I am a bit confused about the issues. AFAIK, if whether the wildcard works or not depends upon the support of your storage format. For example, hdfs, local file system, etc.
2014-09-02 21:57 GMT+08:00 Jordan Kaye <[email protected]>: > I’m having issues using wildcard input paths in Pig. > > > If I run the following commands: > > > A = load ‘something/*.csv’ using PigStorage(‘,’) > > Dump A; > > > I see the output from all csv files in the something folder. > > If, however, I run a store instead: > > > A = load ‘something/*.csv’ using PigStorage(‘,’) > > Store A into ‘somedestination’; > > > The job fails with the following error message: > > Input(s): > > Failed to read data from “/something/*.csv” > > > It looks like the store is attempting to load from the literal path instead > of globbing using the wildcard, but if that’s the case then why does it > work during the dump? Is there another way to accomplish this? > -- Bing Jiang
