blob: aba41df912470b0710356202c58d33c462ea5604 (
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
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
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
|
;; Copyright (c) Jeffrey Straszheim. All rights reserved. The use and
;; distribution terms for this software are covered by the Eclipse Public
;; License 1.0 (http://opensource.org/licenses/eclipse-1.0.php) which can
;; be found in the file epl-v10.html at the root of this distribution. By
;; using this software in any fashion, you are agreeing to be bound by the
;; terms of this license. You must not remove this notice, or any other,
;; from this software.
;;
;; database.clj
;;
;; A Clojure implementation of Datalog -- Support for in-memory database
;;
;; straszheimjeffrey (gmail)
;; Created 21 Feburary 2009
(ns clojure.contrib.datalog.database
(:use clojure.contrib.datalog.util)
(:use clojure.contrib.def)
(:use [clojure.set :only (union intersection difference)])
(:use [clojure.contrib.except :only (throwf)]))
(defstruct relation
:schema ; A set of key names
:data ; A set of tuples
:indexes) ; A map key names to indexes (in turn a map of value to tuples)
;;; DDL
(defmethod print-method ::datalog-database
[db #^Writer writer]
(binding [*out* writer]
(do
(println "(datalog-database")
(println "{")
(doseq [key (keys db)]
(println)
(println key)
(print-method (db key) writer))
(println "})"))))
(defn datalog-database
[rels]
(with-meta rels {:type ::datalog-database}))
(def empty-database (datalog-database {}))
(defmethod print-method ::datalog-relation
[rel #^Writer writer]
(binding [*out* writer]
(do
(println "(datalog-relation")
(println " ;; Schema")
(println " " (:schema rel))
(println)
(println " ;; Data")
(println " #{")
(doseq [tuple (:data rel)]
(println " " tuple))
(println " }")
(println)
(println " ;; Indexes")
(println " {")
(doseq [key (-> rel :indexes keys)]
(println " " key)
(println " {")
(doseq [val (keys ((:indexes rel) key))]
(println " " val)
(println " " (get-in rel [:indexes key val])))
(println " }"))
(println " })"))))
(defn datalog-relation
"Creates a relation"
[schema data indexes]
(with-meta (struct relation schema data indexes) {:type ::datalog-relation}))
(defn add-relation
"Adds a relation to the database"
[db name keys]
(assoc db name (datalog-relation (set keys) #{} {})))
(defn add-index
"Adds an index to an empty relation named name"
[db name key]
(assert (empty? (:data (db name))))
(let [rel (db name)
inx (assoc (:indexes rel) key {})]
(assoc db name (datalog-relation (:schema rel)
(:data rel)
inx))))
(defn ensure-relation
"If the database lacks the named relation, add it"
[db name keys indexes]
(if-let [rel (db name)]
(do
(assert (= (:schema rel) (set keys)))
db)
(let [db1 (add-relation db name keys)]
(reduce (fn [db key] (add-index db name key))
db1
indexes))))
(defmacro make-database
"Makes a database, like this
(make-database
(relation :fred [:mary :sue])
(index :fred :mary)
(relation :sally [:jen :becky])
(index :sally :jen)
(index :sally :becky))"
[& commands]
(let [wrapper (fn [cur new]
(let [cmd (first new)
body (next new)]
(assert (= 2 (count body)))
(cond
(= cmd 'relation)
`(add-relation ~cur ~(first body) ~(fnext body))
(= cmd 'index)
`(add-index ~cur ~(first body) ~(fnext body))
:otherwise (throwf "%s not recognized" new))))]
(reduce wrapper `empty-database commands)))
(defn get-relation
"Get a relation object by name"
[db rel-name]
(db rel-name))
(defn replace-relation
"Add or replace a fully constructed relation object to the database."
[db rel-name rel]
(assoc db rel-name rel))
;;; DML
(defn database-counts
"Returns a map with the count of elements in each relation."
[db]
(map-values #(-> % :data count) db))
(defn- modify-indexes
"Perform f on the indexed tuple-set. f should take a set and tuple,
and return the new set."
[idxs tuple f]
(into {} (for [ik (keys idxs)]
(let [im (idxs ik)
iv (tuple ik)
os (get im iv #{})
ns (f os tuple)]
[ik (if (empty? ns)
(dissoc im iv)
(assoc im iv (f os tuple)))]))))
(defn- add-to-indexes
"Adds the tuple to the appropriate keys in the index map"
[idxs tuple]
(modify-indexes idxs tuple conj))
(defn- remove-from-indexes
"Removes the tuple from the appropriate keys in the index map"
[idxs tuple]
(modify-indexes idxs tuple disj))
(defn add-tuple
"Two forms:
[db relation-name tuple] adds tuple to the named relation. Returns
the new database.
[rel tuple] adds to the relation object. Returns the new relation."
([db rel-name tuple]
(assert (= (-> tuple keys set) (-> rel-name db :schema)))
(assoc db rel-name (add-tuple (db rel-name) tuple)))
([rel tuple]
(let [data (:data rel)
new-data (conj data tuple)]
(if (identical? data new-data) ; optimization hack!
rel
(let [idxs (add-to-indexes (:indexes rel) tuple)]
(assoc rel :data new-data :indexes idxs))))))
(defn remove-tuple
"Two forms:
[db relation-name tuple] removes the tuple from the named relation,
returns a new database.
[rel tuple] removes the tuple from the relation. Returns the new
relation."
([db rel-name tuple] (assoc db rel-name (remove-tuple (db rel-name) tuple)))
([rel tuple]
(let [data (:data rel)
new-data (disj data tuple)]
(if (identical? data new-data)
rel
(let [idxs (remove-from-indexes (:indexes rel) tuple)]
(assoc rel :data new-data :indexes idxs))))))
(defn add-tuples
"Adds a collection of tuples to the db, as
(add-tuples db
[:rel-name :key-1 1 :key-2 2]
[:rel-name :key-1 2 :key-2 3])"
[db & tupls]
(reduce #(add-tuple %1 (first %2) (apply hash-map (next %2))) db tupls))
(defn- find-indexes
"Given a map of indexes and a partial tuple, return the sets of full tuples"
[idxs pt]
(if (empty? idxs)
nil
(filter identity (for [key (keys pt)]
(if-let [idx-map (idxs key)]
(get idx-map (pt key) #{})
nil)))))
(defn- match?
"Is m2 contained in m1?"
[m1 m2]
(let [compare (fn [key]
(and (contains? m1 key)
(= (m1 key) (m2 key))))]
(every? compare (keys m2))))
(defn- scan-space
"Computes a stream of tuples from relation rn matching partial tuple (pt)
and applies fun to each"
[fun db rn pt]
(let [rel (db rn)
idxs (find-indexes (:indexes rel) pt)
space (if (empty? idxs)
(:data rel) ; table scan :(
(reduce intersection idxs))]
(trace-datalog (when (empty? idxs)
(println (format "Table scan of %s: %s rows!!!!!"
rn
(count space)))))
(fun #(match? % pt) space)))
(defn select
"finds all matching tuples to the partial tuple (pt) in the relation named (rn)"
[db rn pt]
(scan-space filter db rn pt))
(defn any-match?
"Finds if there are any matching records for the partial tuple"
[db rn pt]
(if (= (-> pt keys set) (:schema (db rn)))
(contains? (:data (db rn)) pt)
(scan-space some db rn pt)))
;;; Merge
(defn merge-indexes
[idx1 idx2]
(merge-with (fn [h1 h2] (merge-with union h1 h2)) idx1 idx2))
(defn merge-relations
"Merges two relations"
[r1 r2]
(assert (= (:schema r1) (:schema r2)))
(let [merged-indexes (merge-indexes (:indexes r1)
(:indexes r2))
merged-data (union (:data r1)
(:data r2))]
(assoc r1 :data merged-data :indexes merged-indexes)))
(defn database-merge
"Merges databases together"
[dbs]
(apply merge-with merge-relations dbs))
(defn database-merge-parallel
"Merges databases together in parallel"
[dbs]
(preduce merge-relations dbs))
;; End of file
|