Hi Knut,
If you think it's a bug start by raising a ticket then post about it on
the chat.
2c that may help you
All inputs need to be sorted.
I highly recommend that the sorting is done through the sort step as
opposed through the DB
That your chosen sorting method is set through that step (e.g. case
sensitive or not. Collator strength)
Diego
On 30/5/2024 2:18 am, Knut Meyer wrote:
Hi togehter,
In my current project, I think I have stumbled across a bug. Before I
create a ticket, I would like to clarify briefly whether I am wrong in
my assessment.
I use the denormaliser to turn several lines into one. A string field
containing article numbers is used as the group. Despite differences
in capitalisation (specifically ‘LOc 0101’ and ‘LOC 0101’), these two
lines are grouped unexpectedly. I was able to avoid this by creating a
checksum, which was also included in the grouping.
The implementation uses rowMeta.compare to compare the grouping
fields. See line 291
https://github.com/apache/hop/blob/7921253c2d18526e6830ef17d740b952460cc2d3/plugins/transforms/denormaliser/src/main/java/org/apache/hop/pipeline/transforms/denormaliser/Denormaliser.java#L288C3-L292C4
Am I right in thinking that capitalisation should make a difference?
Best regards,
Knut
--
<https://www.bizcubed.com.au> Diego Mainou
Product Manager
M. +61 415 152 091
E. diego.mai...@bizcubed.com.au
www.bizcubed.com.au <https://www.bizcubed.com.au>
<https://www.bizcubed.com.au>