TODO:
- [ ] test writing of .nif files
- [x] implement loading of fields in PType/PSym that might not have been
loaded
- [ ] implement interface logic
- [ ] implement pragma "replays"
- [ ] implement special logic for `converter`
- [ ] implement special logic for `method`
- [ ] test the logic holds up for `export`
- [ ] implement logic to free the memory of PSym/PType if memory
pressure is high
- [ ] implement logic to close memory mapped files if too many are open.
---------
Co-authored-by: demotomohiro <gpuppur@gmail.com>
Co-authored-by: ringabout <43030857+ringabout@users.noreply.github.com>
Co-authored-by: Jacek Sieka <arnetheduck@gmail.com>
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.
fixes#24091, refs #24092
Any instantiations resolving to a generic body type now gives an error.
Due to #24092, this does not error in cases like matching against `type
M` in generics because generic body type symbols are just not
instantiated. But this prevents parameters with type `type M` from being
used, although there doesn't seem to be any code which does this. Just
in case such code exists, we still allow `typedesc` types resolving to
generic body types.
fixes#24228, refs #22022
As described in
https://github.com/nim-lang/Nim/issues/24228#issuecomment-2392462221,
instantiating generic routines inside `typeof` causes all code inside to
be treated as being in a typeof context, and thus preventing compile
time proc folding, causing issues when code is generated for the
instantiated routine. Now, instantiated generic procs are treated as
never being inside a `typeof` context.
This is probably an arbitrary special case and more issues with the
`typeof` behavior from #22022 are likely. Ideally this behavior would be
removed but it's necessary to accomodate the current [proc `declval` in
the package `stew`](https://github.com/status-im/nim-stew/pull/190), at
least without changes to `compileTime` that would either break other
code (making it not eagerly fold by default) or still require a change
in stew (adding an option to disable the eager folding).
Alternatively we could also make the eager folding opt-in only for
generic compileTime procs so that #22022 breaks nothing whatsoever, but
a universal solution would be better. Edit: Done in #24230 via
experimental switch
split from #24198
This is a required refactor for the only good solution I've been able to
think of for #4858 etc. Explanation:
---
`sigmatch` currently [disables
bindings](d6a71a1067/compiler/sigmatch.nim (L1956))
(except for binding to other generic parameters) when matching against
constraints of generic parameters. This is so when the constraint is a
general metatype like `seq`, the type matching will not treat all
following uses of `seq` as the type matched against that generic
parameter.
However to solve #4858 etc we need to bind `or` types with a conversion
match to the type they are supposed to be converted to (i.e. matching
`int literal(123)` against `int8 | int16` should bind `int8`[^1], not
`int`). The generic parameter constraint binding needs some way to keep
track of this so that matching `int literal(123)` against `T: int8 |
int16` also binds `T` to `int8`[^1].
The only good way to do this IMO is to generate a new "binding context"
when matching against constraints, then binding the generic param to
what the constraint was bound to in that context (in #24198 this is
restricted to just `or` types & concrete types with convertible matches,
it doesn't work in general).
---
`semtypinst` already does something similar for bindings of generic
invocations using `LayeredIdTable`, so `LayeredIdTable` is now split
into its own module and used in `sigmatch` for type bindings as well,
rather than a single-layer `TypeMapping`. Other modules which act on
`sigmatch`'s binding map are also updated to use this type instead.
The type is also made into an `object` type rather than a `ref object`
to reduce the pointer indirection when embedding it inside
`TCandidate`/`TReplTypeVars`, but only on arc/orc since there are some
weird aliasing bugs on refc/markAndSweep that cause a segfault when
setting a layer to its previous layer. If we want we can also just
remove the conditional compilation altogether and always use `ref
object` at the cost of some performance.
[^1]: `int8` binding here and not `int16` might seem weird, since they
match equally well. But we need to resolve the ambiguity here, in #24012
I tested disallowing ambiguities like this and it broke many packages
that tries to match int literals to things like `int16 | uint16` or
`int8 | int16`. Instead of making these packages stop working I think
it's better we resolve the ambiguity with a rule like "the earliest `or`
branch with the best match, matches". This is the rule used in #24198.
fixes#24121
Proc arguments can have typeclass type like `Foo | Bar` that `sigmatch`
handles specially before matching them to the param type, because they
wouldn't match otherwise. Not exactly sure why this existed but matching
any typeclass or unresolved type in generic contexts now fails the match
so typing the call is delayed until instantiation.
Also it turns out default values with `tyFromExpr` type depending on
other parameters was never tested, this also needs a patch to make the
`tyFromExpr` type `tfNonConstExpr` so it doesn't try to evaluate the
other parameter at compile time.
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#16376
The way the compiler handled generic proc instantiations in calls (like
`foo[int](...)`) up to this point was to instantiate `foo[int]`, create
a symbol for the instantiated proc (or a symchoice for multiple procs
excluding ones with mismatching generic param counts), then perform
overload resolution on this symbol/symchoice. The exception to this was
when the called symbol was already a symchoice node, in which case it
wasn't instantiated and overloading was called directly ([these
lines](b7b1313d21/compiler/semexprs.nim (L3366-L3371))).
This has several problems:
* Templates and macros can't create instantiated symbols, so they
couldn't participate in overloaded explicit generic instantiations,
causing the issue #16376.
* Every single proc that can be instantiated with the given generic
params is fully instantiated including the body. #9997 is about this but
isn't fixed here since the instantiation isn't in a call.
The way overload resolution handles explicit instantiations by itself is
also buggy:
* It doesn't check constraints.
* It allows only partially providing the generic parameters, which makes
sense for implicit generics, but can cause ambiguity in overloading.
Here is how this PR deals with these problems:
* Overload resolution now always handles explicit generic instantiations
in calls, in `initCandidate`, as long as the symbol resolves to a
routine symbol.
* Overload resolution now checks the generic params for constraints and
correct parameter count (ignoring implicit params). If these don't
match, the entire overload is considered as not matching and not
instantiated.
* Special error messages are added for mismatching/missing/extra generic
params. This is almost all of the diff in `semcall`.
* Procs with matching generic parameters now instantiate only the type
of the signature in overload resolution, not the proc itself, which also
works for templates and macros.
Unfortunately we can't entirely remove instantiations because overload
resolution can't handle some cases with uninstantiated types even though
it's resolved in the binding (see the last 2 blocks in
`texplicitgenerics`). There are also some instantiation issues with
default params that #24005 didn't fix but I didn't want this to become
the 3rd huge generics PR in a row so I didn't dive too deep into trying
to fix them. There is still a minor instantiation fix in `semtypinst`
though for subscripts in calls.
Additional changes:
* Overloading of `[]` wasn't documented properly, it somewhat is now
because we need to mention the limitation that it can't be done for
generic procs/types.
* Tests can now enable the new type mismatch errors with just
`-d:testsConciseTypeMismatch` in the command.
Package PRs:
- using fork for now:
[combparser](https://github.com/PMunch/combparser/pull/7) (partial
generic instantiation)
- merged: [cligen](https://github.com/c-blake/cligen/pull/233) (partial
generic instantiation but non-overloaded + template)
- merged: [neo](https://github.com/andreaferretti/neo/pull/56) (trying
to instantiate template with no generic param)
fixes#21258
When a generic proc is instantiated, if one of the default values
doesn't match the type of the parameter, `seminst` sets the default
parameter node to an `nkEmpty` node with `tyError` type. `sigmatch`
checks for this to give an error message if the default param is
actually used, but only while actively matching the proc signature,
before the proc is even instantiated. The error message also gives very
little information.
Now, we check for this in `updateDefaultParams` at the end of
`semResolvedCall`, after the proc has been instantiated. The `nkEmpty`
node also is given the original mismatching type instead rather than
`tyError`, only setting `tyError` after erroring to prevent cascading
errors. The error message is changed to the standard type mismatch error
also giving the instantiation info of the routine.
fixes#23730
Since #23188 the compiler errors when matching a type variable to an
uninstantiated static value. However sometimes an uninstantiated static
value is given even when only a type match is being performed to the
base type of the static type, in the given issue this case is:
```nim
proc foo[T: SomeInteger](x: T): int = int(x)
proc bar(x: static int): array[foo(x), int] = discard
discard bar(123)
```
To deal with this issue we only error when matching against a type
variable constrained to `static`.
Not sure if the `q.typ.kind == tyGenericParam and
q.typ.genericConstraint == tyStatic` check is necessary, the code above
for deciding whether the variable becomes `skConst` doesn't use it.
fixes#23186
As explained in #23186, generics can transform `genericProc[int]` into a
call `` `[]`(genericProc, int) `` which causes a problem when
`genericProc` is resemmed, since it is not a resolved generic proc. `[]`
needs unresolved generic procs since `mArrGet` also handles explicit
generic instantiations, so delay the resolved generic proc check to
`semFinishOperands` which is intentionally not called for `mArrGet`.
The root issue for
[t6137](https://github.com/nim-lang/Nim/blob/devel/tests/generics/t6137.nim)
is also fixed (because this change breaks it otherwise), the compiler
doesn't consider the possibility that an assigned generic param can be
an unresolved static value (note the line `if t.kind == tyStatic: s.ast
= t.n` below the change in sigmatch), now it properly errors that it
couldn't instantiate it as it would for a type param. ~~The change in
semtypinst is just for symmetry with the code above it which also gives
a `cannot instantiate` error, it may or may not be necessary/correct.~~
Now removed, I don't think it was correct.
Still possible that this has unintended consequences.
refs #22605
Sym choice nodes are now only allowed to pass through semchecking if
contexts ask for them to (with `efAllowSymChoice`). Otherwise they are
resolved or treated as ambiguous. The contexts that can receive
symchoices in this PR are:
* Call operands and addresses and emulations of such, which will subject
them to overload resolution which will resolve them or fail.
* Type conversion operands only for routine symchoices for type
disambiguation syntax (like `(proc (x: int): int)(foo)`), which will
resolve them or fail.
* Proc parameter default values both at the declaration and during
generic instantiation, which undergo type narrowing and so will resolve
them or fail.
This means unless these contexts mess up sym choice nodes should never
leave the semchecking stage. This serves as a blueprint for future
improvements to intermediate symbol resolution.
Some tangential changes are also in this PR:
1. The `AmbiguousEnum` hint is removed, it was always disabled by
default and since #22606 it only started getting emitted after the
symchoice was soundly resolved.
2. Proc setter syntax (`a.b = c` becoming `` `b=`(a, c) ``) used to
fully type check the RHS before passing the transformed call node to
proc overloading. Now it just passes the original node directly so proc
overloading can deal with its typechecking.
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>
* sacrifice "tgenericshardcases" for working statics
* legacy switch for CI, maybe experimental later
* convert to experimental
* apparently untyped needs the experimental switch
* try special case call semcheck
* try fix
* fix compilation
* final cleanup, not experimental, make `when` work
* remove last needed use of untyped
* fix unused warning in test
* remove untyped feature
* 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
* fix#19882 Improve error message when instantiating generics that lack a type
* Update tests/errmsgs/t19882.nim
Co-authored-by: Clay Sweetser <Varriount@users.noreply.github.com>
* simplified proc-like name ident to symbol code
* wip - reworking generic param sem
* wip - closer to removing nkEmpty generic params
* it's hacky but tests pass
* slowly tweaking semProcAux to take on semLambda
* fix pragma superset check proto vs current
* Set the symbol owner earlier
* partial progress reworking proto
found bug where default values between forward and impl lead to overload
resolution issues.
* simplified pragma handling and callConv checks
Co-authored-by: Clyybber <Clyybber@users.noreply.github.com>
* partially working
* cgexprs issue
* It works!
* comment clean-up
* clean-up asserts, comments, and other bits
* add isGenericParams, inline isGeneric queries
* seeing if this is sufficiently consistent
* can use this approach or continue it in a further PR
* commentary about nullary generics and clean-ups
* fixed a mistake in PNode isGenericRoutine
* Some small cleanups
* Small cleanup
* for func lambdas ensure we use lambda pragmas
* add some basic compileTime func tests
* [ci skip] remove comments
Co-authored-by: Clyybber <Clyybber@users.noreply.github.com>
Co-authored-by: Clyybber <darkmine956@gmail.com>
* fixes#15210 [backport:1.2]
* make tests green
* make ordinal work
* makes Swapable test compile
* make Indexable example work
* concepts: 'self' is now 'Self'
* concepts: make Dictionary example compile
* document the new concept implementation
* concepts: make typeDesc work properly
* concepts: allow documentation comments (d'oh)
* removed dead code
* we need even more laziness for the generic caches
* make it bootstrap on older Nims
* wrote more deserialization code
* IC: replay required methods information
* IC: dead code elimination pass
* preparations for a different codegen strategy
* added documentation to the newly written code
* IC: backend code
* IC: backend adjustments
* optimized the compiler a bit
* IC: yet another massive refactoring
* fixes regressions
* cleanups
* cleanups
* ast.nim: cleanups
* IC: no more sym.tab field, stored externally in the module graph
* nimble compiles again
* rodfiles: store bitwidth of integers and the endianness in the cookie because we serialize 'int' directly
* rodfiles: added compilerproc and export sections
* rodfiles: added all the missing sections
* rodfiles: track the missing information
* IC: architecture for lazy loading of proc bodies
* make tests green again
* completed the lazy loading of proc bodies
* symbol lookup integration, part 1
* symbol lookup integration, part 2
* symbol lookup integration, part 3
* make tcompilerapi work again
* rodfiles: fixed config change handling
* reworked ID handling
* the packed AST now has its own ID mechanism
* basic serialization code works
* extract rodfiles to its own module
* rodfiles: store and compare configs
* rodfiles: store dependencies
* store config at the end
* precise dependency tracking
* dependency tracking for rodfiles
* completed loading of PSym, PType, etc
* removed dead code
* bugfix: do not realloc seqs when taking addr into an element
* make IC opt-in for now
* makes tcompilerapi green again
* final cleanups
Co-authored-by: Andy Davidoff <github@andy.disruptek.com>
