 svenbarth
|
0aee319fdb
* do_member_read() takes ownership of the specialization context, so Nil it after the call
|
7 years ago |
 florian
|
93c5eae528
* accept dereferences in generics definitions, resolves #33700 properly
|
7 years ago |
|
svenbarth
|
441fc2fab6
* also parse postfix operators after parsing a _STRING token (required for class functions on type helpers for String)
|
7 years ago |
 nickysn
|
e813a11e85
+ support TP7-compatible @proc^ (as in FillChar(@proc^,...))
|
7 years ago |
|
nickysn
|
88d9dd23ae
* the 'again' parameter of factor_read_id() changed to 'out' instead of 'var'
|
7 years ago |
|
svenbarth
|
22a5e5e4ee
* the symbol parsed for "specialize" does not need to be a type; it could be a procsym or some other type (e.g. a constant of the same name in a third unit)
|
7 years ago |
|
svenbarth
|
e7f29e9e28
* also pass the pattern with which the potential unit/namespace symbol had been found to try_consume_unitsym*()
|
7 years ago |
|
nickysn
|
3318703ece
* moved nf_typedaddr to addrnodeflags (anf_typedaddr)
|
7 years ago |
|
nickysn
|
213e61eae5
+ introduce the anf_ofs address node flag - indicates that the node was
|
7 years ago |
 Jonas Maebe
|
2919d97f91
* support asssigning @class.classmethod to a procvar of object in FPC modes
|
7 years ago |
|
nickysn
|
78e0f6c68b
+ added an i8086-specific FarAddr() function, similar to Addr(), but always
|
7 years ago |
|
svenbarth
|
f6a867ef04
* move handling of Concat to tinlinenode so that it can be easily extended for dynamic arrays
|
7 years ago |
|
svenbarth
|
83f5b27e02
* check helpertype instead of the extendeddef to judge whether inherited should be allowed or not
|
8 years ago |
|
svenbarth
|
6acba684d4
* class helpers: fix calling virtual methods of the extended type using inherited
|
8 years ago |
|
svenbarth
|
0b02dab684
+ new Delphi-compatible intrinsic GetTypeKind() which returns the TTypeKind of a type as a constant value (and thus can be optimized away in If- and Case-statements)
|
8 years ago |
|
svenbarth
|
eb9b4fb71f
* fix for Mantis #31973: resolve dummy symbols earlier and print a nice error message if a generic dummy could not be resolved to an ordinary symbol
|
8 years ago |
|
nickysn
|
efc5e339d0
* use an enum instead of integer constants to represent inline numbers
|
8 years ago |
|
florian
|
b1dff29cbf
* removed unused units
|
8 years ago |
|
svenbarth
|
1aa80f77cb
- remove no longer needed global variable allow_array_constructor
|
8 years ago |
|
svenbarth
|
6d512ddac7
* allow nested array constructors inside normal code blocks (body & except)
|
8 years ago |
|
svenbarth
|
82b2d22ed6
* keep track of allow_array_constructor as part of the array constructor node
|
8 years ago |
 maciej-izak
|
16a11c8b7a
Bug fix for getting pointer to function result (mantis #10933 #19861)
|
8 years ago |
|
svenbarth
|
22e579cc74
* fix for Mantis #31107: disallow calling of ordinary record methods using the record's type.
|
8 years ago |
|
florian
|
2b6ccb3565
* better error recovery
|
8 years ago |
|
svenbarth
|
531829c072
* don't check for hints if we have a specialization with the same name as the to be used generic (though since the typedef would be errordef it wouldn't print anything anyway...)
|
9 years ago |
|
svenbarth
|
50f2aab0ef
* fix for Mantis #22192: handle the case that a specialization has the same name as the generic it's based on as in that case the typedef of the symbol is an errordef; therefor we need to use the symbol's real name
|
9 years ago |
|
svenbarth
|
7dd0130ac2
* generate_specialization_phase2() might also return an errordef, so check for that instead of simply using the symbol; fixes access violations of tgenconstraint{38,39,40}.pp
|
9 years ago |
|
svenbarth
|
aaceb5518f
* fix Mantis #30494 by applying patch from Maciej Izak
|
9 years ago |
|
nickysn
|
c2305809dc
+ added an i8086-embedded target support to the compiler (RTL and makefile
|
9 years ago |
|
svenbarth
|
a2c9c75e97
Convert Insert() and Delete() to intrinsics in preparation for dynamic array support for these two procedures.
|
9 years ago |
