Am Montag, 19. September 2016 09:02:12 UTC+2 schrieb Noel Grandin:
> Looks like we actually store it as a map of <index keys...., row_no> ->
That's an interesting design choice. It avoids the problem I've outlined
above. However, the price to pay is that the ability to search for "partial
keys" in the map is required. For example, you know the index key, but you
don't know the row number (otherwise you would not need to do a search).
This is rather untypical for key-value formats. Thanks for the pointer to
the code, I've taken a first look at it (and I spotted the place where the
map keys are built; you are absolutely right about the format), but I will
need to analyze it a bit more.
You received this message because you are subscribed to the Google Groups "H2
To unsubscribe from this group and stop receiving emails from it, send an email
To post to this group, send email to firstname.lastname@example.org.
Visit this group at https://groups.google.com/group/h2-database.
For more options, visit https://groups.google.com/d/optout.