add labels for regex to dict of data

regex-label-extraction.py

# Common data cleaning pattern: you need to find regex matches in some block of text, stored w/ labels in a 
# dictionary (i.e., from json), extract those matches, and store them as an additional labels.  
# here's a quick and dirty utility function to do so.  can then be used in loop or listcomp over entire dataset
import re
from collections import Counter
def pattern_extractor(instance, pattern, textlabel, newlabel):
    found = re.findall(pattern, instance[textlabel])
    if found:
        if isinstance(found[0], (list, tuple)):
            matches = [filter(None, x)[0] for x in found]
        else:
            matches = filter(None, found)
        instance[newlabel] = sorted(Counter(matches).most_common())
    return instance