* json custom serialization; application for strtabs
* serialize using nesting
* make toJson more feature complete
* add since
* Revert "Improve JSON serialisation of strtabs (#14549)"
This reverts commit 7cb4ef26ad.
* better approach via mixin
* toJson, jsonTo
* fix test
* address comments
* move to jsonutils
* doc
* cleanups
* also test for js
* also test for vm
* Error -> Defect for defects
The distinction between Error and Defect is subjective,
context-dependent and somewhat arbitrary, so when looking at an
exception, it's hard to guess what it is - this happens often when
looking at a `raises` list _without_ opening the corresponding
definition and digging through layers of inheritance.
With the help of a little consistency in naming, it's at least possible
to start disentangling the two error types and the standard lib can set
a good example here.
* base `parseEnum` on a case statement, fixes#14030
* apply simplifactions / clean up, remove `norm` node, use strVal
* export `normalize` in json.nim
* cmp using nimIdentNormalize, error at CT if ambiguous enum found
`nimIdentNormalize` provided by @cooldome.
We track all names of the branches we have created so far and error if
a duplicate is found.
Dummy change to make github react...
* fix docstring of `nimIdentNormalize`
* make `typ` arg `typedesc`, add lineinfo, call norm. only once
* fixes#12015 by also checking kind of `typeNode`
If a tuple field is aliased it'll appear the same as a ref type in a
call to `getType` if only for the kind of the resulting `NimNode` is
checked (that is a `nnkBracketExpr`)
* fix test case due to #12017 and add more realistic test case
Adds an additional test case, which includes generics and is closer to
the real failure I encountered
* remove previous fix and fix differently after all
The previous fix was incomplete, because it failed for generics.
Note that the `of "tuple"` is not actually needed, the
`nnkBracketExpr` branch in the `else` branch would catch it too, but I
decided to introduce it for clarity. However, the latter is actually
needed, because it seems for aliases of `seq` we end up in it.
* update comment about global `%` proc in json test
* Tighten the conversion from tyRange to scalar types.
Introduce the `isIntConv` rule for unsigned types.
Do not allow mixed-signedness conversions between ranges and scalar types.
* More json adjustments
