bugfixes; field discriminant checks; linearScanEnd, unroll, shallow pragmas

This commit is contained in:
Araq 2011-03-23 01:09:52 +01:00
commit 5b789f2da8
45 changed files with 2342 additions and 243 deletions

View file

@ -22,11 +22,13 @@ High priority (version 0.9.0)
Bugs
----
- proc (x: int) is passable to proc (x: var int) !?
- detected by pegs module 64bit: p(result, result) should use a temporary!
- the parser allows empty object case branches
- Exception matching needs to take subclasses into account
- pegs: the anchor '^' does not work because many procs use a linear search
and matchLen()
- BUG: generic assign still buggy
- Optimization: If we use a temporary for the result anyway the code gen
should make use of this fact to generate better code...
To implement
@ -51,18 +53,13 @@ Low priority
- find a way to reintroduce the cleanup() pass for C code generation: this
is hard because of partial evaluation --> symbol files will fix this as
a side effect
- implement better error handling: errornous top level statements are ignored
and not part of the symbal table --> for REPL
--> this needs deletion operation for symbol table!
- floating point checks for EcmaScript
- enhance `` notation for identifier concatenation: `concat` a `these`
- prefer proc in current module over other procs with same overloading result?
- real types for template results
- generalized case statement (requires better transf)
- normalize for the DOM
- tlastmod returns wrong results on BSD (Linux, MacOS X: works)
- nested tuple unpacking
- fast assignment optimization for TPeg
- nested tuple unpacking; no auto-unpacking in 'for' loops!
- better error messages for used keywords as identifiers
- case statement branches should support constant sets