#!/usr/bin/env python

import base64
import sys
import re
import urllib
import urlparse

STATE_NOMATCH = 0
STATE_SAWBRACKET = 1
STATE_INTAG = 2

operation = "archive"

inputfilename = sys.argv[1]
inputfile = urllib.urlopen(inputfilename)
using_http = 0
if inputfilename[:5].lower() == "http:":
    using_http = 1

outputfilename = sys.argv[2]
outputfile = file(sys.argv[2], "w")
done = 0
state = STATE_NOMATCH
buffer = ""
# won't allow escaped quotes inside "src"
#img_re =  re.compile('^\s*img [^>]*src\s*=\s*(?P<quote>["\'])(?P<src>[^(?P=quote)]+)(?P=quote)', re.IGNORECASE)
img_re =  re.compile('^\s*img [^>]*src\s*=\s*(?P<srcwithquotes>(?P<quote>["\'])(?P<src>[^"\']+)(?P=quote))', re.IGNORECASE)
img_re2 = re.compile('^\s*img [^>]*src\s*=(?P<src>[^ "\']+)', re.IGNORECASE)

buffer = inputfile.read(4096)

while not done:

    if state == STATE_NOMATCH:
        match = buffer.find("<")
        if match != -1:
            # left brace found
            outputfile.write(buffer[:match+1])
            buffer = buffer[match+1:]
            state = STATE_SAWBRACKET
            continue
        else:
            # left brace not found
            outputfile.write(buffer)
            buffer = ""
    elif state == STATE_SAWBRACKET:
        # make sure we have the whole tag
        if buffer.find(">") == -1:
            # wait for the next iteration, so we have all the data
            pass
        match = img_re.search(buffer)
        if match != None:
            # "img" found.
            # first dump until the beginning of the value for the 'src' attribute
            outputfile.write(buffer[:match.start('srcwithquotes')])
            buffer = buffer[match.end('srcwithquotes'):]
            imgsrc = match.group('src')
            print imgsrc
            # guess mime type
            extension = imgsrc[imgsrc.rfind(".")+1:]
            if extension == "gif":
                mimetype="image/gif"
            elif extension == "jpg" or extension == "jpeg":
                mimetype="image/jpeg"
            elif extension == "png":
                mimetype="image/png"
            # todo: if not a filename, if we can't find it, etc
            if using_http:
                imgfile = urllib.urlopen(urlparse.urljoin(inputfilename, imgsrc))
            else:
                imgfile = file("./" + imgsrc)
            outputfile.write('"data:' + mimetype + ";base64,")
            base64.encode(imgfile, outputfile)
            outputfile.write('"')
            imgfile.close()
            state = STATE_INTAG
            continue
        else:
            # the tag was not an "img" tag, or did not have the format we require
            state = STATE_INTAG
            continue
    elif state == STATE_INTAG:
        match = buffer.find(">")
        if match != -1:
            # right brace found
            outputfile.write(buffer[:match+1])
            buffer = buffer[match+1:]
            state = STATE_NOMATCH
            continue
        else:
            outputfile.write(buffer)
            buffer = ""

    read = inputfile.read(4096)
    if read == "":
        done = 1
    buffer += read

outputfile.write(buffer)
inputfile.close()
outputfile.close()