PDF page reader

Read a PDF file page by page: the reader emits exactly one record per page, with the full text of the page as content. A configurable amount of context from the surrounding pages is provided alongside each page. For example:

A PDF file
# Page 1
Lorem ipsum dolor sit amet, consectetur adipiscing elit.

# Page 2
Quisque ac dolor massa.

Reader configuration
Chunk size: 20

Output
Page 1; Chunk 0
  Content:          Lorem ipsum dolor sit amet, consectetur adipiscing elit.
  Previous content: (empty)
  Next content:     Quisque ac dolor mas
Page 2; Chunk 0
  Content:          Quisque ac dolor massa.
  Previous content: tur adipiscing elit.
  Next content:     (empty)

Parameters

Chunk size

The number of characters of surrounding context to provide with each page: the Previous content output carries the last Chunk size characters of the previous page, and the Next content output carries the first Chunk size characters of the next page. It does not split the page content itself. When unset, both context outputs are empty.

Output

Page index

The index of the current page being read from the PDF file (starting at 1).

Chunk index

Always 0: pages are not split into chunks.

Content

The full text content of the current page.

Previous content

The last Chunk size characters of the previous page's text. Empty on the first page or when no chunk size is set.

Next content

The first Chunk size characters of the next page's text. Empty on the last page or when no chunk size is set.

Metadata

Metadata of the current file.

Metadata content
- author (optional)
- title (optional)
- subject (optional)
- keywords (optional)
- creator (optional)
- producer (optional)
- creation_date (optional)
- modification_date (optional)

Parameters​

Output​

Parameters

Output