sql经典语句(SQL classic sentences)I. Basic operation1) DESC, the describe role is to display the structure of the data table using the form: desc data table name2) distinct eliminates duplicate data usage: select, distinct, field name, from data table3) order by field 1 ASC, field 2 desc4) nested queries select, emp.empno, emp.ename, emp.job, emp.salFrom scott.empWhere sal>= (select, Sal, from, scott.emp, where, ename='WARD');5) nested queries select, emp.empno, emp.ename, emp.job, in, emp.salFrom scott.empWhere, Sal, in (select, Sal, from, scott.emp, where, ename ='WARD');6) nested queries select, emp.empno, emp.ename, emp.job, any, emp.salFrom scott.empWhere Sal > any (select, Sal, from, scott.emp, where, job='MANAGER');Equivalent to (1) select, Sal, from, scott.emp, where, job ='MANAGER'(2) select, emp.empno, emp.ename, emp.job, emp.salFrom scott.empData found in where Sal > (1) data a, or, Sal > (1), data found in B, or, Sal > (1), CEg:Select, Sal, from, scott.emp, where, job ='MANAGER'results; 12,10,13Equivalent to sal=12,10,13 or SAL> (12, OR, 10, OR, 13)7) the intersection operation is the intersection concept in the set. The sum of the elements that belong to the set A and belongs to the set B is the intersection. In the command edit area, execute the following statement.Eg:(select, djbh, from, ck_rwd_hz) intersect (select, djbh, from, ck_rwd_mx) documents numbered the sameSelect * from ck_rwd_mx a,((select, djbh, from, ck_rwd_hz) intersect (select, djbh, from, ck_rwd_mx)) BWhere a.djbh =b.djbhTwo function1) ceil takes the minimum integer ceil (N) greater than or equal to the value N; select, Mgr, mgr/100, ceil (mgr/100), from, scott.emp;2) floor takes the maximum integer floor (N) less than or equal to the value N; select, Mgr, mgr/100, floor (mgr/100), from, scott.emp;3) the remainder of mod m divisible by mod (m, n) n4) power, m, N, square, mod (m, n)5) round m four, five, keep the n bit mod (m, n)Select round (8.655,2) from dual; 8.66Select round (8.653,2) from dual; 8.656) sign n>0, take 1; n=0, take 0; n<0, take -1;7) AVG averages AVG (field name)8) count statistics total count (field name) select (*) from scott.emp; select count (distinct job) from scott.emp;9) min calculates numeric fields minimum, select, min (SAL), minimum salary from scott.emp;10) max calculates numeric field maximum, select, max (SAL), highest salary from scott.emp;11) sum calculates the sum of numeric fields, select, sum (SAL), the sum of salaries, from, scott.emp;Three input data1) single data entryInsert into data tables (fields 1, fields 2,...) valuse (field name 1 values, field name 2 values,...)Numeric fields can write values directly; character fields add single quotation marks; date fields add single quotes; at the same time pay attention to the order of days and months2) multi line data entryInsert into data table (field name 1, field name 2,...)(select (field name 1 or operation, field name 2 or operation,...) from data table where condition)3) data replication between tablesCreate table scott.testAs(Select, distinct, empno, ename, hiredate, from, scott.emp, where, empno>=7000);Create table spkpk_liu as select * from spkfk; creates tables and copies data, but creates incomplete table informationThis is all the way to full table backups.Usually after the table is built, you need to see if you want to build the index and primary key again.And "create, table, spkpk_liu, as, select * from, spkfk.""After this table is built, many of the parameter values of the table are the default minimum values, such as the initial value of the original table 10M, and the new table is probably only 256K.Formal environment used in the table, generally do not recommend such a table built.In this way, just a little lazy, if you do so,A statement can achieve the purpose of building tables and inserting data.For example, you need to modify the data in table A, and you might want to back up the data from the A table before you modify it.This time you can use create table... As...This makes it easy to retrieve data from the A table in the futureYou can do this when you debug your own program, but you can't create processes, packages, functions like thisFour delete dataDelete deletes data; truncate deletes the entire table data, but retains the structure1) delete recordsDelete from scott.test where empno and empno <=8000 > = 7500;2) delete the entire dataTruncate table scott.test;Similarities and differences between truncate, delete and dropNote: the delete here refers to the delete statement without the where clauseThe same thing: truncate and the delete without the where clause, and drop will delete the data in the tableDifference:1. truncate and delete only delete data and do not delete the structure of the table (definition)The drop statement will delete the structure of the table, the dependent constraints (constrain), the trigger, and the index (index); the stored procedure / function that depends on the table will be retained, but the invalid state will be changedThe 2.delete statement is DML, which is put into the rollback segement before the transaction is committed, and if you have the corresponding trigger, the execution will be triggeredTruncate, drop is DDL, the operation takes effect immediately, and the original data is not put into the rollback segment and cannot be rolled back. The operation does not trigger trigger.. Obviously, the drop statement releases all the space occupied by the table3. speed, in general: drop>, truncate > deleteOn use, you want to delete part of the data rows, with delete,take care to bring the where clause. The rollback section is large enough to be rolled back through the ROBACK, and there is considerable room for recoveryTo delete tables, of course, use dropYou want to keep the table and delete all the data. If you have nothing to do with the transaction, you can use truncate. Truncate, table, XX, delete the entire table of data, there is no room for recovery, the benefits can be arranged in the table debris, release spaceSo it's better to back up the data firstIf you are tidying up the inner fragments of a table, you can use truncate to catch up with reuse stroage and then import / insert data againFive update dataUpdate data sheetSet field name, 1=, new assignment, field name, 2=, new assignment,...Where conditionUpdate scott.empSet, empno=8888, ename='TOM', hiredate='03-9, -2002'Where empno = 7566;Update scott.empSet sal=(select, sal+300, from, scott.emp, where, empno = 8099)Where empno=8099;Decode (condition, value 1, translation value 1, value 2, translation value 2, value n, translation value n, default value)Six data export1 export the database TEST completely, export the user name system password manager to D:\daochu.dmpExpsystem/manager@TESTfile=d:\daochu.dmp full=y2 export the system user in the database to the table of the sys userExpsystem/manager@TESTfile=d:\daochu.dmp only wner= (system, Sys)3 export tables table1 and table2 from the databaseExpsystem/manager@TESTfile=d:\daochu.dmp tables= (table1, table2)4 export the field filed1 in the table table1 in the database to data that starts with "00"Expsystem/manager@TESTfile=d:\daochu.dmp tables= (table1) query=\ "where filed1 like'00%'\""Explmis_wh/lmis@lmisbuffer=10000 only WNER=lmis_wh rows=nfile=d:\lmis_wh_nodata.dmp log=d:\lmis_wh_nodata.logImplmis/lmis@lmisbuffer=10000, fromuser=lmis_wh, touser=lmis, file=d:\lmis_wh_nodata.dmp, log=d:\lmis_wh_nodata.logC:\>implmis/lmis@lmisbuffer=50000000, full=n,file=e:\daochu.dmp, ignore=y, rows=yCommit=y compile=y fromuser=lmis_wh touser=lmisSeven data import1 import data from the D:\daochu.dmp into the TEST database.Impsystem/manager@TEST file=d:\daochu.dmpThere may be a problem with it because some tables already exist, and then it is reported wrong, and the table is not imported.It's OK to add ignore=y at the back.2 import table table1 from d:\daochu.dmpImpsystem/manager@TEST file=d:\daochu.dmp tables= (table1)SQL definition: SQL is a database oriented general data processing language specification, can complete the following functions: data extraction query, insert modify delete data generation, modify and delete database objects, database security, database integrity control and data protection.SQL classification:DDL - Data Definition Language (CREATE, ALTER, DROP, DECLARE)DML - Data Manipulation Language (SELECT, DELETE, UPDATE, INSERT)DCL - data control language (GRANT, REVOKE, COMMIT, ROLLBACK) DDL - database definition language: direct submission. CREATE: used to create database objects.DECLARE: in addition to creating temporary tables that are used only during the process, the DECLARE statements are very similar to the CREATE statements. The only object that can be declared is the table. And must be added to the user temporary tablespace.DROP: you can delete any object created with CREATE (database object) and DECLARE (table).ALTER: allows you to modify information about certain databaseobjects. Cannot modify index.Eight, the following is mainly based on object presentation of basic grammar1, database:Create database: CREATE, DATABASE, database-name, [USING, CODESET, codeset, TERRITORY, territory]Note: the code page problem.Delete database: drop, database, dbname2, table:Create new table:Create, table, tabname (col1, Type1, [not, null], [primary, key], col2, type2, [not, null],...)Create a new table based on the existing table:A:create, table, tab_new, like, tab_oldB:create, table, tab_new, as, select, col1, col2... From tab_old definition onlyModify table:Add a column:Alter, table, tabname, add, column, col, typeNote: column added will not be deleted. The DB2 column after the data type does not change, the only change is to increase the size of the varchar. Add primary key:Alter, table, tabname, add, primary, key (Col)Delete Primary key:Alter, table, tabname, drop, primary, key (Col)Delete table: drop, table, tabnameAlter table BMDOC_LIUFDrop constraint PK1_BMDOC cascade;3, table space:Create table spaces: create, tablespace, tbsname, PageSize, 4K, managed, by, database, using (file, file, size)Adding containers to tablespace: alter, tablespace, tablespace_name, add (file,'filename', size)Note: the operation is irreversible and will not be removed after adding the container. Therefore, when it is added, pay attention to it.Delete tablespace: drop, tablespace, tbsname4, index:Create indexes: create, [unique], index, idxname, on, tabname (col... ).Delete index: drop, index, idxnameNote: the index is not modifiable. If you want to change it, you must delete it.5 views:Create views: create, view, VIEWNAME, as, select, statementDelete view: drop view VIEWNAMENote: the only change to the view is the reference type column, which changes the range of columns. None of the other definitions can be modified. The view becomes invalid when the view is based on the base table drop.DML - the database manipulation language, which does not implicitly submit the current transaction and is committed to the setting of the visual environment.SELECT: querying data from tablesNote: the connection in the condition avoids Cartesian productDELETE: delete data from existing tablesUPDATE: update data for existing tablesINSERT: insert data into existing tablesNote: whether DELETE, UPDATE, and INSERT are submitted directly depends on the environment in which the statement is executed.Pay attention to the full transaction log when executing.2, DELETE: delete records from the tableSyntax format:DELETE, FROM, tablename, WHERE (conditions)3, INSERT: insert a record into the tableSyntax format:INSERT INTO tablename (col1, col2),... ) VALUES (value1, Value2),... );INSERT INTO tablename (col1, col2),... ) VALUES (value1, Value2),... ) (value1, value2,... ),......Insert does not wait for any program and does not cause locking4, UPDATE:Syntax format:UPDATE tabname SET (col1=values1, col2=values2),... (WHERE) (conditions);Note: update is slower and requires indexing on the corresponding column.Nine permissionsDCL - Data Control LanguageGRANT - Grant user permissionsREVOKE - revoke user rightsCOMMIT - commit transactions can permanently modify the databaseROLLBACK - rollback transactions, eliminating all changes made after the last COMMIT command, so that the contents of the database are restored to the state after the last COMMIT execution.1, GRANT: all or administrators assign access rights to other usersSyntax format:Grant [all privileges|privileges,... On tabname VIEWNAME to [public|user |,... .2 and REVOKE: cancel one of the user's access rightsSyntax format:Revoke [all privileges|privileges,... On tabname VIEWNAME from [public|user |,... .Note: any permissions of users of instance level cannot be canceled. They are not authorized by grant, but are permissions implemented by groups.3, COMMIT: permanently records changes made in the transaction to the database.Syntax format:Commit [work]4, ROLLBACK: will undo all changes made since the last submission.Syntax format:Rollback [work]Ten advanced SQL brief introductionFirst, the query between the use of computing wordsA:UNION operatorThe UNION operator derives a result table by combining two other result tables (such as TABLE1 and TABLE2) and eliminating any duplicate rows in the table.When ALL is used with UNION (that is, UNION ALL), the duplicate rows are not eliminated. In the two case, each row of the derived table does not come from TABLE1, or from TABLE2.B:EXCEPT operatorThe EXCEPT operator derives a result table by including all rows in TABLE1, but not in TABLE2, and eliminating all duplicate rows. When ALL is used with EXCEPT (EXCEPT ALL), the duplicate rows are not eliminated.C:INTERSECT operatorThe INTERSECT operator derives a result table by including only rows in TABLE1 and TABLE2 and eliminating all duplicate rows. When ALL is used with INTERSECT (INTERSECT ALL), the duplicate rows are not eliminated.Note: several query results lines using arithmetic words must be consistent.Appendix: introduction to commonly used functions1, type conversion function:Converted to a numeric type:Decimal, double, Integer, smallint, realHex (ARG): 16 hexadecimal representation converted into parameters.Converted to string type:Char, varcharDigits (ARG): returns the string representation of Arg, and Arg must be decimal.Converted to date or time:Date, time, timestamp2, time and date:Year, quarter, month, week, day, hour, minute, secondDayofyear (ARG): returns the daily value of Arg in the yearDayofweek (ARG): returns the daily value of Arg in the weekDays (ARG): the integer representation of the return date, the number of days from the 0001-01-01.Midnight_seconds (ARG): the number of seconds between midnight and arg.Monthname (ARG): returns the month name of arg.Dayname (ARG): the week that returns arg.3 string function:Length, lcase, ucase, ltrim, rtrimCoalesce (arg1, arg2)... ) returns the first non null parameter in the argument set.Concat (arg1, arg2): connect two strings, arg1 and arg2.Insert (arg1, POS, size, arg2): returns a arg1 that removes size characters from POS and inserts arg2 into that location.Left (Arg, length): returns the leftmost length string of arg.Locate (arg1, arg2, <, pos>): find the location of the first occurrence of arg1 in arg2, specify POS, and start looking for the location of the arg1 from the POS of arg2.Posstr (arg1, arg2): returns the position where arg2 first appeared in arg1.Repeat (arg1, num_times): returns the string arg1 repeated num_times times.Replace (arg1, arg2, ARG3): replace all arg2 in arg1 to arg3.Right (Arg, length): returns a string consisting of lengthbytes on the left of the arg.Space (ARG): returns a string containing Arg spaces.Substr (arg1, POS, <, length>): returns the length character at the start of the POS position in arg1, and returns the remaining characters if no length is specified.4. Mathematical function:Abs, count, Max, min, sumCeil (ARG): returns the smallest integer greater than or equal to arg.Floor (ARG): returns the smallest integer less than or equal to the parameter.Mod (arg1,Arg2) returns arg1 by the remainder of the arg2, with the same symbol as arg1.Rand (): returns a random number between 1 and 1.Power (arg1, arg2): returns the arg2 power of arg1.Round (arg1, arg2): four, five into the truncated processing, arg2 is the number of bits, if arg2 is negative, then the number of decimal points before four to five processing.Sigh (ARG): symbolic designator to return arg. -1,0,1 representation.Truncate (arg1, arg2): truncate arg1, arg2 is the number of digits, and if arg2 is negative, keep the arg2 bits before the arg1 decimal point.5, others:Nullif (arg1, arg2): returns if the 2 arguments are equal, otherwise the argument 1 is returned。