How to extract the Table data which spread across the multiple pages( 1-3) and last page has another Table?

[thangarajdeivasikamani]

Hello Team,

I have the pdf as below
page :1

pdf which has 2 columns and 4 rows

First row content occupies full page.

Column Name | description

Test1 | 60 to 70 lines of text

Page: 2
Second row occupies fist half of the page

Test2 | 20 to 30 lines of text

Third row start second half of the page.

Test3 | 30 to 40 lines of text

Page: 3

Third rows ends in3rd page

Test3 | 5 lines

Fourth rows

Test3 | 2 lines

in Page 3 new Table starting as below

5 columns 2 rows

Column Name

               Column 1  | Column 2 | Column 3 | Column 4 | Column 5

Row 1 1 | 2 | 3 | 4 | 5
Row 2 1 0 | 11 | 13 | 14 | 15

Summary:

Case 1:

Table 1 start from page 1 and End at Page 3

Table 2 start from Page 3 and End at page 3

Case 2:

Table 1 start from page 1 and End at Page 3

Table 2 start from Page 3 and End at page 4

[JorjMcKie]

Please be aware that this way of posting questions leads to nothing:
You must provide a PDF or nobody can have any clue what is going on.

[thangarajdeivasikamani]

Test_multipage.pdf

Attached the pdf

[JorjMcKie]

Sorry - you did not attach the PDF!

[thangarajdeivasikamani]

Sorry. Pls recheck

[JorjMcKie]

I extracted the tables on all pages without issue.
what is your problem?

[chandan-artpark]

His problem is not with extracting the tables themselves, but rather understanding how to handle the tables that span across multiple pages. Specifically:

1. How to extract and recombine tables that start on one page and end on another (e.g., Table 1 which spans pages 1 to 3)?
2. How to handle cases where a new table starts on the same page where the previous table ends (e.g., Table 2 starts on page 3 right after Table 1)?

Ps: Even I am dealing with this issue, any insights on this matter is greatly appreciated.

[JorjMcKie]

You need to develop your own code to deal with these situations.
The table finder doesn't know better than you whether or not a table on a page continues the last one on a previous page.
Obvious criteria are same column count. If you are lucky, the table header row may have been repeated. But otherwise ...

But when you determine a continuation, you must use table extractions like table.extract()[:] or table.to_pandas() to join table content.

[JorjMcKie]

There is a table joiner Jupyter notebook here.

[chandan-artpark]

Thank you for the quick replies, will let you know if I find a good solution to this problem, or if I have any further queries.