from BeautifulSoup import BeautifulSoup

tables = dict(
	tab_single='single.html',
)

types = dict(
	VD = 'DREX_'
)

imm = (
	'A', 'I', 'J', 'O',
)
for name in imm:
	types[name] = 'IMM_'

modrm = (
	'C', 'D', 'E', 'F',
	'G', 'M', 'P', 'PR',
	'N', 'Q', 'R', 'S',
	'T', 'V', 'VR', 'U',
	'W', 'X', 'Y'
)
for name in modrm:
	types[name] = 'MODRM_'

sizes = dict(
	b = 'BYTE',
	w = 'WORD',
	d = 'DWORD',
	q = 'QWORD',
	o = 'WORD',
	v = 'VAR',
	z = 'VAR',
	a = 'VAR',
	p = 'POINTER',
	s = 'SIXBYTE',
	t = 'TENBYTE',
)

def parse(val):
	flags = []
	for sub in val.split(','):
		sub = sub.strip()
		if not sub or len(sub) < 2:
			continue
		sub = sub[:-1], sub[-1]
		
		if len(sub) != 2 or sub[1].lower() != sub[1] or sub[0].upper() != sub[0]:
			continue
		if sub[0][0] == '(' or sub[1] == ':':
			continue
		if sub[1] in '0123456789':
			continue
		
		type = types[sub[0]]
		size = sizes[sub[1]]
		
		if not isinstance(type, list) and not isinstance(type, tuple):
			type = [type]
		for subtype in type:
			flags.append(subtype + size)
	
	if not flags:
		flags = '0'
	
	return ' | '.join(flags)

def screenScrape(html):
	ret = dict()
	soup = BeautifulSoup(html)
	
	for table in soup.contents:
		if not hasattr(table, 'name') or table.name != 'table':
			continue
		
		top = True
		lower = []
		for tr in table.tbody:
			if not hasattr(tr, 'name') or tr.name != 'tr':
				continue
			
			i = 0
			for td in tr:
				if not hasattr(td, 'name') or td.name != 'td':
					continue
				
				if top:
					for child in td.font:
						if not child.string or not child.string.strip():
							continue
						try:
							lower.append(int(child.string[1], 16))
						except:
							lower.append(0)
				else:
					lines = []
					for child in td.font:
						if not child.string or not child.string.strip():
							continue
						lines.append(child.string)
					
					if i == 0:
						higher = int(lines[0][0], 16) << 4
					elif i > 8:
						continue
					else:
						key = higher | lower[i]
						if len(lines) < 2:
							continue
						ret[key] = ','.join(lines).replace(' ', ',')
					
					i += 1
			
			top = False
	
	return ret

for name, table in tables.items():
	table = screenScrape(file(table, 'r').read())
	
	print 'UCHAR %10s[0x100] = {' % name
	
	for i in range(0x100):
		if i in table:
			val = parse(table[i])
		else:
			val = 0
		
		if i == 0xFF:
			print '\t/* 0x%02x */ %s' % (i, val)
		else:
			print '\t/* 0x%02x */ %s,' % (i, val)
	
	print '};'
	print