通过shell脚本得到数据字典的信息

在平时的工作中，可能需要查询一些数据字典的信息，比如数据字典对应的基表信息，可以得到更多数据库内部的一些详细信息。
比如user_objects这个数据字典视图，里面可能就包含很多的信息。

首先可以得到，它是一个视图，然后在public上创建了对应的同义词，但是在不用用户使用的使用，查出的结果会不相同，这个其实是根据类似环境变量的一套东西来实现的。
   where l.owner# = userenv('SCHEMAID')，它会锁定当前的schema。

OWNER                           OBJECT_ID DATA_OBJECT_ID OBJECT_NAME                    OBJECT_TYPE
------------------------------ ---------- -------------- ------------------------------ -------------------
SYS                                  3305                USER_OBJECTS                   VIEW
PUBLIC                               3306                USER_OBJECTS                   SYNONYM

synonym_details

OWNER                          SYNONYM_NAME
------------------------------ ------------------------------
PUBLIC                         USER_OBJECTS

view_details

VIEW_NAME                      TEXT
------------------------------ --------------------------------------------------------------------------------
USER_OBJECTS                   select o.name, o.subname, o.obj#, o.dataobj#,
                                      decode(o.type#, 0, 'NEXT OBJECT', 1, 'INDEX', 2, 'TABLE', 3, 'CLUSTER',
                                                     4, 'VIEW', 5, 'SYNONYM', 6, 'SEQUENCE',
                                                     7, 'PROCEDURE', 8, 'FUNCTION', 9, 'PACKAGE',
                                                     11, 'PACKAGE BODY', 12, 'TRIGGER',
                                                     13, 'TYPE', 14, 'TYPE BODY',
                                                     19, 'TABLE PARTITION', 20, 'INDEX PARTITION', 21, 'LOB',
                                                     22, 'LIBRARY', 23, 'DIRECTORY',  24, 'QUEUE',
                                                     28, 'JAVA SOURCE', 29, 'JAVA CLASS', 30, 'JAVA RESOURCE',
                                                     32, 'INDEXTYPE', 33, 'OPERATOR',
                                                     34, 'TABLE SUBPARTITION', 35, 'INDEX SUBPARTITION',
                                                     40, 'LOB PARTITION', 41, 'LOB SUBPARTITION',
                                                     42, NVL((SELECT 'REWRITE EQUIVALENCE'
                                                              FROM sum$ s
                                                              WHERE s.obj#=o.obj#
                                                                    and bitand(s.xpflags, 8388608) = 8388608),
                                                             'MATERIALIZED VIEW'),
                                                     43, 'DIMENSION',
                                                     44, 'CONTEXT', 46, 'RULE SET', 47, 'RESOURCE PLAN',
                                                     48, 'CONSUMER GROUP',
                                                     51, 'SUBSCRIPTION', 52, 'LOCATION',
                                                     55, 'XML SCHEMA', 56, 'JAVA DATA',
                                                     57, 'EDITION', 59, 'RULE',
                                                     60, 'CAPTURE', 61, 'APPLY',
                                                     62, 'EVALUATION CONTEXT',
                                                     66, 'JOB', 67, 'PROGRAM', 68, 'JOB CLASS', 69, 'WINDOW',
                                                     72, 'SCHEDULER GROUP', 74, 'SCHEDULE', 79, 'CHAIN',
                                                     81, 'FILE GROUP', 82, 'MINING MODEL',  87, 'ASSEMBLY',
                                                     90, 'CREDENTIAL', 92, 'CUBE DIMENSION', 93, 'CUBE',
                                                     94, 'MEASURE FOLDER', 95, 'CUBE BUILD PROCESS',
                                                     100, 'FILE WATCHER', 101, 'DESTINATION',
                                                     'UNDEFINED'),
                                      o.ctime, o.mtime,
                                      to_char(o.stime, 'YYYY-MM-DD:HH24:MI:SS'),
                                      decode(o.status, 0, 'N/A', 1, 'VALID', 'INVALID'),
                                      decode(bitand(o.flags, 2), 0, 'N', 2, 'Y', 'N'),
                                      decode(bitand(o.flags, 4), 0, 'N', 4, 'Y', 'N'),
                                      decode(bitand(o.flags, 16), 0, 'N', 16, 'Y', 'N'),
                                      o.namespace,
                                      o.defining_edition
                               from sys."_CURRENT_EDITION_OBJ" o
                               where o.owner# = userenv('SCHEMAID')
                                 and o.linkname is null
                                 and (o.type# not in (1  /* INDEX - handled below */,
                                                     10 /* NON-EXISTENT */)
                                      or
                                      (o.type# = 1 and 1 = (select 1
                                                            from sys.ind$ i
                                                           where i.obj# = o.obj#
                                                             and i.type# in (1, 2, 3, 4, 6, 7, 8, 9))))
                                 and o.name != '_NEXT_OBJECT'
                                 and o.name != '_default_auditing_options_'
                                 and bitand(o.flags, 128) = 0
                               union all
                               select l.name, NULL, to_number(null), to_number(null),
                                      'DATABASE LINK',
                                      l.ctime, to_date(null), NULL, 'VALID', 'N', 'N', 'N', NULL, NULL
                               from sys.link$ l
                               where l.owner# = userenv('SCHEMAID')


实现的脚本如下：

sqlplus -s  / as sysdba < COL OBJECT_NAME FORMAT A30
COL DB_LINK FORMAT A20
COL OWNER FORMAT A30
COL TABLE_OWNER FORMAT A30
COL TABLE_NAME FORMAT A30
SET LINESIZE 200
SET PAGES 100
SET LONG 999999
prompt object_details
SELECT OWNER,OBJECT_ID,DATA_OBJECT_ID,OBJECT_NAME,OBJECT_TYPE FROM DBA_OBJECTS WHERE OBJECT_NAME=UPPER('$1');

prompt synonym_details
SELECT OWNER,SYNONYM_NAME FROM DBA_SYNONYMS WHERE SYNONYM_NAME=upper('$1');

prompt view_details
SELECT VIEW_NAME,TEXT FROM DBA_VIEWS WHERE VIEW_NAME=upper('$1');

 

EOF
exit

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/23718752/viewspace-1256265/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/23718752/viewspace-1256265/