#!/usr/bin/python

URL = "http://www.prefeitura.unicamp.br/pref/servicos/restaurantes-carda.html"
ARQ = "/home/gustavo/public_html/bandex.wml"


# Módulos utilizados:
import sys
import re
from urllib import *
from htmlentitydefs import entitydefs


# Pega a página com cardápio:
print >> sys.stderr, "Aguarde enquanto pego o cardápio de %s ..." % URL ,

page = urlopen( URL )
html = page.read()
page.close()

print >> sys.stderr, "Feito!"



################################################################################
# Expressões regulares: Black Magic                                            #
# Vide: http://www.python.org/doc/2.3/lib/module-re.html                       #
################################################################################

# Limpa algumas tags inúteis (pra nós):
# <span>, <b>, <br>, <div>
html = html.replace( "\r", "" )
html = re.sub( "<(span|b|div|p|tbody) *[^>]*>", "", html )
html = re.sub( "</(span|b|div|p|tbody)>", "", html )

# Pega os dados da página
valores = re.search( "^\s+<table width=\"92%\" border=\"1\" height=\"126\" " + \
                     "align=\"center\" bordercolor=\"#666666\">\s*$\\\n" + \
                     "^\s+<tr>\s*$\\\n" + \
                     "^\s+<td width=\"26%\">\s*$\\\n" + \
                     "^\s+(?P<label0>\S+)\s*$\\\n" + \
                     "^\s+</td>\s*$\\\n" + \
                     "^\s+<td width=\"20%\">\s*$\\\n" + \
                     "^\s+(?P<label1>\w+)\s*$\\\n" + \
                     "^\s+</td>\s*$\\\n" + \
                     "^\s+<td width=\"18%\">\s*$\\\n" + \
                     "^\s+(?P<label2>\w+)\s*$\\\n" + \
                     "^\s+</td>\s*$\\\n" + \
                     "^\s+<td width=\"17%\">\s*$\\\n" + \
                     "^\s+(?P<label3>\w+)\s*$\\\n" + \
                     "^\s+</td>\s*$\\\n" + \
                     "^\s+<td width=\"19%\">\s*$\\\n" + \
                     "^\s+(?P<label4>\d{2}/\d{2})\s*$\\\n" + \
                     "^\s+</td>\s*$\\\n" + \
                     "^\s+</tr>\s*$\\\n" + \
                     "^\s+<tr>\s*$\\\n" + \
                     "^\s+<td width=\"26%\" height=\"36\">\s*$\\\n" + \
                     "^\s+(?P<value0>[^<]+)" + \
                     "^\s+</td>\s*$\\\n" + \
                     "^\s+<td width=\"20%\" height=\"36\">\s*$\\\n" + \
                     "^\s+(?P<value1>[^<]+)" + \
                     "^\s+</td>\s*$\\\n" + \
                     "^\s+<td width=\"18%\" height=\"36\">\s*$\\\n" + \
                     "^\s+(?P<value2>[^<]+)" + \
                     "^\s+</td>\s*$\\\n" + \
                     "^\s+<td width=\"17%\" height=\"36\">\s*$\\\n" + \
                     "^\s+(?P<value3>[^<]+)" + \
                     "^\s+</td>\s*$\\\n" + \
                     "^\s+<td width=\"19%\" height=\"36\">\s*$\\\n" + \
                     "^\s+(?P<value4>[^<]+)" + \
                     "^\s+</td>\s*$\\\n" + \
                     "^\s+</tr>\s*$\\\n" + \
                     "^\s+</table>",
                    html,
                    re.MULTILINE | re.DOTALL )



# Transcreve valores para um dicionário (hash, ie:
#    valor[ "value0" ] == prato do dia)
# E ainda remove espaços duplos, \n e etc...
re_limpa = re.compile( "( {1,}|\\\n)", re.MULTILINE )

def limpa_texto( texto ):
    # Troca &<entidade>; pelo valor correspondente, ie: &atilde; -> ã
    for e in entitydefs:
        if e != "amp": # Não converte &amp; -> &
            texto = texto.replace( "&" + e + ";", entitydefs[ e ] )

    # chama 2 vezes re_limpa.sub( ... re_limpa.sub( .. ) )
    # para tirar os espaços duplos
    texto = re_limpa.sub( " ", re_limpa.sub( " ", texto.strip() ) )
    
    return texto
# limpa_texto()


valor = { }
if valores:
    for v in valores.groupdict():
        valor[ v ] = limpa_texto( valores.group( v ) )

# Fazer alguma coisa com valor, exemplo:
#print "Refeição: %s" % valor[ "label0" ]
#print "Prato principal: %s" % valor[ "value0" ]
#for i in range( 1, 4 ):
#    print "%s: %s" % ( valor[ "label%d" % i ], valor[ "value%d" % i ] )


# No caso do wap, vc gera o documento, vide http://www.wapforum.org
# Eu não manjo disto pq nunca mexi... mas fiz isso baseado numa parte do Freevo
# Ah, e foi mal usar strings mult-line ( """ ... """ ), mas ficar digitando concatenação
# enche o saco...

# Cabeçalho WAP
html = \
"""<?xml version="5.0" encoding="ISO8859-1" ?>
<!DOCTYPE wml PUBLIC "-//WAPFORUM//DTD WML 1.1//EN" "http://www.wapforum.org/DTD/wml_1.1.xml">

<wml>
   <card id="card1" title="Cardápio Bandeijão">
      <p>
         <big>
            %s %s
         </big>
      </p>
      <p>
         <strong>
            Prato Principal:
         </strong>
         %s
         <br />
         <br />
""" % ( valor[ "label0" ], valor[ "label4" ], valor[ "value0" ] )

# Imprime os acompanhamentos:
for i in range( 1, 4 ):
    html += """         <strong>
            %s:
         </strong>
         %s
         <br />
""" % ( valor[ "label%d" % i ],
        valor[ "value%d" % i ] )

# Rodapé WAP
html += \
"""      </p>
   </card>
</wml>
"""


# ae vc faz alguma coisa com esta página, por exemplo grava na sua área...
page = open( ARQ, "w")
page.write( html )
page.close()