#!/usr/bin/env python3
"""Offline helper (runs HERE, stdlib-only): cluster vtable candidates.

A vtable = run of >=3 consecutive LE32 words in rodata, each pointing
into the code sect. Output CSV is consumed by the Ghidra-side script.

Usage:
  python3 ghidra/gen_vtable_csv.py "Lofi-12XT_v1.5.205/.../Lofi-12 XT.bin" /tmp/vtables-1.5.205.csv
"""
import os
import struct
import sys

sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', 'tools'))
from lofi_image import parse_image, find_sect_by_role


def main():
    image, out = sys.argv[1], sys.argv[2]
    d = open(image, 'rb').read()
    info = parse_image(d)
    code = find_sect_by_role(info, 'code')
    ro = find_sect_by_role(info, 'rodata')
    print("code sect%d %08X..%08X" % (code['index'], code['addr'], code['end']))
    print("rodata sect%d %08X..%08X" % (ro['index'], ro['addr'], ro['end']))

    p = ro['payload']
    words = [struct.unpack('<I', p[i:i + 4])[0]
             for i in range(0, len(p) - 3, 4)]
    is_code_ptr = [code['addr'] <= w < code['end'] for w in words]

    runs = []
    i = 0
    n = len(words)
    while i < n:
        if is_code_ptr[i]:
            j = i
            while j < n and is_code_ptr[j]:
                j += 1
            if j - i >= 3:
                runs.append((i * 4, j - i, words[i:j]))
            i = j
        else:
            i += 1

    with open(out, 'w') as f:
        f.write("rodata_off,load_addr,nentries,entries\n")
        for off, cnt, ws in runs:
            f.write("%d,%08X,%d,%s\n"
                    % (off, ro['addr'] + off, cnt,
                       ";".join("%08X" % w for w in ws)))
    to_code = sum(is_code_ptr)
    print("rodata LE words -> code: %d, vtable runs(>=3): %d -> %s"
          % (to_code, len(runs), out))


if __name__ == '__main__':
    main()