Python-docx: How to Insert HTML into Docx?

[tbell511]

Is it possible to insert HTML into a Document using python-docx with styling applied?
The only thing I need to work are italics.

For example how to insert "Today is \Saturday\" with Saturday actually being inserted with italics?

Thanks!

[electron378]

I had a similar issue however our HTML is way more complex and I couldn't find any direct method, so we ended up translating HTML fragments into a run map and then something like:
par = doc.add_paragraph()
for run_item in run_map:
run = par.add_run(run_item["text_fragment"])
if run_item["type"] == "italic":
run.style.italic = True
Something like that, see the docs around run topic for exact methods.

[tbell511]

Wow thank you for the reply! Any way you could provide the link for that documentation?
Thanks,

[electron378]

Have a look at docx.text.run
This is an extract from what i did in a similar case (simplified for italics only):

import re
from docx import Document


test_items = [
    "Trailing <i>tag</i>",
    "<i>Leading</i> tag",
    "This <i>time a tag</i> is in the middle",
]

class HTMLHelper(object):
    """ Translates some html into word runs. """
    def __init__(self):
        self.get_tags = re.compile("(<[a-z,A-Z]+>)(.*?)(</[a-z,A-Z]+>)")

    def html_to_run_map(self, html_fragment):
        """ breakes an html fragment into a run map """
        ptr = 0
        run_map = []
        for match in self.get_tags.finditer(html_fragment):
            if match.start() > ptr:
                text = html_fragment[ptr:match.start()]
                if len(text) > 0:
                    run_map.append((text, "plain_text"))
            run_map.append((match.group(2), match.group(1)))
            ptr = match.end()
        if ptr < len(html_fragment) - 1:
            run_map.append((html_fragment[ptr:], "plain_text"))
        return run_map

    def insert_runs_from_html_map(self, paragraph, run_map):
        """ inserts some runs into a paragraph object. """
        for run_item in run_map:
            run = paragraph.add_run(run_item[0])
            if run_item[1] == "<i>":
                run.italic = True


doc = Document()
html_helper = HTMLHelper()
for test_item in test_items:
    run_map = html_helper.html_to_run_map(test_item)
    print "------------------------------\nTest item:", test_item
    print "\nRun map: ", run_map
    par = doc.add_paragraph()
    print "\n XML before:\n", par._element.xml
    html_helper.insert_runs_from_html_map(par, run_map)
    print "\n XML after:\n", par._element.xml
doc.save("test_run_mapping.docx")

hope it helps

[tbell511]

Thank you so much for spending time to do that! It looks great. I am going to try to implement it now.

[tbell511]

I am getting a type error while trying to make the run fragments.

if match.start > ptr:
TypeError: unorderable types: builtin_function_or_method() > int()]

Any Ideas on how to fix this?

[electron378]

I would expect it to be sorted by now, but for the sake of consistency:
it should be if match.start() > ptr: - simply missing ()

[realsby]

@electron378
its realy good sample but i wish you share full sample, not just for italic.
your sample doesnt work with alone br tag and double tag recursively inside each other immediately :(

[realsby]

I prepare something like this... This is enough for me..

from HTMLParser import HTMLParser
from htmlentitydefs import name2codepoint
from docx import Document

.....
document = Document()
document_html_parser = DocumentHTMLParser(document)

document_html_parser.add_paragraph_and_feed(html_code)
.....
class DocumentHTMLParser(HTMLParser):
    def __init__(self, document):
        HTMLParser.__init__(self)
        self.document = document
        self.paragraph = self.document.add_paragraph()
        self.run = self.paragraph.add_run()

    def add_paragraph_and_feed(self, html):
        self.paragraph = self.document.add_paragraph()
        self.run = self.paragraph.add_run()
        self.feed(html)

    def handle_starttag(self, tag, attrs):
        self.run = self.paragraph.add_run()
        if tag == "i":
            self.run.italic = True
        if tag == "b":
            self.run.bold = True
        if tag == "u":
            self.run.underline = True
        if tag in ["br", "ul", "ol"]:
            self.run.add_break()
        if tag == "li":
            self.run.add_text(u'● ')
        if tag == "p":
            self.run.add_break()
            # self.run.add_break()
            # self.run.add_tab()

    def handle_endtag(self, tag):
        if tag in ["br", "li", "ul", "ol"]:
            self.run.add_break()
        self.run = self.paragraph.add_run()

    def handle_data(self, data):
        self.run.add_text(data)

    def handle_entityref(self, name):
        c = unichr(name2codepoint[name])
        self.run.add_text(c)

    def handle_charref(self, name):
        if name.startswith('x'):
            c = unichr(int(name[1:], 16))
        else:
            c = unichr(int(name))
        self.run.add_text(c)