Closed fei-cheng closed 5 years ago
@fei-cheng what version of SMV and Spark?
I suspect this is due to the split on ';' we do on the input schema string (either input from file or direct string). We can create a dummy sentinel to represent a semicolon (e.g. @delimiter = semi-colon
) and then translate that sentinel to a real ';' on read/write.
I tested on both SMV 2.1.1.1 and the latest version installed by 'pip install', they both have the same error.
Pretty sure you can fix it here: https://github.com/TresAmigosSD/SMV/blob/master/src/main/scala/org/tresamigos/smv/SmvSchema.scala#L490
@fei-cheng you want to give it a try?
and here for the parsing part: https://github.com/TresAmigosSD/SMV/blob/master/src/main/scala/org/tresamigos/smv/SmvSchema.scala#L466
@ninjapapa I will try to fix it
When I changed the delimiter of
CB1200CZ11.csv
from|
to;
, and changed the@delimiter = ;
accordingly, I got below error while running the module in demo.Run command:
smv-run -m stage1.employment.EmploymentByState
Error log: