[PIG-1664] [zebra] leading '_' in directory/file names should be ignored; the "pigtest" build target should include all pig-related zebra tests.

• Type: Bug
• Status: Closed
• Priority: Trivial
• Resolution: Fixed
• Affects Version/s: 0.6.0, 0.7.0
• Fix Version/s: 0.8.0
• Assignee: Yan Zhou
• Reporter: Yan Zhou
• Created: 04/Oct/10 19:56
• Updated: 17/Dec/10 22:46
• Resolved: 04/Oct/10 22:10

Description

Disk entries of names prefixed with '_' is excluded during split generation, and should be excluded from CG's data file processing too. Furthermore, FileInputFormat treats those files as hidden files. Zebra should ignore them to be safe.

At writing, zebra creates temporary directory under the table (PIG-1115). The directory may not be totally cleaned up due to lagging speculative executions. To avoid any confusions resulting from attempt to access the _temporary directory, the disk entries of names with the '' prefix should also be excluded by Zebra processing.

On the other hand, now that Zebra's pig tests run much faster using PIG's local mode, all pig-related should be included in the "pigtest" target to complete PIG's test coverage.