generated from just-the-docs/just-the-docs-template
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmysqlformat.py
291 lines (250 loc) · 12.4 KB
/
mysqlformat.py
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
288
289
290
291
import os
import json
from uplan import *
nodetype_mysql = {
'buffer_result': ['Executor', 'Cache'],
'const': ['Producer', 'Constant'],
'system': ['Producer', 'Constant'],
'ALL': ['Producer', 'Full Table Scan'],
'table': ['Producer', 'Full Table Scan'],
'grouping_operation': ['Folder', 'Group'],
'index_merge': ['Join', 'Index Merge'],
'eq_ref': ['Producer', 'Index Scan'],
'fulltext': ['Producer', 'Index Scan'],
'index_subquery': ['Producer', 'Index Scan'],
'ref': ['Producer', 'Index Scan'],
'ref_or_null': ['Producer', 'Index Scan'],
'unique_subquery': ['Producer', 'Index Scan'],
'index': ['Producer', 'Index-Only Scan'],
'materialized_from_subquery': ['Executor', 'Materialize'],
'nested_loop': ['Join', 'Nested Loop'],
'range': ['Producer', 'Range Scan'],
'ordering_operation': ['Combinator', 'Sort'],
'attached_subqueries': ['Producer', 'Subquery Scan'],
'having_subqueries': ['Producer', 'Subquery Scan'],
'select_list_subqueries': ['Producer', 'Subquery Scan'],
'optimized_away_subqueries': ['Producer', 'Subquery Scan'],
'union_result': ['Combinator', 'Union'],
'duplicates_removal': ['Combinator', 'Unique']
}
properties_mysql = {
'attached_condition': ['Configuration', 'filter'],
'cacheable': ['Status', 'cachable'],
'cost_info': ['Cost', 'cost'],
'data_read_per_join': ['Cardinality', 'row read'],
'dependent': ['Status', 'depend'],
'eval_cost': ['Cost', 'cost eval'],
'filtered': ['Status', 'filtered'],
'key': ['Status', 'index used'],
'key_length': ['Status', 'index length'],
'possible_keys': ['Status', 'index available'],
'prefix_cost': ['Cost', 'cost join'],
'query_cost': ['Cost', 'cost'],
'read_cost': ['Cost', 'cost read'],
'ref': ['Configuration', 'index'],
'rows_examined_per_scan': ['Cardinality', 'row read'],
'rows_produced_per_join': ['Cardinality', 'row'],
'sort_cost': ['Cost', 'cost sort'],
'table_name': ['Configuration', 'name object'],
'used_columns': ['Status', 'column used'],
'used_key_parts': ['Status', 'index used'],
'using_filesort': ['Status', 'filesort'],
'using_temporary_table': ['Status', 'temp table']
}
global_properties_mysql_result = {}
# Reference: https://github.com/mysql/mysql-workbench/blob/8.0/plugins/wb.query.analysis/explain_renderer.py
# handle_**(） return a list of TreeNode
# generate_**(） return a TreeNode
class MySQLTree:
def __init__(self, json):
self._json = json
nodes = self.handle(json)
if not nodes:
nodes.append(TreeNode(['Error', 'Not valid query plan']))
self._root = nodes[0]
def get_root(self):
return self._root
def unexpected(self, node, context=""):
if context:
raise Exception("Unexpected: While parsing JSON output: unexpected node in %s: %s\n" % (context, node))
else:
raise Exception("Unexpected: While parsing JSON output: unexpected node: %s\n" % node)
def generate_table_node(self, table):
newNode = TreeNode(nodetype_mysql[table.get('access_type')])
for key in table:
if key == 'cost_info':
for cost_key in table[key]:
newNode.set_prop(properties_mysql[cost_key], table[key][cost_key])
elif key not in ('access_type', 'materialized_from_subquery', 'attached_subqueries'):
if key in properties_mysql:
newNode.set_prop(properties_mysql[key], table[key])
return newNode
def generate_subquery_node(self, name, subqueries, properties):
newNode = TreeNode(nodetype_mysql[name], properties)
newNode.add_children(subqueries)
return newNode
def generate_nested_loop_node(self, children, row):
newNode = TreeNode(nodetype_mysql['nested_loop'])
newNode.add_children(children)
newNode.set_prop(properties_mysql['rows_produced_per_join'], row)
return newNode
def generate_materialized_join_node(self, name, nested_loop, cost_info, properties):
newNode = TreeNode(nodetype_mysql[name])
if properties.get("using_temporary_table"):
newNode.set_prop(properties_mysql['using_temporary_table'], True)
newNode.add_child(nested_loop)
if cost_info:
newNode.set_prop(properties_mysql["cost_info"], float(cost_info.get("read_cost", 0)) + float(cost_info.get("eval_cost", 0)))
return newNode
def generate_operation_node(self, name, children, cost_info = None, properties = None):
if name in ("grouping_operation", "duplicates_removal", "ordering_operation"):
newNode = TreeNode(nodetype_mysql[name])
newNode.add_children(children)
else:
self.unexpected("Unknown operation: %s\n" % name)
if cost_info:
cost_sum = 0.0
for key in cost_info:
cost_sum += float(cost_info[key])
newNode.set_prop(properties_mysql["cost_info"], cost_sum)
if properties.get("using_temporary_table", False):
newNode.set_prop(properties_mysql['using_temporary_table'], True)
if properties.get("using_filesort", False):
newNode.set_prop(properties_mysql['using_filesort'], True)
return newNode
def handle_table(self, table):
newNodes = []
# rows_examined_per_scan: how many rows are examined per scan of the table.
# rows_produced_per_join: how many rows are produced after filter or join.
if 'rows_examined_per_scan' not in table:
table['rows_examined_per_scan'] = (table.get('rows') or 0)
if 'rows_produced_per_join' not in table:
table['rows_produced_per_join'] = (table.get('rows') or 0) * int(float(table.get('filtered', 0))) / 100
newTable = self.generate_table_node(table)
newNodes.append(newTable)
# Put subqueries into a sibling node.
materialized_from_subquery = table.get("materialized_from_subquery")
if materialized_from_subquery:
materialized_from_subquery_node = self.handle_materialized_from_subquery(materialized_from_subquery)
newNodes.extend(materialized_from_subquery_node)
attached_subqueries_ = table.get("attached_subqueries")
if attached_subqueries_:
attached_subqueries = self.handle_attached_subqueries("attached_subqueries", attached_subqueries_)
newNodes.extend(attached_subqueries)
return newNodes
def handle_nested_loop(self, data):
newNodes = []
# For all children of the nested loop, we generate a nested node for any pair of children.
for node in data:
if type(node) is dict and len(node) == 1 and 'table' in node:
newNodes.extend(self.handle_table(node['table']))
elif type(node) is dict and len(node) == 1 and 'duplicates_removal' in node:
newNodes.extend(self.handle_query_block('duplicates_removal', node['duplicates_removal']))
else:
self.unexpected(node, "nested_loop")
if len(newNodes) >= 2:
row = node['table']['rows_produced_per_join'] # the number of produced rows in the right table is the nubmer of rows after the join
loop_node = self.generate_nested_loop_node(newNodes, row)
newNodes = [loop_node]
return newNodes
def handle_attached_subqueries(self, name, subquery_list):
subqueries = []
for data in subquery_list:
qblock = []
properties = {}
for key, value in list(data.items()):
if key == "query_block":
qblock = self.handle_query_block(key, value)
if "query_cost" in value:
properties[properties_mysql["query_cost"][1]] = [properties_mysql["query_cost"][0], value['cost_info']['query_cost']] # merge the cost info in the query block into the subquery node
elif key in ("dependent", "cacheable", "using_temporary_table"):
properties[properties_mysql[key][1]] = [properties_mysql[key][0], value]
elif key == "table": # handles bug #18997475
qblock = self.handle_table(value)
else:
self.unexpected(key, name)
subqueries.extend(qblock)
return [self.generate_subquery_node(name, subqueries, properties)]
def handle_materialized_from_subquery(self, data):
newNodes = []
properties = {}
for key, value in list(data.items()):
if key == "query_block":
inner_qblock = self.handle_query_block(key, value)
newNodes.extend(inner_qblock)
elif key in ("dependent", "cacheable", "using_temporary_table"):
properties[properties_mysql[key][1]] = [properties_mysql[key][0], value]
else:
self.unexpected(key, "materialized_from_subquery")
newNode = TreeNode(nodetype_mysql["materialized_from_subquery"], properties)
newNode.add_children(newNodes)
return [newNode]
def handle_union_result(self, name, data):
properties = {}
qblocks = []
for key, value in list(data.items()):
if key in ("using_temporary_table", "access_type", "table_name"):
properties[key] = value
elif key == "query_specifications":
for qspec in value:
qblocks.extend(self.handle_query_block("query_block", qspec.get("query_block")))
else:
self.unexpected(key, name)
return [self.generate_subquery_node(name, qblocks, properties)]
# Normally, we ignore the node of query block, and show its children dircetly.
# Other nodes, such as aggregation, buffer_result, are also applicable for this function.
def handle_query_block(self, name, data):
newNodes = []
cost_info = None
properties = {}
# Parse the children of query block
for key, value in list(data.items()):
if key == "nested_loop":
newNodes.extend(self.handle_nested_loop(value))
elif key == "table":
newNodes.extend(self.handle_table(value))
elif key in ("grouping_operation", "ordering_operation", "duplicates_removal"):
value['cost_info'] = cost_info
newNodes.extend(self.handle_query_block(key, value))
elif key in ("using_temporary_table", "using_filesort", "dependent"):
properties[key] = value
elif key == "cost_info":
cost_info = value
elif key == "union_result":
newNodes.extend(self.handle_union_result(key, value))
elif key == "select_list_subqueries":
newNodes.extend(self.handle_attached_subqueries(key, value)) # select (select ...) from ...
elif key == "having_subqueries":
newNodes.extend(self.handle_attached_subqueries(key, value))
elif key == "buffer_result":
newNodes.extend(self.handle_query_block(key, value)) # buffer_result is a materialized JOIN
elif key == "select_id":
pass
elif key == "message":
pass
elif key == "optimized_away_subqueries":
pass
else:
self.unexpected(key, name)
# Handle the other nodes
if name == "query_block":
return newNodes
elif name in ("grouping_operation", "ordering_operation", "duplicates_removal"):
return [self.generate_operation_node(name, newNodes, cost_info, properties)]
elif name == "buffer_result":
return [self.generate_materialized_join_node(name, data, cost_info, properties)]
else:
raise Exception("Unexpected query_block name: " + name)
def handle(self, json_data):
output = []
for key, value in list(json_data.items()):
if key == "query_block":
output.extend(self.handle_query_block(key, value))
return output
# It accepts the query plans via `EXPLAIN FORMAT=JSON SELECT ...`.
def parse_mysql(raw_data):
validated_data = raw_data.replace('\\n', '\n')
validated_data = validated_data[validated_data.find('{'):validated_data.rfind('}')+1]
json_data = json.loads(validated_data)
root = MySQLTree(json_data).get_root()
return [root, global_properties_mysql_result]