closes https://github.com/nim-lang/RFCs/issues/554
Adds a symmetric difference operation to the language bitset type. This
maps to a simple `xor` operation on the backend and thus is likely
faster than the current alternatives, namely `(a - b) + (b - a)` or `a +
b - a * b`. The compiler VM implementation of bitsets already
implemented this via `symdiffSets` but it was never used.
The standalone binary operation is added to `setutils`, named
`symmetricDifference` in line with [hash
sets](https://nim-lang.org/docs/sets.html#symmetricDifference%2CHashSet%5BA%5D%2CHashSet%5BA%5D).
An operator version `-+-` and an in-place version like `toggle` as
described in the RFC are also added, implemented as trivial sugar.
fixes#24288, refs #23625
Since #23625 "cannot evaluate" errors during VM code generation are
"soft" errors with `nim check`, i.e. the code generation isn't halted
(except for the current proc which `return`s which can cause wrong
codegen) and the expression is still attempted to be evaluated. Now,
these errors signal to the VM that the current generated VM code cannot
be evaluated, and so instead of evaluating, an error node is returned.
This keeps the benefit of the "soft" errors without potentially crashing
the compiler on improperly generated VM code. Although maybe the
compiler might not be able to handle the generated error node in some
cases.
This fixes the chame example in #24288 but this is not tested in CI.
Presumably it or the compiler was doing something like `compiles()` on
code that can't run in the VM.
I would accept nicer ways of tracking non-evaluability than
`c.cannotEval = true` but I tried to keep it as harmless as possible.
fixes#24097
For `nkConv` addresses where the conversion is between 2 types that are
equal between backends, treat assignments the same as assignments to the
argument of the conversion. In the VM this seems to be in `genAsgn` and
`genAsgnPatch`, as evidenced by the special logic for `nkDerefExpr` etc.
This doesn't handle ranges after #24037 because `sameBackendType` is
used and not `sameBackendTypeIgnoreRange`. This is so this is
backportable without #24037 and another PR can be opened that implements
it for ranges and adds tests as well. We can also merge
`sameBackendTypeIgnoreRange` with `sameBackendType` since it doesn't
seem like anything that uses it would be affected (only cycle checks and
the VM), but then we still have to add tests.
fixes#21353
```nim
result = newNodeIT(nkTupleConstr, info, t)
result.add(newNodeIT(nkNilLit, info, t))
result.add(newNodeIT(nkNilLit, info, t))
```
The old implementation uses `t` which is the type of the closure
function as its type. It is not correct and generates ((nil, nil), (nil,
nil)) for `default(closures)`. This PR creates `(tyPointer, tyPointer)`
for fake closure types just like what cctypes do.
Follow up of #23927 which solves the build error.
This is still only a partial fix as it doesn't take into account
unordered enums. I'll make a separate issue for those.
---------
Co-authored-by: ringabout <43030857+ringabout@users.noreply.github.com>
`reset`, `wasMoved` and `move` doesn't support primitive types, which
generate `null` for these types. It is now produce `x = default(...)` in
the backend. Ideally it should be done by ast2ir in the future
Theoretical Benefits / Plans:
- Typed assembler-like language.
- Allows for a CPS transformation.
- Can replace the existing C backend by a new C backend.
- Can replace the VM.
- Can do more effective "not nil" checking and static array bounds
checking.
- Can be used instead of the DFA.
- Easily translatable to LLVM.
- Reasonably easy to produce native code from.
- Tiny memory consumption. No pointers, no cry.
**In very early stages of development.**
Todo:
- [x] Map Nim types to IR types.
- [ ] Map Nim AST to IR instructions:
- [x] Map bitsets to bitops.
- [ ] Implement string cases.
- [ ] Implement range and index checks.
- [x] Implement `default(T)` builtin.
- [x] Implement multi string concat.
- [ ] Write some analysis passes.
- [ ] Write a backend.
- [x] Integrate into the compilation pipeline.
Since they are integer types, by mean of allowing cast integer to enums
in VM, we can suppress some enum warnings in the stdlib in the unified
form, namely using a cast expression.
Attempts to move the generic instantiation to the module that uses it.
This should decrease re-compilation times as the source module where the
generic lives doesnt need to be recompiled
---------
Co-authored-by: ringabout <43030857+ringabout@users.noreply.github.com>
Co-authored-by: Andreas Rumpf <rumpf_a@web.de>
* fixes#21704; remove `nfIsRef` for genLit
* remove nfIsRef from the output of macros
* make the logic better
* try again
* act together
* excl nfIsRef
* refactoring in preparation for better, simpler name mangling that works with IC flawlessly
* use new disamb field
* see if this makes tests green
* make tests green again
* test implicitly huge set types
refs https://github.com/nim-lang/RFCs/issues/298
* oh my god
* boot at least
* don't error, fix remaining issues, no 2 len arrays
* fix runnable example
* test assuming 0..255 for int literal
* test refactor, add changelog, test
* fixes#10938; always initialize global variable in VM
* fixes importc vars
* there is a pre-existing issue regarding closure types in the VM
* add tests
* fixes regression #17121; adding doc comment in importc proc makes it silently noop at CT
* Update compiler/vmgen.nim
Co-authored-by: Andreas Rumpf <rumpf_a@web.de>
* Added 'openarray[char]' overloads to 'std/parseutils'
* Removed redundant `start` and `last` params from slice using procs
* Fixed type for parseIdent overload
* fixed one by off with 'substr'
* removed missed start parameters for procedures
* Added 'openarray[char]' overloads to 'std/parseutils'
* Removed redundant `start` and `last` params from slice using procs
* Fixed type for parseIdent overload
* fixed one by off with 'substr'
* removed missed start parameters for procedures
* Fixed VM op to work with new 'opcSlice'
* Corrected captureBetween's logic to work with openarray
* js sys's parsefloat logic now uses openarray
Co-authored-by: Clay Sweetser <Varriount@users.noreply.github.com>
* Implemented opcSlice to make 'toOpenArray' work on the VM
* Added nkOpenArray for VM to reduce bodgeness
* Fixed range issues and erraneous comments
* Range check correctly for openArrays in opcLdArr
* Inverted logic for ldArr checking
* vm now supports slicing strings
* Added string tests
* Removed usage of 'nkOpenArray' and redundant operations
* Refactored vmSlice implementation, removing redundant and incorrect code
* Made tuples go throw opcWrObj for field assignment
* All strkinds should be considered for openarrays
