docxtext.py

import lxml.etree
from zipfile import ZipFile


class DocX(object):
    DOCPATH = 'word/document.xml'

    def __init__(self, fn):
        with ZipFile(fn).open(self.DOCPATH) as f:
            self.doc = lxml.etree.parse(f)
        self.nsmap = self.doc.getroot().nsmap

    @property
    def paragraphs(self):
        paragraphs = []

        for p in self.doc.xpath('.//w:p', namespaces=self.nsmap):
            line = ''
            for text in p.xpath('.//w:t', namespaces=self.nsmap):
                line += text.text
            paragraphs.append(line)
        return paragraphs


if __name__ == '__main__':
    import codecs
    import argparse

    parser = argparse.ArgumentParser(description='.docx plain text extractor')

    parser.add_argument('docx',
                        help='the source docx')
    parser.add_argument('--output',
                        '-o',
                        default=None,
                        help='output file')

    args = parser.parse_args()

    plaintext = '\n'.join(DocX("/tmp/Gregory Pinto's Resume.docx").paragraphs)

    if args.output:
        with codecs.open(args.output, 'w', encoding='utf-8') as f:
            f.write(plaintext)
    else:
        print plaintext