fixes#24898
A type is only overwritten if it is definitely a forward type, partial
object (symbol marked `sfForward`) or a magic type. Maybe worse for
performance but should be more correct. Another option might be to
provide a different value for `prev` for the `preserveSym` case but then
we cannot easily ignore only nominal type nodes.
(cherry picked from commit d966ee3fc3)
fixes#5631, fixes#8938, fixes#18855, fixes#19271, fixes#23885,
fixes#24877
`isTupleRecursive`, previously only called to give an error for illegal
recursions for:
* tuple fields
* types declared in type sections
* explicitly instantiated generic types
did not check for recursions in proc types. It now does, meaning proc
types now need a nominal type layer to recurse over themselves. It is
renamed to `isRecursiveStructuralType` to better reflect what it does,
it is different from a recursive type that cannot exist due to a lack of
pointer indirection which is possible for nominal types.
It is now also called to check the param/return types of procs, similar
to how tuple field types are checked. Pointer indirection checks are not
needed since procs are pointers.
I wondered if this would lead to a slowdown in the compiler but since it
only skips structural types it shouldn't take too many iterations, not
to mention only proc types are newly considered and aren't that common.
But maybe something in the implementation could be inefficient, like the
cycle detector using an IntSet.
Note: The name `isRecursiveStructuralType` is not exactly correct
because it still checks for `distinct` types. If it didn't, then the
compiler would accept this:
```nim
type
A = distinct B
B = ref A
```
But this breaks when attempting to write `var x: A`. However this is not
the case for:
```nim
type
A = object
x: B
B = ref A
```
So a better description would be "types that are structural on the
backend".
A future step to deal with #14015 and #23224 might be to check the
arguments of `tyGenericInst` as well but I don't know if this makes
perfect sense.
(cherry picked from commit 7f0e07492f)
fixes#24887 (really just this [1 line
commit](632c7b3397)
would have been enough to fix the issue but it would ignore the general
problem)
When a type definition is encountered where the symbol already has a
type (not a forward type), the type is left alone (not reset to
`tyForward`) and the RHS is handled differently: The RHS is still
semchecked, but the type of the symbol is not updated, and nominal type
nodes are ignored entirely (specifically if they are the same kind as
the symbol's existing type but this restriction is not really needed).
If the existing type of the symbol is an enum and and the RHS has a
nominal enum type node, the enum fields of the existing type are added
to scope rather than creating a new type from the RHS and adding its
symbols instead.
The goal is to prevent any incompatible nominal types from being
generated during resem as in #24887. But it also restricts what macros
can do if they generate type section AST, for example if we have:
```nim
type Foo = int
```
and a macro modifies the type section while keeping the symbol node for
`Foo` like:
```nim
type Foo = float
```
Then the type of `Foo` will still remain `int`, while it previously
became `float`. While we could maybe allow this and make it so only
nominal types cannot be changed, it gets even more complex when
considering generic params and whether or not they get updated. So to
keep it as simple as possible the rule is that the symbol type does not
change, but maybe this behavior was useful for macros.
Only nominal type nodes are ignored for semchecking on the RHS, so that
cases like this do not cause a regression:
```nim
template foo(): untyped =
proc bar() {.inject.} = discard
int
type Foo = foo()
bar() # normally works
```
However this specific code exposed a problem with forward type handling:
---
In specific cases, when the type section is undergoing the final pass,
if the type fits some overly general criteria (it is not an object,
enum, alias or a sink type and its node is not a nominal type node), the
entire RHS is semchecked for a 2nd time as a standalone type (with `nil`
prev) and *maybe* reassigned to the new semchecked type, depending on
its type kind. (for some reason including nominal types when we excluded
them before?) This causes a redefinition error if the RHS defines a
symbol.
This code goes all the way back to the first commit and I could not find
the reason why it was there, but removing it showed a failure in
`thard_tyforward`: If a generic forward type is invoked, it is left as
an unresolved `tyGenericInvocation` on the first run. Semchecking it
again at the end turns it into a `tyGenericInst`. So my understanding is
that it exists to handle these loose forward types, but it is way too
general and there is a similar mechanism `c.skipTypes` which is supposed
to do the same thing but doesn't.
So this is no longer done, and `c.skipTypes` is revamped (and renamed):
It is now a list of types and the nodes that are supposed to evaluate to
them, such that types needing to be updated later due to containing
forward types are added to it along with their nodes. When finishing the
type section, these types are reassigned to the semchecked value of
their nodes so that the forward types in them are fully resolved. The
"reassigning" here works due to updating the data inside the type
pointer directly, and is how forward types work by themselves normally
(`tyForward` types are modified in place as `s.typ`).
For example, as mentioned before, generic invocations of forward types
are first created as `tyGenericInvocation` and need to become
`tyGenericInst` later. So they are now added to this list along with
their node. Object types with forward types as their base types also
need to be updated later to check that the base type is correct/inherit
fields from it: For this the entire object type and its node are added
to the list. Similarly, any case where whether a component type is
`tyGenericInst` or `tyGenericInvocation` matters also needs to cascade
this (`set` does presumably to check the instantiated type).
This is not complete: Generic invocations with forward types only check
that their base type is a forward type, but not any of their arguments,
which causes #16754 and #24133. The generated invocations also need to
cascade properly: `Foo[Bar[ForwardType]]` for example would see that
`Bar[ForwardType]` is a generic invocation and stay as a generic
invocation itself, but it might not queue itself to be updated later.
Even if it did, only the entire type `Foo[Bar[ForwardType]]` needs to be
queued, updating `Bar[ForwardType]` by itself would be redundant or it
would not change anything at all. But these can be done later.
(cherry picked from commit 525d64fe88)
Yet another one of these. Multiple changes piled up in this one. I've
only minimally cleaned it for now (debug code is still here etc). Just
want to start putting this up so I might get feedback. I know this is a
lot and you all are busy with bigger things. As per my last PR, this
might just contain changes that are not ready.
### concept instantiation uniqueness
It has already been said that concepts like `ArrayLike[int]` is not
unique for each matching type of that concept. Likewise the compiler
needs to instantiate a new proc for each unique *bound* type not each
unique invocation of `ArrayLike`
### generic parameter bindings
Couple of things here. The code in sigmatch has to give it's bindings to
the code in concepts, else the information is lost in that step. The
code that prepares the generic variables bound in concepts was also
changed slightly. Net effect is that it works better.
I did choose to use the `LayedIdTable` instead of the `seq`s in
`concepts.nim`. This was mostly to avoid confusing myself. It also
avoids some unnecessary movings around. I wouldn't doubt this is
slightly less performant, but not much in the grand scheme of things and
I would prefer to keep things as easy to understand as possible for as
long as possible because this stuff can get confusing.
### various fixes in the matching logic
Certain forms of modifiers like `var` and generic types like
`tyGenericInst` and `tyGenericInvocation` have logic adjustments based
on my testing and usage
### signature matching method adjustment
This is the weird one, like my last PR. I thought a lot about the
feedback from my last attempt and this is what I came up with. Perhaps
unfortunately I am preoccupied with a slight grey area. consider the
follwing:
```nim
type
C1 = concept
proc p[T](s: Self; x: T)
C2[T] = concept
proc p(s: Self; x: T)
```
It would be temping to say that these are the same, but I don't think
they are. `C2` makes each invocation distinct, and this has important
implications in the type system. eg `C2[int]` is not the same type as
`C2[string]` and this means that signatures are meant to accept a type
that only matches `p` for a single type per unique binding. For `C1` all
are the same and the binding `p` accepts multiple types. There are
multiple variations of this type classes, `tyAnything` and the like.
The make things more complicated, an implementation might match:
```nim
type
A = object
C3 = concept
proc p(s: Self; x: A)
```
if the implementation defines:
```nim
proc p(x: Impl; y: object)
```
while a concept that fits `C2` may be satisfied by something like:
```nim
proc p(x: Impl; y: int)
proc spring[T](x: C2[T])
```
it just depends. None of this is really a problem, it just seems to
provoke some more logic in `concepts.nim` that makes all of this (appear
to?) work. The logic checks for both kinds of matches with a couple of
caveats. The fist is that some unbind-able arrangements may be matched
during overload resolution. I don't think this is avoidable and I
actually think this is a good way to get a failed compilation. So, first
note imo is that failing during binding is preferred to forcing the
programming to write annoying stub procs and putting insane gymnastics
in the compiler. Second thing is: I think this logic is way to accepting
for some parts of overload resolutions. Particularly in `checkGeneric`
when disambiguation is happening. Things get hard to understand for me
here. ~~I made it so the implicit bindings to not count during
disambiguation~~. I still need to test this more, but the thought is
that it would help curb excessive ambiguity errors.
Again, I'm sorry for this being so many changes. It's probably
inconvenient.
---------
Co-authored-by: Andreas Rumpf <rumpf_a@web.de>
(cherry picked from commit dfab30734b)
fixes#24743, refs #24718
We cannot do this in general for any expression with generic type
because the `typeof` logic is called for things like `type Foo` in:
```nim
type Foo[T] = object
proc init(_: type Foo) = discard
```
We also cannot use `containsUnresolvedType` to work around this specific
case because the base type of `static[auto]` is not unresolved, it is a
typeclass that isn't lifted to a parameter. The behavior of generating
`tyFromExpr` is also consistent with pre-2.0, so we do this in this
special case of `static`.
(cherry picked from commit 569d02e212)
fixes#24715
In generic typechecking, unresolved static param symbols (i.e.
`skGenericParam`) have [the static type
itself](1f8da3835f/compiler/semexprs.nim (L1483-L1485))
as their type when used in an expression. This is not the case when the
static param is resolved (the type is wrapped in static when necessary),
but semchecking of types and generic typechecking expects the type of
the value to be wrapped in `static` (at least `array[N, int]` breaks).
So for now, to solve the issue, `typeof` just skips static types.
(cherry picked from commit 514a25c9a2)
fixes#24644
Another option is to include the symbol names and owners in the type
listing as in #24645 but this is a bit verbose.
(cherry picked from commit 0861dabfa7)
fixes#23952
It reorders `type Foo = enum A, B = -1` to `type Foo = enum B = -1, A`
so that `firstOrd` etc. continue to work.
(cherry picked from commit 294b1566e7)
Previously, the compiler never differentiated between `untyped`/`typed`
argument default values and other types, it considered any parameter
with a type as typed and called `semExprWithType`, which both
typechecked it and disallowed `void` expressions. Now, we perform no
typechecking at all on `untyped` template param default values, and call
`semExpr` instead for `typed` params, which allows expressions with
`void` type.
(cherry picked from commit 4eed341ba5)
fixes#24203
`semTypeNode` is called twice for RHS'es of type sections,
[here](b0e6d28782/compiler/semstmts.nim (L1612))
and
[here](b0e6d28782/compiler/semstmts.nim (L1646)).
Each time `prev` is `s.typ`, but the assertion expects `prev == nil`
which is false since `s.typ` is not nil the second time. To fix this,
the `prev == nil` part of the assertion is removed.
The reason this only happens for types like `seq[int]`, `(int, int)` etc
is because they don't have syms: `semTypeIdent` attempts to directly
[replace the typedesc param
itself](b0e6d28782/compiler/semtypes.nim (L1916))
with the sym of the base type of the resolved typedesc type if it
exists, which means `semTypeNode` doesn't receive the typedesc param sym
to perform the assert.
fixes#11422, refs #8336/#8333, refs #20130
The compiler generates conversion nodes *after* evaluating the branches
of case statements as constants, the reasoning is that case branches
accept constants of different types, like arrays or sets. But this means
that conversion nodes that need to be evaluated like converter calls
don't get evaluated as a constant for codegen. #8336 fixed this by
re-evaluating the node if an `nkHiddenCallConv` was created, and in
#20130 this logic also had to be added for `nkHiddenStdConv` for
cstrings. This logic was only for single case elements, it has now been
added to range elements as well to fix#11422. Additionally, all
conversion nodes are now evaluated for simplicity, but maybe this won't
pass CI.
fixes#24150, refs #22022
An exception is raised in the `semExprWithType` call, which means `dec
c.inTypeofContext` is never called, but `compiles` allows compilation to
continue. This means `c.inTypeofContext` is left perpetually nonzero,
which prevents `compileTime` evaluation for the rest of the program.
To fix this, `defer:` is used for the `dec c.inTypeofContext` call, as
is done for
[`instCounter`](d51d88700b/compiler/seminst.nim (L374))
in other parts of the compiler.
fixes#22661
Range expressions in `of` branches in `case` statements start off as
calls to `..` then become `nkRange` when getting typed. For this reason
the compiler leaves `nkRange` alone when type checking the case
statements again, but it still does the exhaustiveness checking for the
entire case statement, and leaving the range alone means it doesn't
count the values of the range for exhaustiveness. So the counting is now
also done on `nkRange` nodes in the same way as when typechecking it the
first time.
fixes#23506#24065 broke compilation of template parameter default values that
depended on other template parameters. But this was never implemented
anyway, actually attempting to use those default values breaks the
compiler as in #23506. So these are now implemented as well as fixing
the regression.
First, if a default value expression uses any unresolved arguments
(generic or normal template parameters) in a template header, we leave
it untyped, instead of applying the generic typechecking (fixing the
regression). Then, just before the body of the template is about to be
explored, the default value expressions are handled in the same manner
as the body as well. This captures symbols including the parameters, so
the expression is checked again if it contains a parameter symbol, and
marked with `nfDefaultRefsParam` if it does (as an optimization to not
check it later).
Then when the template is being evaluated, when substituting a
parameter, if we try to substitute with a node marked
`nfDefaultRefsParam`, we also evaluate it as we would the template body
instead of passing it as just a copy (the reason why it never worked
before). This way we save time if a default value doesn't refer to
another parameter and could just be copied regardless.
fixes#15959
Another followup of #22029 and #24005, subscript expressions now
recognize when their parameters are generic types, then generating
tyFromExpr. `typeof` also now properly sets `tfNonConstExpr` to make it
usable in proc signatures. `lent` with brackets like `lent[T]` is also
now allowed.
fixes#22342, fixes#22607
Another followup of #22029, `when` expressions in general in generic
type bodies now behave like `nkRecWhen` does since #24042, leaving them
as `tyFromExpr` if a condition is uncertain. The tests for the issues
were originally added but left disabled in #24005.
fixes#16700, fixes#20916, refs #24010
Fixes the instantiation issues for proc param default values encountered
in #24010 by:
1. semchecking generic default param values with `inGenericContext` for
#22029 and followups to apply (the bigger change in semtypes),
2. rejecting explicit generic instantiations with unresolved generic
types inside `inGenericContext` (sigmatch change),
3. instantiating the default param values using `prepareNode` rather
than an insufficient manual method (the bigger change in seminst).
This had an important side effect of references to other parameters not
working since they would be resolved as a symbol belonging to the
uninstantiated original generic proc rather than the later instantiated
proc. There is a more radical way to fix this which is generating ident
nodes with `tyFromExpr` in specifically this context, but instead we
just count them as belonging to the same proc in
`hoistParamsUsedInDefault`.
Other minor bugfixes:
* To make the error message in t20883 make sense, we now give a "cannot
instantiate" error when trying to instantiate a proc generic param with
`tyFromExpr`.
* Object constructors as default param values generated default values
of private fields going through `evalConstExpr` more than once, but the
VM doesn't mark the object fields as `nfSkipFieldChecking` which gives a
"cannot access field" error. So the VM now marks object fields it
generates as `nfSkipFieldChecking`. Not sure if this affects VM
performance, don't see why it would.
* The nkRecWhen changes in #24042 didn't cover the case where all
conditions are constantly false correctly, this is fixed with a minor
change. This isn't needed for this PR now but I encountered it after
forgetting to `dec c.inGenericContext`.
fixes#24041
`when` statements in generic object types normally just leave their
conditions as expressions and still typecheck their branch bodies.
Instead of this, when the condition can be evaluated as a constant as
well as the ones before it and it resolves to `true`, it now uses the
body of that branch without typechecking the remaining ones.
fixes#15314, fixes#24002
The OpenSym behavior first added to generics in #23091 now also applies
to templates, since templates can also capture symbols that are meant to
be replaced by local symbols if the context imports symbols with the
same name, as in the issue #24002. The experimental switch
`templateOpenSym` is added to enable this behavior for templates only,
and the experimental switch `openSym` is added to enable it for both
templates and generics, and the documentation now mainly mentions this
switch.
Additionally the logic for `nkOpenSymChoice` nodes that were previously
wrapped in `nkOpenSym` now apply to all `nkOpenSymChoice` nodes, and so
these nodes aren't wrapped in `nkOpenSym` anymore. This means
`nkOpenSym` can only have children of kind `nkSym` again, so it is more
in line with the structure of symchoice nodes. As for why they aren't
merged with `nkOpenSymChoice` nodes yet, we need some way to signal that
the node shouldn't become ambiguous if other options exist at
instantiation time, we already captured a symbol at the beginning and
another symbol can only replace it if it's closer in scope and
unambiguous.
updated version of #22193
After #22029 and the followups #23983 and #24005 which fixed issues with
it, `tyFromExpr` no longer match any proc params in generic type bodies
but delay all non-matching calls until the type is instantiated.
Previously the mechanism `fauxMatch` was used to pretend that any
failing match against `tyFromExpr` actually matched, but prevented the
instantiation of the type until later.
Since this mechanism is not needed anymore for `tyFromExpr`, it is now
only used for `tyError` to prevent cascading errors and changed to a
bool field for simplicity. A change in `semtypes` was also needed to
prevent calling `fitNode` on default param values resolving to type
`tyFromExpr` in generic procs for params with non-generic types, as this
would try to coerce the expression into a concrete type when it can't be
instantiated yet.
The aliases `tyProxy` and `tyUnknown` for `tyError` and `tyFromExpr` are
also removed for uniformity.
fixes#4228, fixes#4990, fixes#7006, fixes#7008, fixes#8406, fixes
#8551, fixes#11112, fixes#20027, fixes#22647, refs #23854 and #23855
(remaining issue fixed), refs #8545 (works properly now with
`cast[static[bool]]` changed to `cast[bool]`), refs #22342 and #22607
(disabled tests added), succeeds #23194
Parameter and return type nodes in generic procs now undergo the same
`inGenericContext` treatment that nodes in generic type bodies do. This
allows many of the fixes in #22029 and followups to also apply to
generic proc signatures. Like #23983 however this needs some more
compiler fixes, but this time mostly in `sigmatch` and type
instantiations.
1. `tryReadingGenericParam` no longer treats `tyCompositeTypeClass` like
a concrete type anymore, so expressions like `Foo.T` where `Foo` is a
generic type don't look for a parameter of `Foo` in non-generic code
anymore. It also doesn't generate `tyFromExpr` in non-generic code for
any generic LHS. This is to handle a very specific case in `asyncmacro`
which used `FutureVar.astToStr` where `FutureVar` is generic.
2. The `tryResolvingStaticExpr` call when matching `tyFromExpr` in
sigmatch now doesn't consider call nodes in general unresolved, only
nodes with `tyFromExpr` type, which is emitted on unresolved expressions
by increasing `c.inGenericContext`. `c.inGenericContext == 0` is also
now required to attempt instantiating `tyFromExpr`. So matching against
`tyFromExpr` in proc signatures works in general now, but I'm
speculating it depends on constant folding in `semExpr` for statics to
match against it properly.
3. `paramTypesMatch` now doesn't try to change nodes with `tyFromExpr`
type into `tyStatic` type when fitting to a static type, because it
doesn't need to, they'll be handled the same way (this was a workaround
in place of the static type instantiation changes, only one of the
fields in the #22647 test doesn't work with it).
4. `tyStatic` matching now uses `inferStaticParam` instead of just range
type matching, so `Foo[N div 2]` can infer `N` in the same way `array[N
div 2, int]` can. `inferStaticParam` also disabled itself if the
inferred static param type already had a node, but `makeStaticExpr`
generates static types with unresolved nodes, so we only disable it if
it also doesn't have a binding. This might not work very well but the
static type instantiation changes should really lower the amount of
cases where it's encountered.
5. Static types now undergo type instantiation. Previously the branch
for `tyStatic` in `semtypinst` was a no-op, now it acts similarly to
instantiating any other type with the following differences:
- Other types only need instantiation if `containsGenericType` is true,
static types also get instantiated if their value node isn't a literal
node. Ideally any value node that is "already evaluated" should be
ignored, but I'm not sure of a better way to check this, maybe if
`evalConstExpr` emitted a flag. This is purely for optimization though.
- After instantiation, `semConstExpr` is called on the value node if
`not cl.allowMetaTypes` and the type isn't literally a `static` type.
Then the type of the node is set to the base type of the static type to
deal with `semConstExpr` stripping abstract types.
We need to do this because calls like `foo(N)` where `N` is `static int`
and `foo`'s first parameter is just `int` do not generate `tyFromExpr`,
they are fully typed and so `makeStaticExpr` is called on them, giving a
static type with an unresolved node.
fixes#23406, closes#23854, closes#23855 (test code of both compiles
but separate issue exists), refs #23432, follows #23411
In generic bodies, previously all regular `nkCall` nodes like `foo(a,
b)` were directly treated as generic statements and delayed immediately,
but other call kinds like `a.foo(b)`, `foo a, b` etc underwent
typechecking before making sure they have to be delayed, as implemented
in #22029. Since the behavior for `nkCall` was slightly buggy (as in
#23406), the behavior for all call kinds is now to call `semTypeExpr`.
However the vast majority of calls in generic bodies out there are
`nkCall`, and while there isn't a difference in the expected behavior,
this exposes many issues with the implementation started in #22029 given
how much more code uses it now. The portion of these issues that CI has
caught are fixed in this PR but it's possible there are more.
1. Deref expressions, dot expressions and calls to dot expressions now
handle and propagate `tyFromExpr`. This is most of the changes in
`semexprs`.
2. For deref expressions to work in `typeof`, a new type flag
`tfNonConstExpr` is added for `tyFromExpr` that calls `semExprWithType`
with `efInTypeof` on the expression instead of `semConstExpr`. This type
flag is set for every `tyFromExpr` type of a node that `prepareNode`
encounters, so that the node itself isn't evaluated at compile time when
just trying to get the type of the node.
3. Unresolved `static` types matching `static` parameters is now treated
the same as unresolved generic types matching `typedesc` parameters in
generic type bodies, it causes a failed match which delays the call
instantiation.
4. `typedesc` parameters now reject all types containing unresolved
generic types like `seq[T]`, not just generic param types by themselves.
(using `containsGenericType`)
5. `semgnrc` now doesn't leave generic param symbols it encounters in
generic type contexts as just identifiers, and instead turns them into
symbol nodes. Normally in generic procs, this isn't a problem since the
generic param symbols will be provided again at instantiation time (and
in fact creating symbol nodes causes issues since `seminst` doesn't
actually instantiate proc body node types).
But generic types can try to be instantiated early in `sigmatch` which
will give an undeclared identifier error when the param is not provided.
Nodes in generic types (specifically in `tyFromExpr` which should be the
only use for `semGenericStmt`) undergo full generic type instantiation
with `prepareNode`, so there is no issue of these symbols remaining as
uninstantiated generic types.
6. `prepareNode` now has more logic for which nodes to avoid
instantiating.
Subscripts and subscripts turned into calls to `[]` by `semgnrc` need to
avoid instantiating the first operand, since it may be a generic body
type like `Generic` in an expression like `Generic[int]`.
Dot expressions cannot instantiate their RHS as it may be a generic proc
symbol or even an undeclared identifier for generic param fields, but
have to instantiate their LHS, so calls and subscripts need to still
instantiate their first node if it's a dot expression.
This logic still isn't perfect and needs the same level of detail as in
`semexprs` for which nodes can be left as "untyped" for overloading/dot
exprs/subscripts to handle, but should handle the majority of cases.
Also the `efDetermineType` requirement for which calls become
`tyFromExpr` is removed and as a result `efDetermineType` is entirely
unused again.
fixes#23977
The problem is that for *any* body of a generic declaration,
[semstmts](2e4d344b43/compiler/semstmts.nim (L1610-L1611))
sets the sym of its value to the generic type name, and
[semtypes](2e4d344b43/compiler/semtypes.nim (L2143))
just directly gives the referenced type *specifically* when the
expression is a generic body. I'm blaming `semtypes` here because it's
responsible for the type given but the exact opposite behavior
specifically written in makes me think generating an alias type here
maybe breaks something.
fixes#10753, fixes#22021, refs #19365 (was fixed by #22029, but more
faithful test added)
For whatever reason `compileTime` proc calls did not fold if the proc
was generic ([since this folding was
introduced](c25ffbf262 (diff-539da3a63df08fa987f1b0c67d26cdc690753843d110b6bf0805a685eeaffd40))).
I'm guessing the intention was for *unresolved* generic procs to not
fold, which is now the logic.
Non-magic `compileTime` procs also now don't fold at compile time in
`typeof` contexts to avoid possible runtime errors (only the important)
and prevent double/needless evaluation.
refs https://github.com/nim-lang/Nim/pull/23873#discussion_r1687995060,
fixes#23386, fixes#23385, supersedes #23572
Turns the `nfOpenSym` node flag implemented in #23091 and extended in
#23102 and #23873, into a node kind `nkOpenSym` that forms a unary node
containing either `nkSym` or `nkOpenSymChoice`. Since this affects
macros working on generic proc AST, the node kind is now only generated
when the experimental switch `genericsOpenSym` is enabled, and a new
node flag `nfDisabledOpenSym` is set to the `nkSym` or `nkOpenSymChoice`
when the switch is not enabled so that we can give a warning.
Now that the experimental switch has more reasonable semantics, we
define `nimHasGenericsOpenSym2`.
Still have to look this over some. We'll see. I put sink in this branch
simply because I saw `tyVar` there and for no other reason. In any case
the problem appears to be coming from `liftParamType` as it removes the
`sink` type from the formals.
#23869
The following would crash nimsuggest on init:
```nim
import chronos
type
HistoryQuery = object
start: int
limit: int
HistoryResult = object
messages: string
type HistoryQueryHandler* = proc(req: HistoryQuery): Future[HistoryResult] {.async, gcsafe.}
```
fixes#17163, refs #23204
Types that aren't `tyRange` and are bigger than 16 bits, so `int32`,
`uint64`, `int` etc, are disallowed as array index range types.
`tyRange` is excluded because the max array size is backend independent
(except for the specific size of `high(uint64)` which crashes the
compiler) and so there should still be an escape hatch for people who
want bigger arrays.
fixes#23200, fixes#18866#21065 made it so `auto` proc return types remained as `tyAnything` and
not turned to `tyUntyped`. This had the side effect that anything
previously bound to `tyAnything` in the proc type match was then bound
to the proc return type, which is wrong since we don't know the proc
return type even if we know the expected parameter types (`tyUntyped`
also [does not care about its previous bindings in
`typeRel`](ab4278d217/compiler/sigmatch.nim (L1059-L1061))
maybe for this reason).
Now we mark `tyAnything` return types for routines as `tfRetType` [as
done for other meta return
types](18b5fb256d/compiler/semtypes.nim (L1451)),
and ignore bindings to `tyAnything` + `tfRetType` types in `semtypinst`.
On top of this, we reset the type relation in `paramTypesMatch` only
after creating the instantiation (instead of trusting
`isInferred`/`isInferredConvertible` before creating the instantiation),
using the same mechanism that `isBothMetaConvertible` uses.
This fixes the issues as well as making the disabled t15386_2 test
introduced in #21065 work. As seen in the changes for the other tests,
the error messages give an obscure `proc (a: GenericParam): auto` now,
but it does give the correct error that the overload doesn't match
instead of matching the overload pre-emptively and expecting a specific
return type.
tsugar had to be changed due to #16906, which is the problem where
`void` is not inferred in the case where `result` was never touched.
