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
181
182
183
184
185
186
187
188
189
190
191
|
import os
from bsddb3 import db
import Location
import BaseDB
import Bio
_open = open # rename for internal use -- gets redefined below
INDEX_TYPE = "BerkeleyDB/1"
def create(dbname, primary_namespace, secondary_namespaces,
formatname = "unknown"):
os.mkdir(dbname)
config_filename = os.path.join(dbname, "config.dat")
BaseDB.write_config(config_filename = config_filename,
index_type = INDEX_TYPE,
primary_namespace = primary_namespace,
secondary_namespaces = secondary_namespaces,
fileid_info = {},
formatname = formatname
)
dbenv = db.DBEnv(0)
envflags = db.DB_THREAD | db.DB_INIT_MPOOL
dbenv.open(dbname, envflags | db.DB_CREATE)
primary_table = db.DB(dbenv)
primary_table.open("key_%s" % (primary_namespace,), None,
db.DB_BTREE, db.DB_CREATE, 0660)
secondary_tables = {}
for namespace in secondary_namespaces:
x = db.DB(dbenv)
x.open("id_%s" % (namespace,), None, db.DB_BTREE, db.DB_CREATE, 0)
secondary_tables[namespace] = x
for x in secondary_tables.values():
x.close()
primary_table.close()
dbenv.close()
return open(dbname, "rw")
class PrimaryNamespace(BaseDB.DictLookup):
def __init__(self, db, namespace):
self.db = db
self.namespace = namespace
assert namespace == db.primary_namespace
def __getitem__(self, name):
loc = self.db.primary_table[name]
filetag, startpos, length = loc.split("\t")
filename = self.db.fileid_info[filetag][0]
return [
Location.Location(self.namespace,
name,
filename,
long(startpos),
long(length))
]
def keys(self):
return self.db.primary_table.keys()
class SecondaryNamespace(BaseDB.DictLookup):
def __init__(self, db, namespace):
self.db = db
self.namespace = namespace
assert namespace in db.secondary_namespaces
def __getitem__(self, name):
table = self.db._load_namespace(self.namespace)
text = table.get(name, None)
if text is None:
raise KeyError("Cannot find %r key %r" % (self.namespace, name))
data = []
for key in text.split("\t"):
loc = self.db.primary_table[key]
filetag, start, length = loc.split("\t")
filename = self.db.fileid_info[filetag][0]
data.append(Location.Location(self.namespace,
name,
filename,
long(start),
long(length)))
return data
def keys(self):
table = self.db._load_namespace(self.namespace)
return table.keys()
class BerkeleyDB(BaseDB.OpenDB, BaseDB.WriteDB):
def __init__(self, dbname, mode = "r"):
if mode not in ("r", "rw"):
raise TypeError("Unknown mode: %r" % (mode,))
self.__need_flush = 0
BaseDB.OpenDB.__init__(self, dbname, INDEX_TYPE)
self.dbenv = None
dbenv = db.DBEnv()
envflags = db.DB_THREAD | db.DB_INIT_MPOOL
dbenv.open(dbname, envflags)
if mode == "r":
self._dbopen_flags = db.DB_RDONLY
else:
self._dbopen_flags = 0
self.primary_table = db.DB(dbenv)
self.primary_table.open("key_%s" % (self.primary_namespace,),
None,
db.DB_BTREE, self._dbopen_flags, 0660)
self.secondary_tables = {}
self.dbenv = dbenv
def _load_namespace(self, namespace):
dbname = "id_%s" % namespace
# Get the appropriate lookup table
if not self.secondary_tables.has_key(namespace):
# Nope, so load it up
self.secondary_tables[namespace] = db.DB(self.dbenv)
self.secondary_tables[namespace].open(dbname, None,
db.DB_BTREE,
self._dbopen_flags, 0)
return self.secondary_tables[namespace]
def add_record(self, filetag, startpos, length, table):
key_list = table[self.primary_namespace]
if len(key_list) != 1:
raise TypeError(
"Field %s has %d entries but must have only one "
"(must be unique)" % (repr(self.primary_namespace),
len(key_list)))
key = key_list[0]
if self.primary_table.has_key(key):
raise TypeError("Field %r = %r already exists" %
(self.primary_namespace, key))
self.primary_table[key] = "%s\t%s\t%s" % (filetag,
BaseDB._int_str(startpos),
BaseDB._int_str(length))
for namespace in self.secondary_namespaces:
lookup = self._load_namespace(namespace)
# Get the list of secondary identifiers for this identifier
for val in table.get(namespace, ()):
# Go from secondary identifier to list of primary identifiers
if lookup.has_key(val):
lookup[val] = lookup[val] + "\t" + key
else:
lookup[val] = key
self.__need_flush = 1
def flush(self):
if not self.__need_flush:
return
config_filename = os.path.join(self.dbname, "config.dat")
BaseDB.write_config(config_filename = config_filename,
index_type = INDEX_TYPE,
primary_namespace = self.primary_namespace,
secondary_namespaces =
self.secondary_tables.keys(),
fileid_info = self.fileid_info,
formatname = self.formatname,
)
self.__need_flush = 0
def close(self):
self.flush()
self.primary_table.close()
[x.close() for x in self.secondary_tables.values()]
self.dbenv.close()
self.dbenv = self.primary_table = self.fileid_info = \
self.secondary_tables = self.fileid_info = None
def __del__(self):
if self.dbenv is not None:
self.close()
def __getitem__(self, key):
if key not in self.keys():
raise KeyError(key)
if key == self.primary_namespace:
return PrimaryNamespace(self, key)
else:
return SecondaryNamespace(self, key)
open = BerkeleyDB
|
