Export to PDF

[josephernest]

Hi, Congrats for this great project @scanny!

In a .docx file, I need to replace every text {{name}} by a variable in Python, and then export the .docx to PDF. This works:

doc = docx.Document('template.docx')
for paragraph in doc.paragraphs:
    paragraph.text = paragraph.text.replace('{{name}}', 'john')
doc.save('temp.docx')  
# this is very fast until here
# the next line is very slow (3 or 4 seconds for each iteration):
subprocess.call(r"C:\Program Files\LibreOffice\program\soffice.exe --headless --convert-to pdf temp.docx --outdir ." % i)

Is there a faster way, built-in in python-docx, or with another python library, to export a DOCX to PDF?

(I need to do this in batch for thousands of PDF)

[josephernest]

PS: is there maybe a way to connect python-docx to reportlab to produce PDF (instead of having to use soffice.exe --headless) @scanny?

[linuxkd]

Here is an option that might work for you.

1. Setup the above python code to take a filename parameter, and then call the script with a for loop that sends the filename to be iterated on.
2. Modify the code on the doc.save line to name the file based on the input filename. Like mydocument-clean.docx (or whatever makes sense for your use case)
3. Use a bash for loop to iterate through your files, something like this should probably work
   for i in $(\ls -1 *.docx; do python3 yourscriptname.py $i; done

[josephernest]

@linuxkd Thank you for your answer, but I don't really understand it. The slow part is to start a C:\Program Files\LibreOffice\program\soffice.exe --headless --convert-to pdf temp.docx --outdir . process for each file.

I found this solution in the meantime: you can pass multiple .docx in one pass:

C:\Program Files\LibreOffice\program\swriter.exe --headless --convert-to pdf --outdir . temp1.docx temp2.docx ... temp100.docx