[netrexx-pipelines] digest - revisited

[René J. <rvj...@xs...>]

Hi Jeff,

this was left untouched for longer than I wanted. Current state is, DIGEST does what I needed it for and can do different cryptohashes with demonstrable correctness when e.g. comparing to the wikipedia reference examples:

➜  netrexx-code git:(master) ✗ pipe "literal The quick brown fox jumps over the lazy dog | digest md5 | cons"
9e107d9d372bb6826bd81d3542a419d6
➜  netrexx-code git:(master) ✗ pipe "literal The quick brown fox jumps over the lazy dog. | digest md5 | cons"
e4d909c290d0fb1ca068ffaddf22cbd0

But: CMS returns characters and we return a hex string. Problem is that the characters on CMS are always EBCDIC and we do Unicode. For compatibility, are we expected to output EBCDIC on all platforms? Also, the trailing space is tricky on CMS, we have no real answer to that at the moment but it has never bitten.

Things being as they are - like the amount of time I can invest - I am inclined to just document the difference and leaving it like this. I am convinced most people want a hex string anyway, because that is how digests are used.

p.s. I did a small change to the SQLSELECT stage - if it is DB2, you can’t override set properties with the url. Also, the name of the jdbc driver was seriously out of date.

best regards,

René.