Split PDF by Text, Barcode or Separator Sheet

About the Split PDF by Text, Barcode or Separator Sheet

Split PDF by Text turns one long PDF — a month of invoices, a run of payslips, a batch of statements or certificates, a pile of scanned documents — into one file per document, and names each file from what it says. Tell it where a new document starts: wherever a value such as the Invoice No. or Employee ID changes, at every page with certain words, at separator sheets you print and slip between paper documents before scanning, at barcodes or QR codes already on the pages, at blank pages, or every few pages.

You see the plan before anything is written: the first page of every new file with its value, a cut line between files, and the names they will get, such as INV-0042 Northwind Traders.pdf. Correct any split point on the thumbnails, then split: the pages are copied as they are, without re-compressing, and an optional CSV index lists every file with its pages and values. It all happens in your browser; the PDF is never uploaded. Without a pass the result is a free preview: the plan in full and the first files as watermarked page images; the files, their ZIP and the index download with a Premium pass or after unlocking this result.

Common uses

• Sending each customer only their own invoice from a bulk export of the billing system.
• Giving every employee their payslip or tax form from one PDF the payroll software printed.
• Filing scanned post, contracts or medical records as separate documents, named the way you file them.
• Breaking bank statements, certificates or reports out of a combined PDF before uploading them one by one.

Where a new file starts

• A value that changes: the words after a label you choose (Invoice No., Employee ID, Account number, Policy …) are read on every page. A new file starts wherever the value changes; pages without the value, such as an invoice’s second page, stay with the file before them. The value can also be the first group of a pattern (a regular expression such as Employee ID:\s*(\w+)).
• Words on the page: a new file starts at every page with the words (or the pattern), for example Page 1 of.
• Separator sheets: print one sheet per document, put each in front of a document and scan the whole stack. Each sheet carries a QR code with the file name and any fields you add; the sheets are left out of the files unless you keep them.
• Barcodes or QR codes: QR Code, Data Matrix, Code 128 and Code 39 on the pages themselves, a new file where the code changes or at every page with a code, optionally only codes that start with certain characters, and the coded pages left out when they are cover sheets.
• Blank pages: the page after one or more blank pages starts a new file, with the same sensitivity steps as Remove Blank Pages.
• Every N pages, for batches where every document has the same length.

Names and the index

A name pattern is plain text with parts in braces: {value} (the value that starts the file, the code, or the sheet’s name), {n} (01, 02 …), {source} (the PDF’s name), {first}, {last} and {pages}, {date} (the first date written on the file’s first page, kept as written, because 03/04 means another day in another country), {text} (its first line), {after:Label} (the words after Label on its first page, such as {after:Customer}) and {qr:field} (a separator sheet’s field). A file whose part finds nothing is named {source}-{n} and the list says so; two files with the same name get “(2)”.

The CSV index has one row per file: its name, its pages, the value, the code’s text and every separator-sheet field. Values that start like a spreadsheet formula (=, +, -, @) are written with a leading apostrophe, so a spreadsheet shows them as text instead of running them.

Scans and OCR

Text comes from the PDF’s own text layer, which digital invoices and payslips have. A scanned PDF holds only pictures of pages: then choose the language and press Read the text with OCR, and the text is recognised on your device by the same engine as OCR PDF — much slower than reading a text layer, usually a few seconds a page. Separator sheets, barcodes and blank pages need no OCR at all, which makes them the quickest way to split large scanned batches.

Limitations

• Up to 2,000 pages and 300 MB in one PDF. Split larger batches first with Split PDF.
• PDFs with a password to open must be unlocked first with Unlock PDF.
• Text the PDF stores in an unusual way (pages drawn as outlines, text in pictures) has no text layer to read: use OCR, or split at codes, separator sheets or blank pages instead.
• OCR reads clear printed text well; handwriting, very small print and poor scans are read with mistakes, so check the values in the plan before splitting.
• Each file keeps its pages, links, form fields and the bookmarks that point into them; a digital signature cannot survive any split, so signed documents lose their signature.

Privacy

On this device; its files download once

Everything happens in your browser. What you enter or open here is not uploaded or stored by MySmartCoPilot. Reading a scan with OCR downloads the recognition engine and the language data (about 5–7 MB) from this site the first time; finding codes downloads the code reader (under 1 MB). Your PDF is never uploaded.