Lab3_QueryProcessor/executionEngine.py at main · Paul-Motter/Lab3_QueryProcessor · GitHub

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
import math
from typing import Any

import physicalQueryTools
import index

def execute(physPlan: dict[str, Any], indexes: dict[str, Any], relations: dict[str, Any]) -> tuple[int, list[Any], list[str]]:
    # Scan Operation
    if (physPlan.get("op") == "scan"):
        schema = [f"{physPlan.get('relation')}.{attr}" for attr in relations.get("relations").get(physPlan.get("relation")).get("schema")]
        io = 0
        outputTuples = []
        for page in relations.get("relations").get(physPlan.get("relation")).get("pages"):
            # read in the page.
            io = io + 1
            for tuple in page:
                # get each tuple
                outputTuples.append(tuple)
        return io, outputTuples, schema

    # Index Scan Operation
    elif (physPlan.get("op") == "indexScan"):
        schema = [f"{physPlan.get('relation')}.{attr}" for attr in relations.get("relations").get(physPlan.get("relation")).get("schema")]
        # parsing predicate
        predicate = physPlan.get("predicate")
        subPredicates = []
        currentPredicate = []
        for token in predicate:
            if token == "AND":
                subPredicates.append(currentPredicate)
                currentPredicate.clear()
            else:
                currentPredicate.append(token)
        if (len(currentPredicate) == 3):
            subPredicates.append(currentPredicate)
        # find Index
        indexToUse = None
        predicateWithIndex = None
        for subPredicate in subPredicates:
            atributesInSubPredicate = set(subPredicate).intersection(physicalQueryTools.allAttributes)
            indexesInSubpredicate = atributesInSubPredicate.intersection(set(indexes.keys()))
            if (len(indexesInSubpredicate) > 0):
                indexToUse = iter(indexesInSubpredicate).__next__()
                predicateWithIndex: list[str] = subPredicate
                break
        searchKey = predicateWithIndex.copy()
        searchKey.remove(indexToUse)
        searchKey.remove("=")
        searchKey = searchKey[0]
        # get tuples matching the key
        io, tuples = index.indexLookup(indexToUse, searchKey, subPredicates, relations.get("relations").get(indexToUse.split(".")[0]))
        return io, tuples, schema

    elif (physPlan.get("op") == "filter"):
        subtotalIO, tuples, schema = execute(physPlan.get("child"), indexes, relations)
        predicate = physPlan.get("predicate")
        survivingTuples = []
        # getSubPredicates
        subPredicates = []
        currentPredicate = []
        for token in predicate:
            if token == "AND":
                subPredicates.append(currentPredicate)
                currentPredicate.clear()
            else:
                currentPredicate.append(token)
        if (len(currentPredicate) == 3):
            subPredicates.append(currentPredicate)
        # filter the tuples by all predicates.
        for t in tuples:
            predicateFail = False
            for predicate in subPredicates:
                lhs = predicate[0]
                rhs = predicate[2]
                # if lhs is an attr get the attr
                if (lhs in physicalQueryTools.allAttributes):
                    lhs = lhs.split(".")
                    lhs = t[relations.get("relations").get(lhs[0]).get("schema").index(lhs[1])]
                # if rhs is an attr get the attr
                if (rhs in physicalQueryTools.allAttributes):
                    rhs = rhs.split(".")
                    rhs = t[relations.get("relations").get(rhs[0]).get("schema").index(rhs[1])]
                # predicate fails and tuple is not added.
                if (lhs != rhs):
                    predicateFail = True
                    break
            if (not predicateFail):
                # io = io + 1
                survivingTuples.append(t)
        # io = math.ceil(io/2.0)
        return subtotalIO, survivingTuples, schema

    elif (physPlan.get("op") == "project"):
        subtotalIO, tuples, schema = execute(physPlan.get("child"), indexes, relations)
        projAttrs = physPlan.get("attrs")
        keepTupleIndexes = []
        newSchema = []
        # find the indexes to keep
        for attr in schema:
            if (attr in projAttrs):
                keepTupleIndexes.append(schema.index(attr))
                newSchema.append(attr)
        # construct new tuple.
        newTuples = []
        for t in tuples:
            newT = []
            for i in keepTupleIndexes:
                newT.append(t[i])
            newTuples.append(newT)
        # convert to pages
        return subtotalIO, newTuples, newSchema

    elif (physPlan.get("op") == "hashedIndexJoin"):
        subtotalIOL, tuplesL, schemaL = execute(physPlan.get("left"), indexes, relations)
        io = subtotalIOL
        # subtotalIOR, tuplesR, schemaR = execute(physPlan.get("right"), indexes, relations)
        condition = physPlan.get("condition")
        conditionIndexL = schemaL.index(condition[0])
        indexToUse = condition[2]

        schemaR = [x.strip("'") for x in relations.get("relations").get(indexToUse.split(".")[0]).get("schema")]
        conditionIndexR = schemaR.index(condition[2].split(".")[1])

        schemaR.pop(conditionIndexR)
        newSchema = schemaL+schemaR
        # IO cost is B(R)+T(R)*lookupCost
        newTuples = []
        pagesAccessed = set()
        for tL in tuplesL:
            lookupIO, tuplesR = index.indexLookup(indexToUse, tL[conditionIndexL], [], relations.get("relations").get(indexToUse.split(".")[0]), checkPredicates=False, pagesAccessed=pagesAccessed)
            io = io + lookupIO
            for tR in tuplesR:
                # if (tL[conditionIndexL] == tR[conditionIndexR]):
                newTuple = tL+tR[:conditionIndexR] + tR[conditionIndexR+1:]
                newTuples.append(newTuple)

        return io, newTuples, newSchema

    elif (physPlan.get("op") == "blockNestedLoopJoin"):
        io = 0
        subtotalIOL, tuplesL, schemaL = execute(physPlan.get("left"), indexes, relations)
        subtotalIOR, tuplesR, schemaR = execute(physPlan.get("right"), indexes, relations)
        condition = physPlan.get("condition")
        conditionIndexL = schemaL.index(condition[0])
        conditionIndexR = schemaR.index(condition[2])
        schemaR.pop(conditionIndexR)
        newSchema = schemaL+schemaR
        # IO cost is B(R)+B(R)*B(S)
        pageSize = relations.get("page_size")
        # io = math.ceil(len(tuplesL)/pageSize) + math.ceil(len(tuplesL)/pageSize)*math.ceil(len(tuplesR)/pageSize)
        io = subtotalIOL + len(tuplesL)*subtotalIOR
        newTuples = []

        for tL in tuplesL:
            for tR in tuplesR:
                if (tL[conditionIndexL] == tR[conditionIndexR]):
                    newTuple = tL+tR[:conditionIndexR] + tR[conditionIndexR+1:]
                    newTuples.append(newTuple)
        # return subtotalIOL+subtotalIOR+io, newTuples, newSchema
        return io, newTuples, newSchema

    elif (physPlan.get("op") == "hashJoin"):
        subtotalIOL, tuplesL, schemaL = execute(physPlan.get("left"), indexes, relations)
        subtotalIOR, tuplesR, schemaR = execute(physPlan.get("right"), indexes, relations)
        condition = physPlan.get("condition")
        conditionIndexL = schemaL.index(condition[0])
        conditionIndexR = schemaR.index(condition[2])
        schemaR.pop(conditionIndexR)
        newSchema = schemaL+schemaR

        # IO B(R) + B(S)
        io = subtotalIOR + subtotalIOL
        # Build phase
        hashTable = {}
        for tL in tuplesL:
            key = tL[conditionIndexL]
            if (hashTable.get(key) != None):
                hashTable[key].append(tL)
            else:
                hashTable[key] = [tL]
        # Probe Phase
        newTuples = []
        for tR in tuplesR:
            key = tR[conditionIndexR]
            if (hashTable.get(key) != None):
                for tL in hashTable.get(key):
                    newTuple = tL+tR[:conditionIndexR] + tR[conditionIndexR+1:]
                    newTuples.append(newTuple)
        return io, newTuples, newSchema