未验证 提交 044a252e 编写于 作者: S Shengliang Guan 提交者: GitHub

Merge pull request #18565 from taosdata/lihui3.0/test

test: modify limit2.sim to limit.py
......@@ -529,6 +529,7 @@
,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/leastsquares.py -R
,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/length.py
,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/length.py -R
,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/limit.py
,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/log.py
,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/log.py -R
,,y,system-test,./pytest.sh python3 ./test.py -f 2-query/lower.py
......
import taos
import sys
import time
import socket
import os
import threading
import math
from util.log import *
from util.sql import *
from util.cases import *
from util.dnodes import *
from util.common import *
# from tmqCommon import *
class TDTestCase:
def __init__(self):
self.vgroups = 2
self.ctbNum = 10
self.rowsPerTbl = 10000
def init(self, conn, logSql, replicaVar=1):
self.replicaVar = int(replicaVar)
tdLog.debug(f"start to excute {__file__}")
tdSql.init(conn.cursor(), False)
def create_database(self,tsql, dbName,dropFlag=1,vgroups=2,replica=1):
if dropFlag == 1:
tsql.execute("drop database if exists %s"%(dbName))
tsql.execute("create database if not exists %s vgroups %d replica %d"%(dbName, vgroups, replica))
tdLog.debug("complete to create database %s"%(dbName))
return
def create_stable(self,tsql, paraDict):
colString = tdCom.gen_column_type_str(colname_prefix=paraDict["colPrefix"], column_elm_list=paraDict["colSchema"])
tagString = tdCom.gen_tag_type_str(tagname_prefix=paraDict["tagPrefix"], tag_elm_list=paraDict["tagSchema"])
# tdLog.debug(colString)
# tdLog.debug(tagString)
sqlString = f"create table if not exists %s.%s (%s) tags (%s)"%(paraDict["dbName"], paraDict["stbName"], colString, tagString)
tdLog.debug("%s"%(sqlString))
tsql.execute(sqlString)
return
def create_ctable(self,tsql=None, dbName='dbx',stbName='stb',ctbPrefix='ctb',ctbNum=1,ctbStartIdx=0):
for i in range(ctbNum):
sqlString = "create table %s.%s%d using %s.%s tags(%d, 'tb%d', 'tb%d', %d, %d, %d)"%(dbName,ctbPrefix,i+ctbStartIdx,dbName,stbName,i+ctbStartIdx,i+ctbStartIdx,i+ctbStartIdx,i+ctbStartIdx,i+ctbStartIdx,i+ctbStartIdx)
tsql.execute(sqlString)
tdLog.debug("complete to create %d child tables by %s.%s" %(ctbNum, dbName, stbName))
return
def insert_data(self,tsql,dbName,ctbPrefix,ctbNum,rowsPerTbl,batchNum,startTs,tsStep):
tdLog.debug("start to insert data ............")
tsql.execute("use %s" %dbName)
pre_insert = "insert into "
sql = pre_insert
# t = time.time()
# startTs = int(round(t * 1000))
#tdLog.debug("doing insert data into stable:%s rows:%d ..."%(stbName, allRows))
for i in range(ctbNum):
rowsBatched = 0
sql += " %s%d values "%(ctbPrefix,i)
for j in range(rowsPerTbl):
if (i < ctbNum/2):
sql += "(%d, %d, %d, %d,%d,%d,%d,true,'binary%d', 'nchar%d') "%(startTs + j*tsStep, j%10, j%10, j%10, j%10, j%10, j%10, j%10, j%10)
else:
sql += "(%d, %d, NULL, %d,NULL,%d,%d,true,'binary%d', 'nchar%d') "%(startTs + j*tsStep, j%10, j%10, j%10, j%10, j%10, j%10)
rowsBatched += 1
if ((rowsBatched == batchNum) or (j == rowsPerTbl - 1)):
tsql.execute(sql)
# print("===sql: %s"%(sql))
rowsBatched = 0
if j < rowsPerTbl - 1:
sql = "insert into %s%d values " %(ctbPrefix,i)
else:
sql = "insert into "
#end sql
if sql != pre_insert:
#print("insert sql:%s"%sql)
tsql.execute(sql)
# print("===sql: %s"%(sql))
tdLog.debug("insert data ............ [OK]")
return
def prepareTestEnv(self):
tdLog.printNoPrefix("======== prepare test env include database, stable, ctables, and insert data: ")
paraDict = {'dbName': 'lm2_db0',
'dropFlag': 1,
'vgroups': 2,
'stbName': 'lm2_stb0',
'colPrefix': 'c',
'tagPrefix': 't',
'colSchema': [{'type': 'INT', 'count':1},{'type': 'BIGINT', 'count':1},{'type': 'FLOAT', 'count':1},{'type': 'DOUBLE', 'count':1},{'type': 'smallint', 'count':1},{'type': 'tinyint', 'count':1},{'type': 'bool', 'count':1},{'type': 'binary', 'len':10, 'count':1},{'type': 'nchar', 'len':10, 'count':1}],
'tagSchema': [{'type': 'INT', 'count':1},{'type': 'nchar', 'len':20, 'count':1},{'type': 'binary', 'len':20, 'count':1},{'type': 'BIGINT', 'count':1},{'type': 'smallint', 'count':1},{'type': 'DOUBLE', 'count':1}],
'ctbPrefix': 'lm2_tb',
'ctbStartIdx': 0,
'ctbNum': 10,
'rowsPerTbl': 10000,
'batchNum': 3000,
'startTs': 1537146000000,
'tsStep': 600000}
paraDict['vgroups'] = self.vgroups
paraDict['ctbNum'] = self.ctbNum
paraDict['rowsPerTbl'] = self.rowsPerTbl
tdLog.info("create database")
self.create_database(tsql=tdSql, dbName=paraDict["dbName"], dropFlag=paraDict["dropFlag"], vgroups=paraDict["vgroups"], replica=self.replicaVar)
tdLog.info("create stb")
self.create_stable(tsql=tdSql, paraDict=paraDict)
tdLog.info("create child tables")
self.create_ctable(tsql=tdSql, dbName=paraDict["dbName"],stbName=paraDict["stbName"],ctbPrefix=paraDict["ctbPrefix"],ctbNum=paraDict["ctbNum"],ctbStartIdx=paraDict["ctbStartIdx"])
self.insert_data(tsql=tdSql, dbName=paraDict["dbName"],ctbPrefix=paraDict["ctbPrefix"],ctbNum=paraDict["ctbNum"],rowsPerTbl=paraDict["rowsPerTbl"],batchNum=paraDict["batchNum"],startTs=paraDict["startTs"],tsStep=paraDict["tsStep"])
return
def tmqCase1(self):
tdLog.printNoPrefix("======== test case 1: ")
paraDict = {'dbName': 'lm2_db0',
'dropFlag': 1,
'vgroups': 2,
'stbName': 'lm2_stb0',
'colPrefix': 'c',
'tagPrefix': 't',
'colSchema': [{'type': 'INT', 'count':1},{'type': 'BIGINT', 'count':1},{'type': 'FLOAT', 'count':1},{'type': 'DOUBLE', 'count':1},{'type': 'smallint', 'count':1},{'type': 'tinyint', 'count':1},{'type': 'bool', 'count':1},{'type': 'binary', 'len':10, 'count':1},{'type': 'nchar', 'len':10, 'count':1}],
'tagSchema': [{'type': 'INT', 'count':1},{'type': 'nchar', 'len':20, 'count':1},{'type': 'binary', 'len':20, 'count':1},{'type': 'BIGINT', 'count':1},{'type': 'smallint', 'count':1},{'type': 'DOUBLE', 'count':1}],
'ctbPrefix': 'lm2_tb',
'ctbStartIdx': 0,
'ctbNum': 10,
'rowsPerTbl': 10000,
'batchNum': 3000,
'startTs': 1537146000000,
'tsStep': 600000}
val1 = 1
val2 = paraDict["ctbNum"] - 1
# select count(*), t1, t2, t3, t4, t5, t6 from $stb where t1 > $val1 and t1 < $val2 group by t1, t2, t3, t4, t5, t6 order by t1 asc limit 1 offset 0
sqlStr = f"select count(*), t1, t2, t3, t4, t5, t6 from %s where t1 > %d and t1 < %d group by t1, t2, t3, t4, t5, t6 order by t1 asc limit 1 offset 0"%(paraDict["stbName"], val1, val2)
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(1)
tdSql.checkData(0, 0, paraDict["rowsPerTbl"])
tdSql.checkData(0, 1, 2)
tdSql.checkData(0, 2, "tb2")
tdSql.checkData(0, 3, "tb2")
tdSql.checkData(0, 4, 2)
tdSql.checkData(0, 5, 2)
# select count(*), t3, t4 from $stb where t2 like '%' and t1 > 2 and t1 < 5 group by t3, t4 order by t3 desc limit 2 offset 0
sqlStr = f"select count(*), t3, t4 from %s where t2 like '%%' and t1 > 2 and t1 < 5 group by t3, t4 order by t3 desc limit 2 offset 0"%(paraDict["stbName"])
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(2)
tdSql.checkData(0, 0, paraDict["rowsPerTbl"])
tdSql.checkData(0, 1, "tb4")
tdSql.checkData(0, 2, 4)
tdSql.checkData(1, 1, "tb3")
tdSql.checkData(1, 2, 3)
# select count(*) from $stb where t2 like '%' and t1 > 2 and t1 < 5 group by t3, t4 order by t3 desc limit 1 offset 1
sqlStr = f"select count(*) from %s where t2 like '%%' and t1 > 2 and t1 < 5 group by t3, t4 order by t3 desc limit 1 offset 1"%(paraDict["stbName"])
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(1)
## TBASE-348
tdSql.error(f"select count(*) from %s where t1 like 1"%(paraDict["stbName"]))
ts0 = paraDict["startTs"]
delta = paraDict["tsStep"]
tsu = paraDict["rowsPerTbl"] * delta
tsu = tsu - delta
tsu = tsu + ts0
tb = paraDict["ctbPrefix"] + '0'
# select _wstart, max(c1) from $tb where ts >= $ts0 and ts <= $tsu interval(5m) fill(value, -1) limit 10 offset 1
sqlStr = f"select _wstart, max(c1) from %s where ts >= %d and ts <= %d interval(5m) fill(value, -1) limit 10 offset 1"%(tb, ts0, tsu)
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(10)
tdSql.checkData(0, 0, "18-09-17 09:05:00.000")
tdSql.checkData(0, 1, -1)
tdSql.checkData(1, 1, 1)
tdSql.checkData(9, 0, "18-09-17 09:50:00.000")
tdSql.checkData(9, 1, 5)
tb5 = paraDict["ctbPrefix"] + '5'
sqlStr = f"select max(c1), min(c2) from %s where ts >= %d and ts <= %d interval(5m) fill(value, -1, -2) limit 10 offset 1"%(tb5, ts0, tsu)
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(10)
tdSql.checkData(0, 0, -1)
tdSql.checkData(0, 1, -2)
tdSql.checkData(1, 0, 1)
tdSql.checkData(1, 1, -2)
tdSql.checkData(9, 0, 5)
tdSql.checkData(9, 1, -2)
### [TBASE-350]
## tb + interval + fill(value) + limit offset
tb = paraDict["ctbPrefix"] + '0'
limit = paraDict["rowsPerTbl"]
offset = limit / 2
sqlStr = f"select max(c1), min(c2), sum(c3), avg(c4), stddev(c5), spread(c6), first(c7), last(c8), first(c9) from %s where ts >= %d and ts <= %d interval(5m) fill(value, -1, -2 ,-3, -4 , -5, -6 ,-7 ,'-8', '-9') limit %d offset %d"%(tb, ts0, tsu, limit, offset)
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(limit)
tdSql.checkData(0, 1, 0)
sqlStr = f"select max(c1) from lm2_tb0 where ts >= 1537146000000 and ts <= 1543145400000 interval(5m) fill(value, -1000) limit 8200"
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(8200)
sqlStr = f"select max(c1) from lm2_tb0 where ts >= 1537146000000 and ts <= 1543145400000 interval(5m) fill(value, -1000) limit 100000;"
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
sqlStr = f"select max(c1) from lm2_tb0 where ts >= 1537146000000 and ts <= 1543145400000 interval(5m) fill(value, -1000) limit 10 offset 8190;"
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(10)
tdSql.checkData(0, 0, 5)
tdSql.checkData(1, 0, -1000)
tdSql.checkData(2, 0, 6)
tdSql.checkData(3, 0, -1000)
sqlStr = f"select max(c1) from lm2_tb0 where ts >= 1537146000000 and ts <= 1543145400000 interval(5m) fill(value, -1000) limit 10 offset 10001;"
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(10)
tdSql.checkData(0, 0, -1000)
tdSql.checkData(1, 0, 1)
tdSql.checkData(2, 0, -1000)
tdSql.checkData(3, 0, 2)
sqlStr = f"select max(c1) from lm2_tb0 where ts >= 1537146000000 and ts <= 1543145400000 interval(5m) fill(value, -1000) limit 10000 offset 10001;"
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(9998)
sqlStr = f"select max(c1) from lm2_tb0 where ts >= 1537146000000 and ts <= 1543145400000 interval(5m) fill(value, -1000) limit 100 offset 20001;"
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(0)
# tb + interval + fill(linear) + limit offset
limit = paraDict["rowsPerTbl"]
offset = limit / 2
sqlStr = f"select _wstart,max(c1), min(c2), sum(c3), avg(c4), stddev(c5), spread(c6), first(c7), last(c8), first(c9) from %s where ts >= %d and ts <= %d interval(5m) fill(linear) limit %d offset %d"%(tb,ts0,tsu,limit, offset)
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(limit)
tdSql.checkData(0, 1, 0)
tdSql.checkData(1, 1, 0)
tdSql.checkData(1, 3, 0.500000000)
tdSql.checkData(3, 5, 0.000000000)
tdSql.checkData(4, 6, 0.000000000)
tdSql.checkData(4, 7, 1)
tdSql.checkData(5, 7, None)
tdSql.checkData(6, 8, "binary3")
tdSql.checkData(7, 9, None)
limit = paraDict["rowsPerTbl"]
offset = limit / 2
sqlStr = f"select max(c1), min(c2), sum(c3), avg(c4), stddev(c5), spread(c6), first(c7), last(c8), first(c9) from %s where ts >= %d and ts <= %d interval(5m) fill(prev) limit %d offset %d"%(tb,ts0,tsu,limit, offset)
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(limit)
limit = paraDict["rowsPerTbl"]
offset = limit / 2 + 10
sqlStr = f"select _wstart,max(c1), min(c2), sum(c3), avg(c4), stddev(c5), spread(c6), first(c7), last(c8), first(c9) from %s where ts >= %d and ts <= %d and c1 = 5 interval(5m) fill(value, -1, -2 ,-3, -4 , -5, -6 ,-7 ,'-8', '-9') limit %d offset %d"%(tb,ts0,tsu,limit, offset)
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(limit)
tdSql.checkData(0, 1, 5)
tdSql.checkData(0, 2, 5)
tdSql.checkData(0, 3, 5.000000000)
tdSql.checkData(0, 4, 5.000000000)
tdSql.checkData(0, 5, 0.000000000)
tdSql.checkData(0, 7, 1)
tdSql.checkData(0, 8, "binary5")
tdSql.checkData(0, 9, "nchar5")
tdSql.checkData(1, 8, None)
tdSql.checkData(1, 9, None)
limit = paraDict["rowsPerTbl"]
offset = limit * 2 - 11
sqlStr = f"select _wstart,max(c1), min(c2), sum(c3), avg(c4), stddev(c5), spread(c6), first(c7), last(c8), first(c9) from %s where ts >= %d and ts <= %d and c1 = 5 interval(5m) fill(value, -1, -2 ,-3, -4 , -5, -6 ,-7 ,'-8', '-9') limit %d offset %d"%(tb,ts0,tsu,limit, offset)
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(10)
tdSql.checkData(0, 1, -1)
tdSql.checkData(0, 2, -2)
tdSql.checkData(1, 1, 5)
tdSql.checkData(1, 2, 5)
tdSql.checkData(1, 3, 5.000000000)
tdSql.checkData(1, 5, 0.000000000)
tdSql.checkData(1, 6, 0.000000000)
tdSql.checkData(1, 8, "binary5")
tdSql.checkData(1, 9, "nchar5")
### [TBASE-350]
## stb + interval + fill + group by + limit offset
sqlStr = f"select max(c1), min(c2), sum(c3), avg(c4), first(c7), last(c8), first(c9) from lm2_tb0 where ts >= 1537146000000 and ts <= 1543145400000 partition by t1 interval(5m) fill(value, -1, -2, -3, -4 ,-7 ,'-8', '-9') limit 2 offset 10"
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(2)
limit = 5
offset = paraDict["rowsPerTbl"] * 2
offset = offset - 2
sqlStr = f"select max(c1), min(c2), sum(c3), avg(c4), first(c7), last(c8), first(c9) from lm2_tb0 where ts >= 1537146000000 and ts <= 1543145400000 partition by t1 interval(5m) fill(value, -1, -2, -3, -4 ,-7 ,'-8', '-9') order by t1 limit %d offset %d"%(limit, offset)
print("====sql:%s"%(sqlStr))
tdSql.query(sqlStr)
tdSql.checkRows(1)
tdSql.checkData(0, 0, 9)
tdSql.checkData(0, 1, 9)
tdSql.checkData(0, 2, 9.000000000)
tdSql.checkData(0, 3, 9.000000000)
tdSql.checkData(0, 4, 1)
tdSql.checkData(0, 5, "binary9")
tdSql.checkData(0, 6, "nchar9")
#add one more test case
sqlStr = f"select max(c1), last(c8) from lm2_db0.lm2_tb0 where ts >= 1537146000000 and ts <= 1543145400000 interval(5m) fill(linear) limit 10 offset 4089;"
tdLog.printNoPrefix("======== test case 1 end ...... ")
def run(self):
tdSql.prepare()
self.prepareTestEnv()
self.tmqCase1()
def stop(self):
tdSql.close()
tdLog.success(f"{__file__} successfully executed")
event = threading.Event()
tdCases.addLinux(__file__, TDTestCase())
tdCases.addWindows(__file__, TDTestCase())
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册