fix/ref rfc 9535 #215
fix/ref rfc 9535 #215
Conversation
Signed-off-by: Vincent Biret <vincentbiret@hotmail.com>
Signed-off-by: Vincent Biret <vincentbiret@hotmail.com>
|
|
||
| Users SHALL refrain from using additional capabilities offered by the tool's JSONPath implementation to avoid any interoperability issues with their Overlay Documents. | ||
|
|
||
| In case the tool being use is not fully compliant with RFC9535, users MIGHT have to update some JSONPath query expressions to accommodate for the implementation in use. |
There was a problem hiding this comment.
I'm not sure this statement is needed / appropriate. When a tool doesn't comply with this spec, I think it would be better to recommend that users find a compliant tool than adjust their usage to fit the non-compliant tool.
|
|
||
| ### Example of updates to JSONPath | ||
|
|
||
| This example JSONPath query expression: | ||
|
|
||
| ```jsonpath | ||
| $.paths.*.get.parameters[?@.name=='filter' && @.in=='query'] | ||
| ``` | ||
|
|
||
| Might require additional optional parenthesis with some implementations like so: | ||
|
|
||
| ```jsonpath | ||
| $.paths.*.get.parameters[?(@.name=='filter' && @.in=='query)'] | ||
| ``` | ||
|
|
There was a problem hiding this comment.
This too seems inappropriate for the spec, though I agree it is useful information. Could we put this in the README instead, as a subsection of "Tools that Support Overlays"?
There was a problem hiding this comment.
I also considered putting this as an appendix in the beginning of writing this. But since I have parts that are normative above, I put it in the full specification. Maybe a better approach would be to keep the informational parts and the example as an appendix, and get rid of the normative parts of this section?
|
|
||
| In case the tool being use is not fully compliant with RFC9535, users MIGHT have to update some JSONPath query expressions to accommodate for the implementation in use. | ||
|
|
||
| ### Example of updates to JSONPath |
There was a problem hiding this comment.
I would call this a restriction and not an "update" 😎
This whole section should rather go into the README, as proposed by @mikekistler.
Co-authored-by: Ralf Handl <ralf.handl@gmail.com>
3 participants
fixes #88