Jinja2 for word templating

问题

I would like to use jinja2 for word templating like mentioned is this short article. The problem I'm facing is as follows, if I put {{title}} in my word-file the resulting xml can look like this:

<w:r><w:t>{{</w:t></w:r><w:proofErr w:type="gramStart"/><w:r><w:t>title</w:t></w:r><w:proofErr w:type="gramEnd"/><w:r><w:t>}}</w:t></w:r></w:p>

so it is impossible for jinja to replace this accordingly. Is there a possibility to prevent word from splitting {{title}} in separate text elements? (if I copy from a text-editor it works fine)

回答1:

This is an issue that is in word, relating to the proofErr tag.

You have two solutions:

If you want to stick with Jinja2, you should always write your tags in one stroke. Eg, never hit backspace or edit a tag. You can also copy/paste it from an other editor.

I wrote a library, Docxtemplater that works even if the text-elements are splitted, eg it would replace:

<w:r>
 <w:t>
   {{
 </w:t>
</w:r>
<w:proofErr w:type="gramStart"/>
 <w:r>
  <w:t>title</w:t>
 </w:r>
<w:proofErr w:type="gramEnd"/>
 <w:r>
  <w:t>}}</w:t>
 </w:r>

by:

<w:r>
  <w:t>
    Your title
  </w:t>
</w:r>
<w:proofErr w:type="gramStart"/>
<w:r>
  <w:t></w:t>
</w:r>
<w:proofErr w:type="gramEnd"/>
<w:r>
  <w:t></w:t>
</w:r>

Docxtemplater can be programmed over a CLI or in JS Browser/Node.JS

回答2:

https://pypi.org/project/docxtpl/ is also a good alternative.

It extends python-docx, and allows you to just drop in Jinja2 tags wherever you want in an existing .docx document, rather than templating from scratch.

Example:

pip install docxtpl

Usage:

from docxtpl import DocxTemplate

doc = DocxTemplate("my_word_template.docx")
context = { 'company_name' : "Dr. Stubbs Orthopedics and Prosthetics" }
doc.render(context)
doc.save("generated_doc.docx")

If your template my_word_template.docx file looks like this:

[