|
1
|
|
|
#!/usr/bin/env python |
|
2
|
|
|
# -*- encoding: utf-8 -*- |
|
3
|
|
|
# vim: set et sw=4 ts=4 sts=4 ff=unix fenc=utf8: |
|
4
|
|
|
# Author: Binux<[email protected]> |
|
5
|
|
|
# http://binux.me |
|
6
|
|
|
# Created on 2014-10-11 23:54:50 |
|
7
|
|
|
|
|
8
|
|
|
import json |
|
9
|
|
|
import time |
|
10
|
|
|
|
|
11
|
|
|
from pymongo import MongoClient |
|
12
|
|
|
|
|
13
|
|
|
from pyspider.database.base.taskdb import TaskDB as BaseTaskDB |
|
14
|
|
|
from .mongodbbase import SplitTableMixin |
|
15
|
|
|
|
|
16
|
|
|
|
|
17
|
|
|
class TaskDB(SplitTableMixin, BaseTaskDB): |
|
18
|
|
|
collection_prefix = '' |
|
19
|
|
|
|
|
20
|
|
|
def __init__(self, url, database='taskdb'): |
|
21
|
|
|
self.conn = MongoClient(url) |
|
22
|
|
|
self.conn.admin.command("ismaster") |
|
23
|
|
|
self.database = self.conn[database] |
|
24
|
|
|
self.projects = set() |
|
25
|
|
|
|
|
26
|
|
|
self._list_project() |
|
27
|
|
|
# we suggest manually build index in advance, instead of indexing |
|
28
|
|
|
# in the startup process, |
|
29
|
|
|
# for project in self.projects: |
|
30
|
|
|
# collection_name = self._collection_name(project) |
|
31
|
|
|
# self.database[collection_name].ensure_index('status') |
|
32
|
|
|
# self.database[collection_name].ensure_index('taskid') |
|
33
|
|
|
|
|
34
|
|
|
def _create_project(self, project): |
|
35
|
|
|
collection_name = self._collection_name(project) |
|
36
|
|
|
self.database[collection_name].ensure_index('status') |
|
37
|
|
View Code Duplication |
self.database[collection_name].ensure_index('taskid') |
|
|
|
|
|
|
38
|
|
|
self._list_project() |
|
39
|
|
|
|
|
40
|
|
|
def _parse(self, data): |
|
41
|
|
|
if '_id' in data: |
|
42
|
|
|
del data['_id'] |
|
43
|
|
|
for each in ('schedule', 'fetch', 'process', 'track'): |
|
44
|
|
|
if each in data: |
|
45
|
|
|
if data[each]: |
|
46
|
|
|
if isinstance(data[each], bytearray): |
|
47
|
|
|
data[each] = str(data[each]) |
|
48
|
|
|
data[each] = json.loads(data[each], encoding='utf8') |
|
49
|
|
|
else: |
|
50
|
|
|
data[each] = {} |
|
51
|
|
|
return data |
|
52
|
|
|
|
|
53
|
|
|
def _stringify(self, data): |
|
54
|
|
|
for each in ('schedule', 'fetch', 'process', 'track'): |
|
55
|
|
|
if each in data: |
|
56
|
|
|
data[each] = json.dumps(data[each]) |
|
57
|
|
|
return data |
|
58
|
|
|
|
|
59
|
|
|
def load_tasks(self, status, project=None, fields=None): |
|
60
|
|
|
if not project: |
|
61
|
|
|
self._list_project() |
|
62
|
|
|
|
|
63
|
|
|
if project: |
|
64
|
|
|
projects = [project, ] |
|
65
|
|
|
else: |
|
66
|
|
|
projects = self.projects |
|
67
|
|
|
|
|
68
|
|
|
for project in projects: |
|
69
|
|
|
collection_name = self._collection_name(project) |
|
70
|
|
|
for task in self.database[collection_name].find({'status': status}, fields): |
|
71
|
|
|
yield self._parse(task) |
|
72
|
|
|
|
|
73
|
|
|
def get_task(self, project, taskid, fields=None): |
|
74
|
|
|
if project not in self.projects: |
|
75
|
|
|
self._list_project() |
|
76
|
|
|
if project not in self.projects: |
|
77
|
|
|
return |
|
78
|
|
|
collection_name = self._collection_name(project) |
|
79
|
|
|
ret = self.database[collection_name].find_one({'taskid': taskid}, fields) |
|
80
|
|
|
if not ret: |
|
81
|
|
|
return ret |
|
82
|
|
|
return self._parse(ret) |
|
83
|
|
|
|
|
84
|
|
|
def status_count(self, project): |
|
85
|
|
|
if project not in self.projects: |
|
86
|
|
|
self._list_project() |
|
87
|
|
|
if project not in self.projects: |
|
88
|
|
|
return {} |
|
89
|
|
|
collection_name = self._collection_name(project) |
|
90
|
|
|
|
|
91
|
|
|
# when there are too many data in task collection , aggregate operation will take a very long time, |
|
92
|
|
|
# and this will cause scheduler module startup to be particularly slow |
|
93
|
|
|
|
|
94
|
|
|
# ret = self.database[collection_name].aggregate([ |
|
95
|
|
|
# {'$group': { |
|
96
|
|
|
# '_id' : '$status', |
|
97
|
|
|
# 'total': { |
|
98
|
|
|
# '$sum': 1 |
|
99
|
|
|
# } |
|
100
|
|
|
# } |
|
101
|
|
|
# }]) |
|
102
|
|
|
|
|
103
|
|
|
# Instead of aggregate, use find-count on status(with index) field. |
|
104
|
|
|
def _count_for_status(collection, status): |
|
105
|
|
|
total = collection.find({'status': status}).count() |
|
106
|
|
|
return {'total': total, "_id": status} if total else None |
|
107
|
|
|
|
|
108
|
|
|
c = self.database[collection_name] |
|
109
|
|
|
ret = filter( |
|
110
|
|
|
lambda x: x, |
|
111
|
|
|
map( |
|
112
|
|
|
lambda s: _count_for_status(c, s), [self.ACTIVE, self.SUCCESS, self.FAILED] |
|
113
|
|
|
) |
|
114
|
|
|
) |
|
115
|
|
|
|
|
116
|
|
|
result = {} |
|
117
|
|
|
if isinstance(ret, dict): |
|
118
|
|
|
ret = ret.get('result', []) |
|
119
|
|
|
for each in ret: |
|
120
|
|
|
result[each['_id']] = each['total'] |
|
121
|
|
|
return result |
|
122
|
|
|
|
|
123
|
|
|
def insert(self, project, taskid, obj={}): |
|
124
|
|
|
if project not in self.projects: |
|
125
|
|
|
self._create_project(project) |
|
126
|
|
|
obj = dict(obj) |
|
127
|
|
|
obj['taskid'] = taskid |
|
128
|
|
|
obj['project'] = project |
|
129
|
|
|
obj['updatetime'] = time.time() |
|
130
|
|
|
return self.update(project, taskid, obj=obj) |
|
131
|
|
|
|
|
132
|
|
|
def update(self, project, taskid, obj={}, **kwargs): |
|
133
|
|
|
obj = dict(obj) |
|
134
|
|
|
obj.update(kwargs) |
|
135
|
|
|
obj['updatetime'] = time.time() |
|
136
|
|
|
collection_name = self._collection_name(project) |
|
137
|
|
|
return self.database[collection_name].update( |
|
138
|
|
|
{'taskid': taskid}, |
|
139
|
|
|
{"$set": self._stringify(obj)}, |
|
140
|
|
|
upsert=True |
|
141
|
|
|
) |
|
142
|
|
|
|
binux /
pyspider
Loading history...