Dont track unknown props in buffer if ignoreAllUnknown is true
#3082
Conversation
|
Sounds good, in general. Performance optimization would be nice. My only/main concern is that beside regular well-defined properties, there are at least 2 kinds of "unknown" property types:
and specifically I am not 100% sure if there is test coverage; for cases where "ignore all unknown" is enabled and one of above is used. Anyway: it would be great if you could add a new test under |
|
ok, I added a few test cases. |
|
@dhofftgt Excellent, thank you & apologies for slow follow up here. The only thing then, before I merge this in, would be CLA (unless I have gotten one before -- just let me know if so, I didn't notice one). If needed, it's a 1-page doc here: https://github.com/FasterXML/jackson/blob/master/contributor-agreement.pdf and the common way is to print it, fill & sign, scan/photo, email to Thank you once again for contributing this. |
|
@cowtowncoder I work at Target and I had them sign a CCLA. Do I still need to sign an individual CLA? |
|
Nope: CCLA is fine. |
|
@dhofftgt I assume CCLA will be sent in near future, and if so, that's fine and I'll keep checking my fasterxml.com account. But just in case it was sent earlier, I don't think I have seen it: I mention this because occasionally clas have ended up on gmail spam folder for some reason only known to gods of gmail. :) |
|
@cowtowncoder should have been sometime around 3/16 when it was sent. I will follow up today to see how/when it was sent or resend it. |
|
@cowtowncoder Just heard back, it was sent to clas@fasterxml.com. It's being forwarded to info@fasterxml.com now. |
|
Ah. Yes, received it. Thanks! |
cowtowncoder
changed the title
ignoreAllUnknown is true
…rXML#3082) Dont track unknown props in buffer if ignoreAllUnknown is true. (cherry picked from commit 23f9c61)
(cherry picked from commit 027e795)
In performance testing I was seeing that keeping track of unknown properties had an impact. If ignoreUnknownProperties is true, this buffer is ignored anyway.