* compinnr.inc include file converted to a unit
* inline number field size stored in ppu increased from byte to longint
* inlines in the parse tree (when written with the -vp option) now printed with
their enum name, instead of number
git-svn-id: trunk@36174 -
only used in combination with subscriptn/vecn to determine whether the
target is assignable (and we return true immediately if you subscript
or index an implicit pointer type -- also for classrefdef now)
o disallow assigning something to fields of records returned by functions
(that would be assigned to a temp and get lost anyway)
git-svn-id: trunk@35313 -
where it used to be set to true (dereferencing a pointer always means
that regardless of what the pointer was, the resut is assignable)
(mantis #22979)
o also removed gotpointer, since it was only used in combination with
gotderef
git-svn-id: trunk@35308 -
the scope penalty relative to the nearest symtable that contains one of
the applicable overloads, rather than relative to the nearest symtable
that simply contains a method with this name (based on patch by
Maciej Izak, mantis #25607)
git-svn-id: trunk@35089 -
and getters for the ttemp*node classes instead
o this will allow descendants to prevent certain flags from being added
or removed. E.g. for LLVM, certain temps must never be put in registers
because it cannot typecast a value in a register from a non-record/array
type to an array type without forcing it to memory (so if that is done
on an lvalue, the result will be written to the memory temp instead of
to the register)
git-svn-id: trunk@34358 -
htypechk.pas, tcallcandidates:
* create_candidate_list: check the correct tableoptions for sto_has_overload flag
+ added test
git-svn-id: trunk@33211 -
htypechk.pas, tcallcandidates:
* create_candidate_list: don't check whether the pd is a specialization if the owner is valid; happens if a generic method is used more than once (which should happen here and then :P )
+ added test
git-svn-id: trunk@33037 -
* extend tcallnode with the ability to pass a tspecializationcontext so that tcallcandidates can do a final specialization
* the final procdef is registered at the end of tcallnode.pass_typecheck
git-svn-id: trunk@31763 -
dynamic arrays and fixed-length static arrays when looking for
overloads of array constructors, while these are not valid in
such cases (and it also gave an error afterwards when trying to
actually use them). This caused a lot of spurious "can't select
which overloaded routine to call" errors when using many JVM
routines
git-svn-id: trunk@29393 -
htypechk.pas, tcallcandidates:
* collect_overloads_in_struct: we need to check the deftyp of the extended def for record- or objectdef, not the deftyp of the helper which will always be objectdef anyway.
+ added test
git-svn-id: trunk@28572 -
defcmp.pas:
+ add new function "def_is_related" which combines the "is_related" overloads of "tobjectdef", "trecorddef" and "tstringdef" (it returns "false" for other def types which is what "tdef.is_related" did)
* compare_defs_ext & compatible_childmethod_resultdef: change call from "x.is_related" to "def_is_related(x,...)"
symtype.pas, tdef:
- remove "is_related" method
symdef.pas:
- remove "is_related" in "tobjectdef", "trecorddef" and "tstringdef"
* tobjectdef.needs_inittable: for checking whether a Corba interface somehow inherits from a IInterface don't use "is_related" anymore (we want to avoid the dependency after all), but mimic the necessary functionality of "def_is_related"
htypechk.pas, nadd.pas, ncal.pas, ncnv.pas, ngtcon.pas, nld.pas, optvirt.pas, pdecobj.pas, pdecvar.pas, pexpr.pas, pgenutil.pas:
* change call from "x.is_related" to "def_is_related(x,...)"
symtable.pas
+ use unit defcmp
* change call from "x.is_related" to "def_is_related(x,...)"
jvm/njvmcnv.pas, jvm/njvmflw.pas:
* change call from "x.is_related" to "def_is_related(x,...)"
git-svn-id: trunk@25847 -
symconst.pas:
+ add "sto_has_operator" to "tsymtableoption"
pdecsub.pas:
+ include the flag "sto_has_operator" for all operator declarations and their owning symtables
htypechk.pas, tcallcandidates:
* create_candidate_list: only check for operator overloads if the record does indeed declare some
* collect_overloads_in_units: only check for operator overloads if the unit does indeed declare some
utils/ppudump.pp:
+ respect the new "sto_has_operator" flag
git-svn-id: trunk@23688 -
* check "A op B" and "B op A" again for operators that can be commutative (all binary ones except shl, shr, div, mod, **, / and -)
* also check for Nil for classrefdefs if left side is a pointer (allows "TClass var" <>/= Nil again, after the above changes)
* don't allow overloads for "implicit pointer type <>/= pointer" and the other way around (this fixes non compiling Objective Pascal test tobjc21.pp and also the new toperator87.pp test)
* some formating corrections
+ added test for "TObject <> Pointer"
+ added test for "TClass <>/= Nil"
git-svn-id: trunk@21983 -
The changes are built up in a way that all operators on two given types are not allowed if a default implementation for that operator exists. This implies that there is a possibility to have a overload work for (in this order) type A and B, but not for B and A. This is for example the case for Set + Enum (which seems to have a default implementation) where Enum + Set is allowed. The added tests try to detect as many default implementations as feasible, but can't cover everything...
git-svn-id: trunk@21975 -
