When we have a collection of files to read with PDI and they all have the same format development is easy: list all files, run a transformation for each one, Text file input, etc. But if different files have different formats and there's no way to tell a priori what the formats may be things get tricky. In this talk we'll see an example of a collection of heterogeneous CSV files being parsed by a single PDI transformation with the Metadata Injection step. Sample files and PDI transformations will be available on github

YOU MAY ALSO LIKE:

• New community Plugin CTE (SkillsCast recorded in March 2015)
• Datascience with PDI/Kettle (SkillsCast recorded in November 2018)
• Using Python for Fraud Detection with PDI (SkillsCast recorded in November 2018)