Measure space that content uses in a PDF file.

[douglasmiranda]

As a tl;dr; of this essay, I'll just provide with something I say at the end right here in the beggining:

More specifically if I could just be able to measure how much of the main body of pdf pages (not including header and footer) is filled with content I would be happy. Like main_body = 100% space of page, content fills 85% of page main_body.

Hi, so what I'm trying to do is, reading a pdf file, (that could be two column), identify those blocks of content and check the space being filled with content inside the body of that file.

Inspired by these articles:

• https://artifex.com/blog/extracting-text-from-multi-column-pages-a-practical-pymupdf-guide
• https://artifex.com/blog/extract-text-from-a-multi-column-document-using-pymupdf-inpython

Following the examples, I checked I can access the rect.height.

import pymupdf
from multi_column import column_boxes

doc = pymupdf.open("pdfs/2-column.pdf")
for page in doc:
    bboxes = column_boxes(page, footer_margin=40, header_margin=40, no_image_text=True)
    for rect in bboxes:
        print(rect.height)
        print(page.get_text(clip=rect, sort=True))

multi_column being: https://github.com/pymupdf/PyMuPDF-Utilities/blob/master/text-extraction/multi_column.py

Which leads me to believe that I can just run through my pages, (ignoring header and footer with the parameters of column_boxes), get the content blocks (rect) height (for this example I only care for height). And at the end I'll have how much of the page main body (ignoring header and footer) is filled with textual content.

Like, lets say I can measure the first block of content in the first image, having 15 centimeters.

Now for the actual question:

Is that value rect.height what I am looking for?

• https://pymupdf.readthedocs.io/en/latest/rect.html#Rect.height

If yes, what type of measure is that? Because in that case I would try to convert that unit of measurement to some other I want.

---

More specifically if I could just be able to measure how much of the main body of pdf pages (not including header and footer) is filled with content I would be happy. Like main_body = 100% space of page, content fills 85% of page main_body.

[JorjMcKie]

All measures are in points, 1 point = 1/72 inch.
Whether or not header/footer is included depends on your margins value.
Note: There is no way to compute what is a header or footer. So you must specify which parts to ignore via the margins parameter.

[douglasmiranda]

Oh I see, so if I know my PDFs are a4 size (595.28 points x 841.89 points).

My real content space is the total a4 size minus margins, footer and header. Then when iterating over the bboxes, accessing each rect width and height I'll be able to get what I need.

If I come up with some code I'll paste in the comments.

[douglasmiranda]

I did a proof of concept, it's actually not hard.

As long all you want is read the text blocks inside a PDF page, and calculate how much % they are occupying of the content area.

So, in my case, all I need to know is how much % of page height the text blocks occupy.

So if I get text blocks page.get_text("blocks") and inspect a single one I have something like (0, 387, 31, 396, 'text xyz', 25, 0), and 396 being the value I need if all I need is how far from the top my text block is. For reference that is y1 if you think in pymupdf.Rect.

My full proof of concept is here:

https://gist.github.com/douglasmiranda/5105e1ee71fecf2bdd923e6bef65f95a

Notes:

• it's easy to change this to actually get the whole area (width*height) filled with content. ref
• if your document has a simple formatting of text you might even just need to check the latest text block of each page, no need for a for loop
• if your document has multiple columns, this code may need rework. "may need" because in some 2 column documents you may be dealing with those "justified" columns where in the last page the content is equally divided in 2 columns for the rest of the content.