Experimental, no dependencies, python compiler for the x86-64 architecture on linux, utilizing vscc for bytecode generation. This project is largely unfinished and is primarly a proof of concept.
git clone --recurse-submodules https://github.com/dmaivel/pyvscc.git
cd pyvscc
mkdir build
cd build
cmake ..
cmake --build . --config Releaseusage: pyvscc [-h] [-i FILE_PATH] [-e ENTRY_POINT] [-m SIZE] [-s SIZE] [-o] [-p]
options:
-h display help information
-i [FILE_PATH] input file path
-e [ENTRY_POINT] specify entry function (if not specified, searches for any function containing 'main')
-m [SIZE] max amount of bytes program may allocate (default: 4096 bytes)
-s [SIZE] amount of bytes variables/functions with an unspecified type take up (default: 8 bytes)
-o enable optimizations
-p print performance information
Currently, this compiler is able to compile and execute the following example effortlessly (tabs must be U+0009):
def play(y):
if y == 3:
return 11
return 10
def main():
x = play(4)
if x != 11:
print('hello world!\n')
return xClick to reveal: disassembly of compiled output
; pyimpl_strlen(i64)
0: 55 push rbp
1: 48 89 e5 mov rbp,rsp
4: 48 83 ec 11 sub rsp,0x11
8: 48 89 7d f8 mov QWORD PTR [rbp-0x8],rdi
c: 48 c7 45 f0 00 00 00 mov QWORD PTR [rbp-0x10],0x0
13: 00
14: 48 8b 45 f8 mov rax,QWORD PTR [rbp-0x8]
18: 8a 00 mov al,BYTE PTR [rax]
1a: 88 45 ef mov BYTE PTR [rbp-0x11],al
1d: 80 7d ef 00 cmp BYTE PTR [rbp-0x11],0x0
21: 0f 84 0f 00 00 00 je 0x36
27: 48 83 45 f8 01 add QWORD PTR [rbp-0x8],0x1
2c: 48 83 45 f0 01 add QWORD PTR [rbp-0x10],0x1
31: e9 de ff ff ff jmp 0x14
36: 48 8b 45 f0 mov rax,QWORD PTR [rbp-0x10]
3a: 48 83 c4 11 add rsp,0x11
3e: 5d pop rbp
3f: c3 ret
; pyimpl_print_str(i64)
40: 55 push rbp
41: 48 89 e5 mov rbp,rsp
44: 48 83 ec 10 sub rsp,0x10
48: 48 89 7d f8 mov QWORD PTR [rbp-0x8],rdi
4c: 48 8b 7d f8 mov rdi,QWORD PTR [rbp-0x8]
50: e8 ab ff ff ff call 0x0
55: 48 89 45 f0 mov QWORD PTR [rbp-0x10],rax
59: 48 c7 c7 01 00 00 00 mov rdi,0x1
60: 48 8b 75 f8 mov rsi,QWORD PTR [rbp-0x8]
64: 48 8b 55 f0 mov rdx,QWORD PTR [rbp-0x10]
68: 48 c7 c0 01 00 00 00 mov rax,0x1
6f: 0f 05 syscall
71: 48 8b 45 f0 mov rax,QWORD PTR [rbp-0x10]
75: 48 83 c4 10 add rsp,0x10
79: 5d pop rbp
7a: c3 ret
; play(i64)
7b: 55 push rbp
7c: 48 89 e5 mov rbp,rsp
7f: 48 83 ec 08 sub rsp,0x8
83: 48 89 7d f8 mov QWORD PTR [rbp-0x8],rdi
87: 48 83 7d f8 03 cmp QWORD PTR [rbp-0x8],0x3
8c: 0f 85 0a 00 00 00 jne 0x9c
92: b8 0b 00 00 00 mov eax,0xb
97: e9 05 00 00 00 jmp 0xa1
9c: b8 0a 00 00 00 mov eax,0xa
a1: 48 83 c4 08 add rsp,0x8
a5: 5d pop rbp
a6: c3 ret
; main()
a7: 55 push rbp
a8: 48 89 e5 mov rbp,rsp
ab: 48 83 ec 18 sub rsp,0x18
af: 48 c7 c7 04 00 00 00 mov rdi,0x4
b6: e8 c0 ff ff ff call 0x7b
bb: 48 89 45 f8 mov QWORD PTR [rbp-0x8],rax
bf: 48 83 7d f8 0b cmp QWORD PTR [rbp-0x8],0xb
c4: 0f 84 18 00 00 00 je 0xe2
ca: 48 8d 05 2b 00 00 00 lea rax,[rip+0x2b] # 0xfc
d1: 48 89 45 f0 mov QWORD PTR [rbp-0x10],rax
d5: 48 8b 7d f0 mov rdi,QWORD PTR [rbp-0x10]
d9: e8 62 ff ff ff call 0x40
de: 48 89 45 e8 mov QWORD PTR [rbp-0x18],rax
e2: 48 8b 45 f8 mov rax,QWORD PTR [rbp-0x8]
e6: 48 83 c4 18 add rsp,0x18
ea: 5d pop rbp
eb: c3 ret
...The following explicit types may be used: byte, word, dword, and qword, each representing a sigle byte, two bytes, four bytes, and eight byte values respectively.
def main(x: qword) -> dword:
# ...