hadoop - 位置变量上的 pig 过滤器

Question

我正在尝试过滤位置变量。

X = FILTER C BY($14 matches '.*USD.*');
STORE X into '$output' using PigStorage(',');

上面的语句不起作用，但如果我尝试只输出 $14

E = FOREACH C GENERATE FLATTEN($14);
STORE C into '$output' using PigStorage(',');

它工作正常

样本数据:

304a285281be,1383027928890968764,receiver,10C,655362,C2,USD811289,1,0,0,ebay_checkout,cc,cc,USD2659,USD120
304a285281be,1383027928890968764,receiver,10C,655362,C2,USD811289,1,0,0,ebay_checkout,cc,cc,USD2659,USD0
304a285281be,1383027928890968764,receiver,10C,655362,C2,USD811289,1,0,0,ebay_checkout,cc,cc,USD2659,GBP0

样本输出

304a285281be,1383027928890968764,receiver,10C,655362,C2,USD811289,1,0,0,ebay_checkout,cc,cc,USD2659,USD0
304a285281be,1383027928890968764,receiver,10C,655362,C2,USD811289,1,0,0,ebay_checkout,cc,cc,USD2659,GBP0

Answer 1

在 'BY' 和 '(' 之间添加一个空格

    X = FILTER C BY (FLATTEN($14) matches '.*USD.*');
    STORE X into '$output' using PigStorage(',');