Match HistoryBook::ReadyWorld()

[jonschz]

Closes #1369.

Matches to about 53 % on BETA10, though most of the mismatch comes from incorrect registers / entropy. The structural match isn't perfect, but quite decent.

I got LEGO1 to 91 %. Some of it appears to be entropy, but I can't make the beginning work:

0x1008271e : mov ecx, 'A_Bitmap' (STRING)
-0x10082723 : -cmp si, 0x1a
0x10082727 : lea edx, [esp + 0x40]
0x1008272b : mov cl, byte ptr [ecx + 8]
0x1008272e : mov dword ptr [edx], eax
0x10082730 : mov dword ptr [edx + 4], ebx
0x10082733 : mov byte ptr [edx + 8], cl
-0x10082736 : -jge 0x25
0x10082738 : lea eax, [esp + 0x40]
0x1008273c : mov ecx, dword ptr [esp + 0x18]
0x10082740 : push eax
0x10082741 : push 'MxStillPresenter' (STRING)
0x10082746 : call LegoWorld::Find(char const *, char const *) (FUNCTION)

It looks as if the code does something like

MxS16 i = 0;
char bitmap[] = "A_Bitmap";

for (; i < 26; i++) {
// [...]
}

but that only makes the match worse.

[foxtacles]

Matched it to 100% by adding another if (i < 26) check in the loop. Doesn't make any sense but I guess that's what they did. Could be that they added a new inline function to set the entry in m_alphabet which has that check in the body. Probably the same in RegistrationBook::ReadyWorld