[James-ZHANG]

Fails gracefully when allocation fails during array/map construction

Description

Return values of cbor_array_push, _cbor_map_add_key and _cbor_map_add_value are unchecked. An unchecked false return value here implies potential out-of-bound memory access.

Checklist

• I have read followed CONTRIBUTING.md
  • I have added tests
  • I have updated the documentation
  • I have updated the CHANGELOG
• Are there any breaking changes? If so, are they documented?
  No
• Does this PR introduce any platform specific code? If so, is this captured in the description?
  No
• Security: Does this PR potentially affect security? If so, is this captured in the description?
  No
• Performance: Does this PR potentially affect performance? If so, is this captured in the description?
  No

[PJK]

Apologies for the delay and thanks a ton for the PR!

Isn't there still a problem with the memory cleanup? I think that when a nested _cbor_builder_append fails, all the partially built items on the ctx->stack should be freed up, but creation_failed is never checked as we "unwind the stack"

[James-ZHANG]

all the partially built items on the ctx->stack should be freed up

Those are already freed up in cbor_load.

I think the invariant of _cbor_builder_append is it shouldn't have any memory leak: the passed item either becomes part of stack->root; or in case of failure, gets freed properly (and in that case it leaves partially built stack items behind).

Added a function comment emphasizing that.

[PJK]

Makes sense, thank you! This could use better docs, filed #226