[Python] How can I extract a table from a PDF document where the cells contain multi-line text...

I'm working on extracting bank transaction data from a PDF shown below using Python. Each transaction includes two dates, amounts (Money In, Money Out, Balance), and a description. The challenge is that the descriptions span a variable number of lines, and their position isn't consistent relative to the transaction line:

• Sometimes the description appears one line above the date/amount line.
• Other times, it appears one or two lines below.
• In some cases, it's a single line, while in others, it can span two or even three lines.

This inconsistency causes issues when saving the data to Excel — the descriptions end up misaligned or split across rows.

This is what the PDF file looks like:



Extracted data:



def extract_data():
"""Extracts data from the specified pdf file"""
content = ""
with pdfplumber.open(fpath) as pdf:
for page in pdf.pages:
text = page.extract_text(
x_tolerance=float(xTolerance),
y_tolerance=float(yTolerance),
layout=True,
x_density=float(xDensity),
y_density=float(yDensity)
)
content += text

Continue reading...