- AST nodes cannot be shared, because type conversion changes them in
place, and what's appropriate for one use is by no means appropriate for
all uses.
- Added ZCCCompiler class as a place to generate IR and symbols from an
AST. Right now, all it does is simplify constant expressions into
constant values.
- Do type promotion on the AST where appropriate.
- Added true and false tokens to the parser driver.
- Since the VM doesn't directly support the GT, GTEQ, and NEQ comparisons,
don't use them in the trees either. Instead, wrap them as LTEQ, LT, and
EQEQ inside a BoolNot operator.
- I can't believe I completely forgot to let the parser handle true and
false literals.
- Consolidate all the %include blocks in zcc-parse.lemon into a single
one, because Lemon all of a sudden decided it didn't like me having more
than one in the grammar file.
- Added a PBool type to represent boolean values with.
- Since the tokenizer never gives the parser negative numbers but always a
unary minus followed by a positive number, it seems reasonable to make
the parser smart enough to turn these into negative constants without
generating extra tree nodes.
- And since we're doing it for unary -, we might as well do it for unary +
as well and avoid extra nodes when we know we don't need them.
- For an enum like this:
enum { value1 = SOME_NUM*2, value2 };
Generate an increment expression for value2 of the form
(add (id value1) 1)
and not
(add (* SOME_NUM 2) 1)
- Instead of representating enumeration values with a special node type,
use the same ZCC_ConstantDef nodes that const_def produces. These are
created at the same scope as the ZCC_Enum, rather than being contained
entirely within it. To mark the end of enums for a single instance of
ZCC_Enum, a ZCC_EnumTerminator node is now appended to the chain of
ZCC_ConstantDefs.
- Instead of having ZCC_ExprString, ZCC_ExprInt, and ZCC_ExprFloat,
just use a single ZCC_ExprConstant. It should simplify type
promotion and constant folding in the future.
- Constants can fill out the type field right away. Other expressions will need
to wait until a later pass, after names have been resolved, so they get
initialized to NULL.
- To simplify code generation genericizing, add three new opcodes
* NOP: No-Operation
* LANG: Load Angle - load a BAM angle into a float reg as degrees
* SANG: Save Angle - store a float reg into a BEM angle, converting from degrees
- Added versions of the trig operations supported by FLOP that can work
with degrees directly instead of radians.
- Reorder FLOPs into more sensible groupings.
- More forced line breaks.
- Don't use extra parentheses in places where it can still be unambiguous
without them.
- Don't print the UserType for basic-types that aren't ZCC_UserType.
- Being able to omit optional function arguments is not such a nonsensical
thing after all. However, the previous grammar was still inadequate for
representing this in a useful way.
- Fixed: Trying to define a function without any parameters would silently
discard the function, because the declarator tested FuncParams instead
of FuncName to decide if it was a function.
- Fixed: state_call needs to enclose func_expr_list in LPAREN/RPAREN
itself, because func_expr_list doesn't include them. This means it also
needs a separate production to accept calls without a parameter list.
- The variable_name production now accepts an optional array size
argument. (Not yet passed to the AST.)
- The notation for using dotted ID lists as types has been changed from
[id1.id2] to .id1.id2, beacuse the former conflicts with the notation
for arrays.
- Changed Actor's Damage property into an actual function. All access to the damage property
must now be done through GetMissileDamage. actor->GetMissileDamage(0, 1) is equivalent
to the former actor->Damage, for the case where actor->Damage was not an expression. (I
suppose I will probably need to make a thunk for DECORATE expressions that want to read it.)
- Cleaned up some decorate expression evaluation functions that are no longer used.
SVN r3919 (scripting)
one name in its label. This avoids the alloca and security cookie calls.
- Consolidated lots of copy-pasted code in thingdef_expression.cpp to install helper functions
into FindDecorateBuiltinFunction().
SVN r3907 (scripting)
become ungodly slow when using mods with complex DECORATE. The GCC debug builds run just
fine, however. Hopefully this is something that can be fixed later with an assembly-optimized
version of the main VM loop, because I don't relish the thought of being stuck with GDB
for debugging.)
- Fixed: The ACS_Named* action specials were erroneously defined as taking strings instead of
names.
- Fixed: Copy-paste error caused FxMultiNameState::Emit to generate code that called
DecoNameToClass instead of DecoFindMultiNameState.
- Updated FxActionSpecialCall::Emit for named script specials.
- Fixed inverted asserts for FxMinusSign::Emit and FxUnaryNotBitwise::Emit.
SVN r3893 (scripting)
Maybe TODO: Find a simple LISP pretty printer that isn't written in LISP so that the output
isn't all on one big long line.
- zcc-parse.lemon now returns the complete AST through the parse state.
- Fixed crash in PrintEnumNode().
SVN r3773 (scripting)
function will pass its results directly to this function's caller. Eventually, this should
be changed to do a proper tail call for scripted functions.
SVN r3769 (scripting)
of the type declaration rather than split it up after the variable name.
- More AST work. At this rate, maybe I'll finally be building a complete AST by the end of the
month!
SVN r2499 (scripting)
me realize I would be better off creating a type-agnostic AST for the entire input instead
of trying to tentatively create types before they're seen.)
SVN r2354 (scripting)
running with the checked VM can be quite slow, since it has asserts everywhere. Some other
fixes were needed before the code actually worked:
- A_CallSpecial needs to have its arguments cast to ints.
- Some functions that set pnum/paramnum directly did not decrement it by 1. This also applies
to A_Jump, though it just uses the value of paramnum instead of changing it.
- Renamed pnum in the PARAM macros to paramnum, since pnum is already used in a few other
places for something different, so this makes searching for it easier.
This has not been tested especially thoroughly, but a first glance seems to indicate success.
SVN r2163 (scripting)
can actually be used. In the end, all temporaries should be assigned to
unique virtual registers, and a register allocator can alias them to real
registers, but as long as we don't do any CSE, this is good enough for now.
SVN r1920 (scripting)
Currently, they're only good for disassembly and pasting into a proper text editor
for viewing.
- Fixed some problems with the FxExpression emitters that were revealed by actually
using them on the standard actors.
SVN r1911 (scripting)
it isn't even used anywhere yet. In retrospect, I probably should have targeted an intermediate
representation and done codegen with that instead, since that would be something I can reuse.
SVN r1908 (scripting)
- Derive PClass from dobject.cpp. This has one major ramification: Since the PClass
is not allocated until runtime, you cannot initialize any static/global data
structures with pointers to PClasses using RUNTIME_CLASS. Attempting to do so
will just initialize with a NULL pointer. Instead, you can initialize using
the address of the pointer returned by RUNTIME_CLASS and dereference that. By
the time you have an opportunity to dereference it, it will no longer be NULL.
- Sync CmakeLists.txt.
- Random fixes for problems GCC spotted.
SVN r1852 (scripting)
