PDF documents

edit

Convert a PDF document to images

Each page gives one image. --dpi sets the resolution, 150 dots per inch by default:

Meziantou.Framework.FileConversion.Tool report.pdf --output-format png --dpi 200

This writes report-1.png, report-2.png... next to the document, numbered with as many digits as the number of pages so that they sort in order. Output file names shows how to change the names.

--pages selects the pages, for every input: --pages 1 for the first page, --pages 1-3,7,10- for the pages 1 to 3, 7, and from 10 to the end.

OptionDescription
--dpiThe resolution of the images, in dots per inch. 150 by default
--pdf-passwordThe password of an encrypted document
--pdf-render-annotationsRender the annotations (form fields, stamps, comments). true by default

Convert images to a PDF document

The images of all the inputs go to one document, one image per page, in the order of the inputs:

Meziantou.Framework.FileConversion.Tool cover.jpg "pages/*.png" --output book.pdf

By default, each page has the size of its image, at the resolution the image declares (96 dots per inch when it declares none). --page-size places the images on pages of a fixed size, scaled to fit inside the margins and centered:

Meziantou.Framework.FileConversion.Tool "scans/*.jpg" --output scans.pdf --page-size A4 --page-margin 10
OptionDescription
--page-sizefit (the size of each image, the default), A3, A4, A5, Letter, Legal, or a size such as 210x297mm, 8.5x11in or 595x842pt
--page-marginThe margin around the images, in millimeters. 0 by default
--page-orientationAuto (the default) turns the page of a landscape image; Portrait and Landscape keep one orientation
--pdf-image-encodingAuto (the default) copies the JPEG files as they are and compresses the other images without loss; Jpeg compresses every image as JPEG (with --jpeg-quality); Lossless compresses every image without loss

To write one document per image, rather than one document for all of them, use --per-input:

Meziantou.Framework.FileConversion.Tool "scans/*.jpg" --output-format pdf --per-input

Merge PDF documents

The pages of PDF documents are copied as they are, with their text, their fonts and their links:

Meziantou.Framework.FileConversion.Tool part1.pdf part2.pdf --output merged.pdf
Meziantou.Framework.FileConversion.Tool report.pdf --output summary.pdf --pages 1-2

The structure of a tagged document is not kept, and neither are the form fields of the copied pages: to keep the values of a filled-in form, --pdf-flatten-forms draws the fields in the content of their pages first. The values can then no longer be changed.

Meziantou.Framework.FileConversion.Tool filled-form.pdf --output form-copy.pdf --pdf-flatten-forms

Split a PDF document

--per-page writes each page to its own document, named like the images of a PDF document (report-1.pdf, report-2.pdf...). It works for every input, and for comic book archives and TIFF files too:

Meziantou.Framework.FileConversion.Tool report.pdf --output-format pdf --per-page
Meziantou.Framework.FileConversion.Tool report.pdf --output "chapters/{page:00}.pdf" --per-page --pages 3-

Options of the PDF output

These options apply to every PDF output: images, EPUB books and comic book archives converted to PDF, and PDF documents merged, split or copied. To change a PDF document, convert it to a new PDF file.

Save as PDF/A

PDF/A is the PDF format for archiving: the document embeds what it needs, such as its fonts, so that it looks the same in every reader in the long term. --pdf-conformance sets the level:

Meziantou.Framework.FileConversion.Tool "scans/*.jpg" --output archive.pdf --pdf-conformance PdfA2B
Meziantou.Framework.FileConversion.Tool report.pdf --output report-archive.pdf --pdf-conformance PdfA2U
LevelDescription
PdfA2BPDF/A-2b: the pages look the same in every reader. The usual choice
PdfA2UPDF/A-2u: PDF/A-2b, and the text can be extracted as Unicode
PdfA3BPDF/A-3b: PDF/A-2b, and the document can embed files of any format
PdfA3UPDF/A-3u: PDF/A-3b, and the text can be extracted as Unicode

The conversion fixes what it can: it removes the JavaScript and the actions PDF/A does not allow, embeds substitute fonts for the standard fonts that are not embedded, and rebuilds the metadata. When a requirement cannot be met, the tool lists it and writes no file. The colors are declared as sRGB; a document whose colors are CMYK needs the ICC profile of its printing condition, set with --pdf-output-intent. The tool checks the requirements it knows: to be sure that a file conforms, check it with a validator such as veraPDF.

PDF/A does not allow encryption, so --pdf-conformance cannot be used with --pdf-output-password and --pdf-owner-password.

Make the file smaller

--pdf-max-image-dpi resamples the images whose resolution on the page is higher than this one. It makes a scanned document much smaller. The JPEG images are compressed as JPEG again, the other images without loss:

Meziantou.Framework.FileConversion.Tool scan.pdf --output scan-small.pdf --pdf-max-image-dpi 150 --pdf-optimize

--pdf-optimize also removes the resources that the pages do not use and the page thumbnails.

Fast web view

--pdf-linearize writes the pages in order, with an index that lets a viewer show the first pages of a document before it is fully downloaded. A linearized file cannot be encrypted.

Encrypt the document

--pdf-output-password encrypts the document (AES-256) with a password that is needed to open it. --pdf-owner-password sets a second password, which allows every operation; with it, --pdf-permissions lists the operations that the first password allows:

Meziantou.Framework.FileConversion.Tool report.pdf --output protected.pdf --pdf-output-password secret
Meziantou.Framework.FileConversion.Tool report.pdf --output protected.pdf --pdf-owner-password owner-secret --pdf-permissions Print,CopyContent

Without --pdf-output-password, the document opens without a password, and readers allow only the --pdf-permissions unless the owner password is given. The permissions are None, Print, PrintHighQuality, Modify, CopyContent, Annotate, FillForms, ExtractForAccessibility, Assemble and All, separated by commas. Readers enforce them, not the encryption: they do not protect the content against a tool that ignores them.

To remove the password of a document, open it with --pdf-password: the output is not encrypted.

Meziantou.Framework.FileConversion.Tool protected.pdf --output report.pdf --pdf-password secret
OptionDescription
--pdf-conformanceNone (the default), PdfA2B, PdfA2U, PdfA3B or PdfA3U
--pdf-output-intentThe ICC profile of the output intent of a PDF/A output. sRGB by default
--pdf-max-image-dpiResample the images above this resolution, in dots per inch
--pdf-optimizeRemove the unused resources and the page thumbnails
--pdf-linearizeLinearize the file for fast web view
--pdf-output-passwordEncrypt the document with the password needed to open it
--pdf-owner-passwordEncrypt the document with the password that allows every operation
--pdf-permissionsThe operations allowed without the owner password. Needs --pdf-owner-password
--pdf-flatten-formsDraw the form fields of the PDF inputs in their pages, so that their values are kept