hadoop - pig 替换命令

Question

我有一个文件，其数据如下

11/16/2015,"others (phone,health,etc.)",cont'd attempts,"resource,inc.",dg

我只想删除双引号内的逗号。

预期结果

11/162015,"others(phone health etc.)",cont'd attempts,"resource inc.",dg 

到目前为止，我尝试了什么

Foreach a generate replace ($1,',',''); 
Foreach a generate regex_extract($1,'[\,]+',1);

但是它们都不起作用。

Answer 1

所有拳头都使用 REGULAR EXP 来分隔元组中的字段，然后应用 REPLACE

试试这个代码:

a = load '<path>' as line;

b = foreach a generate FLATTEN(REGEX_EXTRACT_ALL(line,'(.*)[,]["](.*)["][,](.*)[,]["](.*)["][,](.*)'));

c = foreach b generate $0,REPLACE($1,',',' '),$2,REPLACE($3,',',' '),$4;
dump c;