#!/usr/bin/python python
"""Traverse filespec of *.html looking for select tag and write out report of options."""
import os
import sys
from xml.sax.saxutils import escape
def process(filename, fp):
print "* Processing:", filename, "\n"
# parse the file
select_of_htmlFile = open(filename)
fp.write("Filename: " + filename + "
\n")
prefix = ""
inSelect = 0
line = select_of_htmlFile.readline( )
while line:
line = " " + line.strip( )
if line.find(" 0:
if inSelect:
fp.write(prefix + "
\n")
inSelect = 1
if line.find("id=") > 0:
v = line.index("id=")
dm = line[(v+3):(v+4)]
v2 = line[(v+4):].index(dm) + v + 4
fp.write("Dropdown: " + line[(v+4):v2] + "
\n")
if line.find("") > 0:
if inSelect:
fp.write(prefix + "\n")
inSelect = 0
elif line.find("document.write") > 0:
inSelect = inSelect
elif line.find(" 0 and line.find("<") > 0 and inSelect:
v2 = line.index("") + v2
v2 = line[v:].index("<") + v
fp.write(prefix + line[(v+1):v2] + "
\n")
prefix = ""
line = select_of_htmlFile.readline( )
select_of_htmlFile.close( )
if inSelect:
fp.write(prefix + "\n")
prefix = ""
inSelect = 0
def finder(fp, dirname, names):
"""Add files in the directory dirname to a list."""
for name in names:
if name.endswith(".html"):
path = os.path.join(dirname, name)
if os.path.isfile(path):
process(path, fp)
def main( ):
print "[extract_select_option_values.py started]"
xmlFd = open("extract_select_option_values.htm", "w")
xmlFd.write("\n")
xmlFd.write("\n")
os.path.walk(sys.argv[1], finder, xmlFd)
xmlFd.write("\n")
xmlFd.write("\n")
xmlFd.close( )
print "[extract_select_option_values.py finished]"
if __name__ == "__main__":
main( )