Hadoop MapReduce job on file containing HTML tags

I have a bunch of large HTML files and I want to run a Hadoop MapReduce job on them to find the most frequently used words. I wrote both my mapper and reducer in Python and used Hadoop streaming to run them.

Here is my mapper:

#!/usr/bin/env python

import sys
import re
import string

def remove_html_tags(in_text):
'''
Remove any HTML tags that are found. 

'''
    global flag
    in_text=in_text.lstrip()
    in_text=in_text.rstrip()
    in_text=in_text+"\n"

    if flag==True: 
        in_text="<"+in_text
        flag=False
    if re.search('^<',in_text)!=None and re.search('(>\n+)$', in_text)==None: 
        in_text=in_text+">"
        flag=True
    p = re.compile(r'<[^<]*?>')
    in_text=p.sub('', in_text)
    return in_text

# input comes from STDIN (standard input)
global flag
flag=False
for line in sys.stdin:
    # remove leading and trailing whitespace, set to lowercase and remove HTMl tags
    line = line.strip().lower()
    line = remove_html_tags(line)
    # split the line into words
    words = line.split()
    # increase counters
    for word in words:
       # write the results to STDOUT (standard output);
       # what we output here will be the input for the
       # Reduce step, i.e. the input for reducer.py
       #
       # tab-delimited; the trivial word count is 1
       if word =='': continue
       for c in string.punctuation:
           word= word.replace(c,'')

       print '%s\t%s' % (word, 开发者_运维知识库1)

Here is my reducer:

#!/usr/bin/env python

from operator import itemgetter
import sys

# maps words to their counts
word2count = {}

# input comes from STDIN
for line in sys.stdin:
    # remove leading and trailing whitespace
    line = line.strip()

    # parse the input we got from mapper.py
    word, count = line.split('\t', 1)
    # convert count (currently a string) to int
    try:
        count = int(count)
        word2count[word] = word2count.get(word, 0) + count
    except ValueError:
        pass

sorted_word2count = sorted(word2count.iteritems(), 
key=lambda(k,v):(v,k),reverse=True)

# write the results to STDOUT (standard output)
for word, count in sorted_word2count:
    print '%s\t%s'% (word, count)

Whenever I just pipe a small sample small string like 'hello world hello hello world ...' I get the proper output of a ranked list. However, when I try to use a small HTML file, and try using cat to pipe the HTML into my mapper, I get the following error (input2 contains some HTML code):

rohanbk@hadoop:~$ cat input2 | /home/rohanbk/mapper.py | sort | /home/rohanbk/reducer.py
Traceback (most recent call last):
  File "/home/rohanbk/reducer.py", line 15, in <module>
    word, count = line.split('\t', 1)
ValueError: need more than 1 value to unpack

Can anyone explain why I'm getting this? Also, what is a good way to debug a MapReduce job program?

---

You can reproduce the bug even with just:

echo "hello - world" | ./mapper.py  | sort | ./reducer.py

The issue is here:

if word =='': continue
for c in string.punctuation:
           word= word.replace(c,'')

If word is a single punctuation mark, as would be the case for the above input (after it is split), then it is converted to an empty string. So, just move the check for an empty string to after the replacement.