Pagecraft

pdf-to-csv

PDF to CSV

Pull the tables out of a PDF as comma-separated values, one file per table.

Reads
a PDF — application/pdf
Writes
any number of files — CSV
Tool page
/pdf-to-csv

Example

Get the rows of a printed table into a spreadsheet.

Given

  • accounts.pdf — a PDF holding a ruled table

From code

            import { loadDocument, runOperation } from '@pagecraft/core';

const accounts = await loadDocument({ bytes, name: 'accounts.pdf' });

const [output] = await runOperation('pdf-to-csv', [accounts], {
  pages: 'all',
});

From a shell

            pagecraft pdf-to-csv accounts.pdf --pages all

What comes back

One CSV file, containing “Region”.

This example is run by the test suite. If it stopped being true, the build would fail.

Options

pages--pages

string, a page selection: '1-3, 7, 9-', 'even', 'last'

Pages

Which pages to take the tables from.

Default: 'all'

bom--bom

boolean

Mark it as UTF-8

Start the file with a byte order mark, without which Excel mangles every accent.

Default: true

What it cannot do

  • A PDF does not record what was a table; it records text in places. A table is found by the lines drawn around it, or by its columns lining up, and a table that has neither is read as ordinary text.
  • CSV has no way to say that two cells are one, so a merged cell arrives as its text in the place it starts and empty cells where it reached. Use PDF to Excel to keep the merge itself.
  • One table becomes one file. A table split across two pages becomes two files, because nothing in the file says the second is a continuation of the first.
  • A scanned page holds a picture of a table. Run OCR over it first.