Acquire / file · All operations

Extract text from documents

A variant of Read another file format

Pull the text, tables or metadata out of PDF, Word and other documents.

Status

In arcform

Not built yet

arcform does not do this yet. Its place on the roadmap

What other tools call it

Meridian's classification; an independent blind rating agreed on 95.7% of 324 rows (variant, version 3).

Alteryx Designer (1)
  • PDF to Textread 2026-09-27

    Extract text from PDF files, using OCR where pages are images

Apache Hop (1)
  • Apache Tikaread 2026-09-27

    Extract text content and metadata from documents of many formats with Apache Tika

Duckle (1)
  • PDF Pagesread 2026-09-25

    one row per PDF page with its text layer and document metadata

Power Query M (1)
  • Pdf.Tablesread 2026-09-27

    Extract the tables found in a PDF file

RapidMiner Studio (2)
  • Read Documentread 2026-09-27

    Read a file as a text document, optionally discarding XML or HTML tags

  • Read Office Fileread 2026-09-27

    Read a Word or PowerPoint file as a text document

VisiData (1)

Background

The Apache Tika toolkit detects and extracts metadata and text from over a thousand different file types, all through a single interface.[1] Portable Document Format (PDF) is a file format developed by Adobe in 1993 to present documents independently of software, hardware and operating systems.[2] PDF's emphasis on preserving visual appearance poses challenges to the targeted extraction of information such as text, images, tables and metadata.[2]

When to use it

A text document scanned to PDF without its text being recognised by optical character recognition (OCR) is an image, with no fonts or text properties.[2] In Alteryx, the PDF to Text tool can extract text characters directly from PDF files, and can also apply OCR to images that contain text.[3] In Alteryx's PDF to Text tool, extracting text characters only is up to 10x faster than OCR and is generally more accurate.[3] Read a document's text layer first, and turn to OCR only for the pages where that layer comes back empty. (Meridian's view)

A tagged PDF includes document structure and semantics information to enable reliable text extraction and accessibility.[2] Tagged PDF is optional, and is not required where a PDF file is intended only for print.[2] On a file that is not tagged, check the order of the extracted paragraphs before trusting it. (Meridian's view)

The method

Within a PDF's text strings, characters are shown using character codes that map to glyphs in the current font using an encoding.[2] For fonts with non-standard glyphs, a PDF needs a ToUnicode table if semantic information about the characters is to be preserved.[2] When text from a PDF comes out as gibberish, suspect a font with no mapping to Unicode before blaming the extractor. (Meridian's view)

Power Query's Pdf.Tables returns any tables found in a PDF, and by default combines similar tables on consecutive pages into a single table.[4] Pdf.Tables' EnforceBorderLines option chooses whether border lines are always cell boundaries or one hint among many, and by default they are a hint.[4]

Apache Tika originated as part of the Apache Nutch codebase, to provide content identification and extraction when crawling, and was separated out in 2007.[5] Apache Tika can get text from images by using the OCR software Tesseract.[5]

Sources

  1. [1]Apache Tika – Apache Tikadocumentation
  2. [2]PDF - Wikipediasecondary source
  3. [3]PDF to Textdocumentation
  4. [4]Pdf.Tables - PowerQuery M | Microsoft Learndocumentation
  5. [5]Apache Tika - Wikipediasecondary source