Extract data from non-HTML documents

The Crawler can extract data from files like PDF and Word documents. To do this, it uses Apache Tika to extract a document’s content and transform it into a basic HTML file.

Limitations

Because it’s difficult to translate non-HTML documents into HTML, there are limitations:

PDF documents can break if it’s exported with an unknown font.
The transformed HTML has little semantic value: headings, paragraphs, and lists in the original document might not be marked in the HTML. This makes good relevancy hard to achieve.
Document indexing is slower than classic HTML indexing.
Language detection isn’t available.

Enable document extraction

To enable document extraction, add the fileTypesToMatch parameter to at least one of your crawler’s actions. For a list of supported file types, see Supported file types. The document’s transformed HTML is stored in the recordExtractor.$ parameter. The file type is stored in the recordExtractor.fileType parameter.

JavaScript

({
  // [...]
  actions: [
    {
      indexName: "crawler-example",
      pathsToMatch: ["https://www.example.com/**"],
      fileTypesToMatch: ["pdf", "doc"],
      recordExtractor: ({ url, $, fileType }) => {
        console.log($.html(), fileType);
      },
    },
  ],
});

Sample crawler configuration

For an example configuration for document extraction, see config.documents.js on GitHub.

Supported file types

The metadata shown in these examples might not be included on every page.

PDF documents

Associated extension: .pdf
fileTypesToMatch: pdf

For example, in this .pdf file, Tika exposes the following HTML which the crawler then passes to your recordExtractor.

A PDF document with the text: Test PDF file content

HTML

<html xmlns="http://www.w3.org/1999/xhtml">
<head>
  <meta name="date" content="2018-07-17T13:35:40Z"/>
  <meta name="pdf:PDFVersion" content="1.3"/>
  <meta name="pdf:docinfo:title" content="test-docx-file.pages"/>
  <meta name="xmp:CreatorTool" content="Pages"/>
  <meta name="access_permission:modify_annotations" content="true"/>
  <meta name="access_permission:can_print_degraded" content="true"/>
  <meta name="dcterms:created" content="2018-07-17T13:35:40Z"/>
  <meta name="Last-Modified" content="2018-07-17T13:35:40Z"/>
  <meta name="dcterms:modified" content="2018-07-17T13:35:40Z"/>
  <meta name="dc:format" content="application/pdf; version=1.3"/>
  <meta name="Last-Save-Date" content="2018-07-17T13:35:40Z"/>
  <meta name="pdf:docinfo:creator_tool" content="Pages"/>
  <meta name="access_permission:fill_in_form" content="true"/>
  <meta name="pdf:docinfo:modified" content="2018-07-17T13:35:40Z"/>
  <meta name="meta:save-date" content="2018-07-17T13:35:40Z"/>
  <meta name="pdf:encrypted" content="false"/>
  <meta name="dc:title" content="test-docx-file.pages"/>
  <meta name="modified" content="2018-07-17T13:35:40Z"/>
  <meta name="Content-Type" content="application/pdf"/>
  <meta name="X-Parsed-By" content="org.apache.tika.parser.DefaultParser"/>
  <meta name="X-Parsed-By" content="org.apache.tika.parser.pdf.PDFParser"/>
  <meta name="meta:creation-date" content="2018-07-17T13:35:40Z"/>
  <meta name="created" content="Tue Jul 17 13:35:40 UTC 2018"/>
  <meta name="access_permission:extract_for_accessibility" content="true"/>
  <meta name="access_permission:assemble_document" content="true"/>
  <meta name="xmpTPg:NPages" content="1"/>
  <meta name="Creation-Date" content="2018-07-17T13:35:40Z"/>
  <meta name="access_permission:extract_content" content="true"/>
  <meta name="access_permission:can_print" content="true"/>
  <meta name="producer" content="Mac OS X 10.13.5 Quartz PDFContext"/>
  <meta name="access_permission:can_modify" content="true"/>
  <meta name="pdf:docinfo:producer" content="Mac OS X 10.13.5 Quartz PDFContext"/>
  <meta name="pdf:docinfo:created" content="2018-07-17T13:35:40Z"/>
  <title>test-docx-file.pages</title>
</head>
<body>
  <div class="page">
    <p/>
    <p>Test PDF file content</p>
    <p/>
  </div>
</body>
</html>

Microsoft Word documents

Associated extensions: .doc, .docx
fileTypesToMatch: doc

For example, in this .doc file, Tika exposes the following HTML, which the crawler then passes to its recordExtractor.

A Microsoft Word document with the text: Test DOC file content

HTML

<html xmlns="http://www.w3.org/1999/xhtml">
<head>
  <meta name="X-Parsed-By" content="org.apache.tika.parser.DefaultParser"/>
  <meta name="X-Parsed-By" content="org.apache.tika.parser.microsoft.OfficeParser"/>
  <meta name="Content-Type" content="application/msword"/>
  <title>
  </title>
</head>
<body>
  <div class="header"/>
    <p class="body">Test DOC file content</p>
  <div class="footer"/>
</body>
</html>

OpenDocument text documents

Associated extension: .odt
fileTypesToMatch: odt

Microsoft Excel spreadsheets

Associated extensions: .xls, .xlsx
fileTypesToMatch: xls

For example, in this .xls file, Tika exposes the following HTML, which the crawler then passes to its recordExtractor.

A Microsoft Excel spreadsheet with the text: Test XML file content

HTML

<html xmlns="http://www.w3.org/1999/xhtml">
  <head>
    <meta name="X-Parsed-By" content="org.apache.tika.parser.DefaultParser"/>
    <meta name="X-Parsed-By" content="org.apache.tika.parser.microsoft.OfficeParser"/>
    <meta name="Content-Type" content="application/vnd.ms-excel"/>
    <title>
    </title>
  </head>
  <body>
    <div class="page">
      <h1>Feuille 1</h1>
      <table>
        <tbody>
          <tr>
            <td>Test XLS file content</td>
          </tr>
        </tbody>
      </table>
      <div class="outside">&amp;C&amp;"Helvetica,Regular"&amp;12&amp;K000000&amp;P</div>
    </div>
  </body>
</html>

OpenDocument spreadsheets

Associated extension: .ods
fileTypesToMatch: ods

Microsoft PowerPoint documents

Associated extensions: .ppt, .pptx
fileTypesToMatch: ppt

For example, in this .ppt file, Tika exposes the following HTML, which the crawler then passes to its recordExtractor.

A Microsoft PowerPoint slide with the text: Test PPT file content

HTML

<html xmlns="http://www.w3.org/1999/xhtml">
  <head>
    <meta name="X-Parsed-By" content="org.apache.tika.parser.DefaultParser"/>
    <meta name="X-Parsed-By" content="org.apache.tika.parser.microsoft.OfficeParser"/>
    <meta name="Content-Type" content="application/vnd.ms-powerpoint"/>
    <title>
    </title>
  </head>
  <body>
    <div class="slideShow">
      <div class="slide">
        <div class="slide-master-content"/>
        <div class="slide-content">
          <p>Test PPT file content</p>
          <p/>
        </div>
      </div>
      <div class="ocr"/>
    </div>
  </body>
</html>

OpenDocument presentations

Associated extension: .odp
fileTypesToMatch: odp

Emails

Associated extension: .msg
fileTypesToMatch: email

The file type email includes all documents related to email. The Crawler supports the Outlook Mail Message (.msg) format. For example, Tika converts this email into the following HTML:

Screenshot of an email with 'From: from@domain.com', 'To: to@domain.com', 'Subject: this is a mail to test message file', and message text.

HTML

<html xmlns="http://www.w3.org/1999/xhtml">
<head>
  <meta name="date" content="2017-06-01T15:24:31Z" />
  <meta name="Message:To-Email" content="to@domain.com" />
  <meta name="dc:description" content="this is a mail to test msg file" />
  <meta name="subject" content="this is a mail to test msg file" />
  <meta name="dc:creator" content="from@domain.com" />
  <meta name="Message:From-Email" content="from@domain.com" />
  <meta name="dcterms:created" content="2017-06-01T15:24:31Z" />
  <meta name="Message-To" content="to@domain.com" />
  <meta name="dcterms:modified" content="2017-06-01T15:24:31Z" />
  <meta name="Last-Modified" content="2017-06-01T15:24:31Z" />
  <meta name="Message-Recipient-Address" content="to@domain.com" />
  <meta name="Message:Raw-Header:X-Unsent" content="1" />
  <meta name="Message:Raw-Header:Subject" content="this is a mail to test msg file" />
  <meta name="meta:mapi-message-class" content="NOTE" />
  <meta name="Message:To-Display-Name" content="to@domain.com" />
  <meta name="Last-Save-Date" content="2017-06-01T15:24:31Z" />
  <meta name="Message:Raw-Header:MIME-Version" content="1.0" />
  <meta name="meta:save-date" content="2017-06-01T15:24:31Z" />
  <meta name="dc:title" content="this is a mail to test msg file" />
  <meta name="Message:Raw-Header:Message-ID" content="<c58b1b52f61f4789ba40339c6e993440>" />
  <meta name="modified" content="2017-06-01T15:24:31Z" />
  <meta name="Content-Type" content="application/vnd.ms-outlook" />
  <meta name="X-Parsed-By" content="org.apache.tika.parser.DefaultParser" />
  <meta name="X-Parsed-By" content="org.apache.tika.parser.microsoft.OfficeParser" />
  <meta name="creator" content="from@domain.com" />
  <meta name="Message:Raw-Header:From" content="from@domain.com" />
  <meta name="meta:author" content="from@domain.com" />
  <meta name="meta:creation-date" content="2017-06-01T15:24:31Z" />
  <meta name="meta:mapi-from-representing-email" content="from@domain.com" />
  <meta name="Creation-Date" content="2017-06-01T15:24:31Z" />
  <meta name="Message-Cc" content="" />
  <meta name="Message-Bcc" content="" />
  <meta name="meta:mapi-from-representing-name" content="from@domain.com" />
  <meta name="Message:Raw-Header:To" content="to@domain.com" />
  <meta name="Message:From-Name" content="from@domain.com" />
  <meta name="Author" content="from@domain.com" />
  <meta name="Message-From" content="from@domain.com" />
  <meta name="Message:To-Name" content="" />
  <title>this is a mail to test msg file</title>
</head>
<body>
  <h1>this is a mail to test msg file</h1>
  <dl>
    <dt>From</dt>
    <dd>from@domain.com</dd>
    <dt>To</dt>
    <dd>to@domain.com</dd>
    <dt>Recipients</dt>
    <dd>to@domain.com</dd>
  </dl>
  <div class="message-body">
    <p>This message was sent using a msg file </p>
  </div>
</body>
</html>

Get started

Extract data

Enrich data

Troubleshooting

APIs

Netlify plugin

Extract data from non-HTML documents

Limitations

Enable document extraction

Sample crawler configuration

Supported file types

PDF documents

Microsoft Word documents

OpenDocument text documents

Microsoft Excel spreadsheets

OpenDocument spreadsheets

Microsoft PowerPoint documents

OpenDocument presentations

Emails

​Limitations

​Enable document extraction

​Sample crawler configuration

​Supported file types

​PDF documents

​Microsoft Word documents

​OpenDocument text documents

​Microsoft Excel spreadsheets

​OpenDocument spreadsheets

​Microsoft PowerPoint documents

​OpenDocument presentations

​Emails

Limitations

Enable document extraction

Sample crawler configuration

Supported file types

PDF documents

Microsoft Word documents

OpenDocument text documents

Microsoft Excel spreadsheets

OpenDocument spreadsheets

Microsoft PowerPoint documents

OpenDocument presentations

Emails