On Sat, Jul 4, 2026 at 4:59 AM Florents Tselai <[email protected]> wrote:
> > > > On Sat, Jun 27, 2026 at 8:40 AM Florents Tselai <[email protected]> > wrote: > >> >> >> On Fri, Jun 26, 2026 at 11:19 PM Zsolt Parragi <[email protected]> >> wrote: >> >>> + binary_jbv.type = jbvBinary; >>> + binary_jbv.val.binary.data = >>> &jb_result->root; >>> + binary_jbv.val.binary.len = >>> VARSIZE(jb_result); >>> >>> Shouldn't that use VARSIZE_ANY_EXHDR or VARSIZE(jb_result) - VARHDRSZ? >>> >> >> I think you're right; let's see if Corey comes back with anything else >> before I send out a new version. >> > > Here's a v5 that fixes the VARSIZE_ANY_EXHDR Zsolt mentioned above > I've also rearranged the test cases in jsonpath.sql and jsonb_jsonpath.sql > Also added parser error test cases that were missing for translate() in > jsonpath.sql > Apologies for the delay. jsonpath.h: I see that all new enum values are now addedto the end. Good. jsonpath.c: I see that the new cases are clustered together where it makes sense. The individual cases are a bit jumbled when viewed all at once, so while I don't think this makes for bad end-result code, I do think the patch should be split into 3, one for each new function, or at least split translate() out to its own patch if you think the code that split/join share makes separating them inelegant. The more morsel-sized the patch(es), the easier it is for the committer. jsonb_jsonpath.out/sql: Comment on empty string delimiter test makes sense now. The silent mode tests seem adequate. I see that the commitfest entry has been marked ready for committer. I think that's slightly premature, so I'm updating the status for now.
