Redis. Please check the Redis' status."