RTT2 EBOOT Dump

start = 0x2100D4

def is_jis_char(s):
    '''test passed string: true if two bytes that decode to one SJIS char'''
    if len(s) != 2:
        return False
    try:
        if len(s.decode('cp932')) == 1:
            return True
    except UnicodeDecodeError:
        return False

with open('NPJH50119.BIN', 'rb') as f:
    f.seek(0x2100d4)                #Read in our .data block
    filedata = f.read(0x1a2250)
slist = []
output = []
decode_fail = []
pos = 0
count = 0
while pos < len(filedata):
    if (filedata[pos] > 0x1F and filedata[pos] < 0x80) or (     #ASCII char
        filedata[pos] == 0xA) or (                              #Newline char
        is_jis_char(filedata[pos:pos+2])):                      #SJIS char
        count += 1
        end = filedata.find(b'\x00', pos)
        try:                                        #Test decode
            s = filedata[pos:end].decode('cp932')
        except UnicodeDecodeError:
            decode_fail.append(hex(pos + start))
            pos = end
            pos += 4 - pos % 4                      #Word-align pointer
            continue                                #Search for next string
        output.append((hex(pos + start), s))        #Save string and position
        pos = end
    pos += 4 - pos % 4                              #Word-align pointer
with open('o.txt', 'w', encoding='utf-8') as f:
    f.write('\n'.join([a + '\t' + b.replace('\n', '\n\t') for a, b in output]))
with open('f.txt', 'w') as f:
    f.write('\n'.join(decode_fail))