CBOR: Improve recognition of CBOR data format and update to Jackson 2.4.3 #8637
Conversation
|
@s1monw I think you wanted this ASAP |
|
|
||
| for (int i = 0; i < length; i++) { | ||
| for (int i = 1; i < length; i++) { |
There was a problem hiding this comment.
for other reviewers wondering where this change comes from, the first char is already checked earlier in this method
|
Just left one question/comment about the detection of short json documents, otherwise it looks good! |
| } | ||
| if (first == (CBORConstants.BYTE_OBJECT_INDEFINITE & 0xff)){ | ||
|
|
||
| if (isCBORObjectHeader((byte)first, (byte)second, (byte)third, (byte)fourth)) { |
There was a problem hiding this comment.
Maybe you should also check that third and fourth are different from -1, otherwise this indicator of a missing byte (-1) is going to be converted to a legal byte (255) via the cast?
There was a problem hiding this comment.
Agreed. Adjusted so that it also includes the loop as well.
|
@pickypg Just left a minor comment, I think it's close |
|
LGTM |
pickypg
force-pushed
the
feature/recognize-cbor-7640
branch
from
18ef4ba
to
d5e5eee
Compare
- Update pom to 2.4.3 from 2.4.2 - Enable the CBOR data header (aka tag) from the CBOR Generator to provide binary identification like the Smile format - Check for the CBOR header and ensure that the data sent in represents a "major type" that is an object - Cleans up `JsonVsCborTests` unused imports
pickypg
force-pushed
the
feature/recognize-cbor-7640
branch
from
d5e5eee
to
488b4ff
Compare
|
Merged 7523d0b (forgot to change message to |
clintongormley
changed the title
clintongormley
changed the title
clintongormley
added
v1.3.6
v1.5.0
and removed
v1.3.6
v2.0.0-beta1
v1.5.0
v1.4.1
labels
|
PR reverted for now |
|
@pickypg any news on this? |
|
@s1monw Nothing substantial. It looks like I'll have time today to dive deeper. |
JsonVsCborTestsunused importsCloses #7640