hadoop - pig 自定义函数加载多个字符 ^^(双胡萝卜)定界符

Question

我是 PIG 的新手，有人可以帮助我如何加载一个包含多个字符(在我的例子中为“^^”)作为列分隔符的文件。

例如我有以下列的文件aisforapple^^bisforball^^cisforcat^^disfordoll^^andeisforelephantfisforfish^^gisforgreen^^hisforhat^^iisforicecreem^^andjisforjarkisforking^^lisforlion^^misformango^^nisfornose^^andoisfororange

问候

Answer 1

正则表达式最适合这种多字符

input.txt
aisforapple^^bisforball^^cisforcat^^disfordoll^^andeisforelephant
fisforfish^^gisforgreen^^hisforhat^^iisforicecreem^^andjisforjar
kisforking^^lisforlion^^misformango^^nisfornose^^andoisfororange

PigScript
A = LOAD 'input.txt' AS line;
B = FOREACH A GENERATE FLATTEN(REGEX_EXTRACT_ALL(line,'(.*)\\^\\^(.*)\\^\\^(.*)\\^\\^(.*)\\^\\^(.*)')) AS (f1,f2,f3,f4,f5);
DUMP B;

Output:
(aisforapple,bisforball,cisforcat,disfordoll,andeisforelephant)
(fisforfish,gisforgreen,hisforhat,iisforicecreem,andjisforjar)
(kisforking,lisforlion,misformango,nisfornose,andoisfororange)

解释:

For better understanding i break the regex into multiple lines
(.*)\\^\\^ ->Any character match till ^^ and stored into f1,(double backslash for special characters) 
(.*)\\^\\^ ->Any character match till ^^ and stored into f2,(double backslash for special characters) 
(.*)\\^\\^ ->Any character match till ^^ and stored into f3,(double backslash for special characters) 
(.*)\\^\\^ ->Any character match till ^^ and stored into f4,(double backslash for special characters) 
(.*)       ->Any character match till the end of string and stored into f5