帮忙分析下这个日志，抽取其中的SQL语句

doujikai

aawk是什么版本的awk？

yinyuemi

回复 11# doujikai


    Sorry!多写了个a

wendy_85

aawk -vRS="LOG:  statement|FATAL" '/: select/{print gensub(/\n| +|:/," ","g",$0)}'  file  
  select * from KDB_BIGINT_IDENTITY order by KDB_ID
  select * from   KDB_FLOAT  where KDB_ID =1;
  select * from   TEST_MERGE_INTO_DEST;
  select a.KDB_ID,a.KDB_INT1  from KDB_INT a, KDB_FLOAT b  where a.KDB_ID=b.KDB_ID;

不知道问什么 我觉得写得是一点问题没有 但是我执行后的结果是这样的

  select * from KDB_BIGINT order by KDB_ID
LOG  parse _PLAN0EE4CFC0  select * from KDB_BIGINT order by KDB_ID
  select current_schema()
  select * from KDB_BIGINT_IDENTITY order by KDB_ID
  selectwhere KDB_ID =1;
  selectTEST_MERGE_INTO_DEST;
  selectwhere a.KDB_ID=b.KDB_ID;T b

换行的SQL语句都乱了，再请看看 我弄了半天也没搞明白

yinyuemi

回复 13# wendy_85


    可能是分割符的问题,试试这个：

1. awk -vRS="LOG:  statement|FATAL" '/: select/{print gensub(/\n|\t+|:/," ","g",$0)}'

复制代码

wendy_85

回复 14# yinyuemi


好悲剧 还是不行~

wendy_85

awk -vRS="LOG:  statement|FATAL" '/: select/{print gensub(/\n|\t+|:/," ","g",$0)}' result

  select * from KDB_BIGINT order by  KDB_ID
LOG   parse _PLAN0EE4CFC0  select * from KDB_BIGINT order by  KDB_ID
  select current_schema()
  select * from KDB_BIGINT_IDENTITY order by  KDB_ID
where KDB_ID =1;
TEST_MERGE_INTO_DEST;
where a.KDB_ID=b.KDB_ID;T b

还是长这个样子。

wendy_85

回复 14# yinyuemi


    谢谢  我觉得是我这个日志文件的问题~