Pdftk man page (just a bit improved)

Irvi Firqotul Aini edited this page Nov 2, 2015 · 7 revisions

Dump from http://www.pdflabs.com/tools/pdftk-the-pdf-toolkit/, but with some notes on the supported features.

Note : Using this gem you won't need to right any of these commands (that is the point of a wrapper), keep reading if you are interested in contributing.


This terse, technical documentation of pdftk features is your pdftk manual. Pdftk is a command-line program, so you should use your computer terminal or command prompt to try these examples.

pdftk < input PDF files | - | PROMPT > [ input_pw < input PDF owner passwords | PROMPT > ] [ < operation > < operation arguments > ] [ output < output filename | - | PROMPT > ] [ < options ... > ]

Note : But it could be useful to understand the core of the processing. Some feature cannot be implemented in this gem. I'm speaking about interactive feature, no PROMPT will be provided, and pdftk will always run in silent mode.


A list of the input PDF files. If you plan to combine these PDFs (without using handles) then list files in the order you want them combined. Use - to pass a single PDF into pdftk via stdin. Input files can be associated with handles, where a handle is a single, upper-case letter:

< input PDF handle >=< input PDF filename >

Handles are often omitted. They are useful when specifying PDF passwords or page ranges, later. For example: A=input1.pdf B=input2.pdf

[ input_pw < input PDF owner passwords | PROMPT > ]

Input PDF owner passwords, if necessary, are associated with files by using their handles: < input PDF handle >=< input PDF file owner password >

If handles are not given, then passwords are associated with input files by order. Most pdftk features require that encrypted input PDF are accompanied by the owner password. If the input PDF has no owner password, then the user password must be given, instead. If the input PDF has no passwords, then no password should be given.

When running in do_ask mode, pdftk will prompt you for a password if the supplied password is incorrect or none was given.

Note : If the input file is replaced by the hyphen - then the tool will wait for stdin input. That is what we use to work on File, Tempfile, StringIO objects. It explains one of the most important limitation of pdftk. We can use use only one stream at a time (one for each input and output).


[ < operation > < operation arguments > ] If this optional argument is omitted, then pdftk runs in ’filter’ mode. Filter mode takes only one PDF input and creates a new PDF after applying all of the output options, like encryption and compression. Available operations are:

  • cat (partially supported)
  • shuffle (partially supported)
  • burst (fully supported)
  • generate_fdf (fully supported) (this operation doesn't always provide readable FDF, we support this feature by our means)
  • fill_form (fully supported)
  • background (fully supported)
  • multibackground (fully supported)
  • stamp (fully supported)
  • multistamp (fully supported)
  • dump_data (fully supported)
  • dump_data_utf8 (fully supported)
  • dump_data_fields (fully supported)
  • dump_data_fields_utf8 (fully supported)
  • update_info (fully supported)
  • update_info_utf8 (fully supported)
  • attach_files (partially supported)
  • unpack_files (fully supported)

Some operations take additional arguments, described below.

cat [ < page ranges > ]

Assembles ("catenates") pages from input PDFs to create a new PDF. Use cat to merge PDF pages or to split PDF pages from documents. You can also use it to rotate PDF pages. Page order in the new PDF is specified by the order of the given page ranges. Page ranges are described like this: [ < input PDF handle > ] [ < begin page number > [ -< end page number > [ < qualifier > ] ] ] [ < page rotation > ] For example: pdftk A=in1.pdf B=in2.pdf cat A1 B2-20even Aend output out.pdf

If no handle is given, pdftk uses the first input PDF. Otherwise, the handle identifies one of the input PDF files. The beginning and ending page numbers are one-based references to pages in the PDF file. The qualifier can be even or odd, and the page rotation can be N, S, E, W, L, R, or D.

If the handle is omitted from the page range, then the pages are taken from the first input PDF. The even qualifier causes pdftk to use only the even-numbered PDF pages, so 1-6even yields pages 2, 4 and 6 in that order. 6-1even yields pages 6, 4 and 2 in that order. The odd qualifier works similarly to the even.

The page rotation setting can cause pdftk to rotate pages and documents. Each option sets the page rotation as follows (in degrees): N: 0, E: 90, S: 180, W: 270, L: -90, R: +90, D: +180. L, R, and D make relative adjustments to a page’s rotation.

If no arguments are passed to cat, then pdftk merges all input PDFs in the order they were given to create the output. NOTES:

  • < end page number > can be less than < begin page number >.
  • The keyword end can be used to reference the final page of a document instead of a page number.
  • Reference a single page by omitting the ending page number in the page range.
  • The handle can be used alone to represent the entire PDF document, e.g., B1-end is the same as B.

Page range examples without handles:

  • 1-endE - rotate entire document 90 degrees
  • 5 11 20
  • 5-25oddW - take odd pages in range, rotate 90 degrees
  • 6-1

Page Range Examples Using Handles: Say A=in1.pdf B=in2.pdf, then:

  • A1-21 - take range from in1.pdf
  • Bend-1odd - take all odd pages from in2.pdf in reverse order
  • A72 - take a single page from in1.pdf
  • A1-21 Beven A72 - assemble pages from both in1.pdf and in2.pdf
  • AW - rotate entire document 90 degrees
  • B - use all of in2.pdf
  • A2-30evenL - take the even pages from the range, remove 90 degrees from each page’s rotation
  • A A - catenate in1.pdf with in1.pdf
  • AevenW AoddE - apply rotations to even pages, odd pages from in1.pdf
  • AW BW BD - catenate rotated documents

shuffle [ <'page ranges> ]

Collates pages from input PDFs to create a new PDF. Works like the cat operation except that it takes one page at a time from each page range to assemble the output PDF. If one range runs out of pages, it continues with the remaining ranges. Ranges can use all of the features described above for cat, like reverse page ranges, multiple ranges from a single PDF, and page rotation. This feature was designed to help collate PDF pages after scanning paper documents.


Splits a single, input PDF document into individual pages. Also creates a report named doc_data.txt which is the same as the output from dump_data. If the output section is omitted, then PDF pages arenamed: pg_%04d.pdf, e.g.: pg_0001.pdf, pg_0002.pdf, etc. To name these pages yourself, supply a printf-styled format string in the output section. For example, if you want pages named: page_01.pdf, page_02.pdf, etc.,pass output page_%02d.pdf to pdftk. Encryption can be applied to the output by appending output options such as owner_pw, e.g.: pdftk in.pdf burst owner_pw foopass


Reads a single, input PDF file and generates an FDF file suitable for fill_form. It saves this FDF file using the output filename. If no output filename is give, it outputs the FDF to stdout. Does not create a new PDF.

fill_form < FDF data filename | XFDF data filename | - | PROMPT >

Fills the single input PDF’s form fields with the data from an FDF file, XFDF file or stdin. Enter the data filename after fill_form, or use - to pass the data via stdin, like so: pdftk form.pdf fill_form data.fdf output form.filled.pdf After filling a form,the form fields remain interactive unless you also use the flatten output option. flatten merges the form fields with the PDF pages. You can use flatten alone, too, but only on a single PDF: pdftk form.pdf fill_form data.fdf output out.pdf flatten or: pdftk form.filled.pdf output out.pdf flatten If the input FDF file includes Rich Text formatted data in addition to plain text, then the RichText data is packed into the form fields as well as the plain text. Pdftk also sets a flag that cues Acrobat/Reader to generate new field appearances based on the Rich Text data. That way, when the user opens the PDF, the viewer will create the Rich Text fields on the spot. If the user’s PDF viewer does not support Rich Text, then the user will see the plain text data instead. If you flatten this form before Acrobat has a chance to create (and save) new field appearances, then the plain text field data is what you’ll see in the flattened PDF.

background < background PDF filename | - | PROMPT >

Applies a PDF watermark to the background of a single input PDF. Pass the background PDF’s filename after background like so: pdftk in.pdf background back.pdf output out.pdf Pdftk uses only the first page from the background PDF and applies it to every page of the input PDF. This page is scaled and rotated as needed to fit the input page. You can use - to pass a background PDF into pdftk via stdin. If the input PDF does not have a transparent background (such as a PDF created from page scans) then the resulting background won’t be visible — use the stamp operation instead.

multibackground < multibackground PDF filename | - | PROMPT >

Same as the background operation, but applies each page of the background PDF to the corresponding page of the input PDF. If the input PDF has more pages than the stamp PDF, then the final stamp page is repeated across these remaining pages in the input PDF.

stamp < stamp PDF filename | - | PROMPT >

This behaves just like the background operation except it overlays the stamp PDF page on top of the input PDF document’s pages. This works best if the stamp PDF page has a transparent background.

multistamp < multistamp PDF filename | - | PROMPT >

Same as the stamp operation, but applies each page of the stamp PDF to the corresponding page of the input PDF. If the input PDF has more pages than the stamp PDF, then the final stamp page is repeated across these remaining pages in the input PDF. dump_data Reads a single, input PDF file and reports various statistics, metadata, bookmarks (a/k/a outlines), and page labels to the given output filename or (if no output is given) to stdout. Non-ASCII characters are encoded as XML numerical entities. Does not create a new PDF.


Same as dump_data excepct that the output is encoded as UTF-8.


Reads a single, input PDF file and reports form field statistics to the given output filename or (if no output is given) to stdout. Non-ASCII characters are encoded as XML numerical entities. Does not create a new PDF.


Same as dump_data_fields except that the output is encoded as UTF-8.

update_info < info data filename | - | PROMPT >

Changes the metadata stored in a single PDF’s Info dictionary to match the input data file. The input data file uses the same syntax as the output from dump_data. Non-ASCII characters should be encoded as XML numerical entities. This does not change the metadata stored in the PDF’s XMP stream, if it has one. For example: pdftk in.pdf update_info in.info output out.pdf update_info_utf8 <info data filename | - | PROMPT> Same as update_info except that the input is encoded as UTF-8.

attach_files < attachment filenames | PROMPT >

[ to_page < page number | PROMPT > ] Packs arbitrary files into a PDF using PDF’s file attachment features. More than one attachment can be listed after attach_files. Attachments are added at the document level unless the optional to_page option is given, in which case the files are attached to the given page number (the first page is 1, the final page is end). For example: pdftk in.pdf attach_files table1.html table2.html to_page 6 output out.pdf


Copies all of the attachments from the input PDF into the current folder or to an output directory given after output. For example: pdftk report.pdf unpack_files output ~/atts/ or, interactively: pdftk report.pdf unpack_files output PROMPT


[ output < output filename | - | PROMPT > ] The output PDF filename can’t be set to the name of an input filename. Use - to output to stdout. When using the dump_data operation, use output to set the name of the output data file. When using the unpack_files operation, use output to set the name of an output directory. When using the burst operation, you can use output to control the resulting PDF page filenames (described above).


[ encrypt_40bit | encrypt_128bit ]

If an output PDF user or owner password is given, output PDF encryption strength defaults to 128 bits. This can be overridden by specifying encrypt_40bit.

Encrypt a PDF using 128-bit strength (the default), withhold all permissions (the default)

pdftk 1.pdf output 1.128.pdf owner_pw foopass

Same as above, except password baz must also be used to open output PDF

pdftk 1.pdf output 1.128.pdf owner_pw foo user_pw baz

Same as above, except printing is allowed (once the PDF is open)

pdftk 1.pdf output 1.128.pdf owner_pw foo user_pw baz allow printing

[ allow < permissions > ]

Permissions are applied to the output PDF only if an encryption strength is specified or an owner or user password is given. If permissions are not specified, they default to ’none,’ which means all of the following features are disabled. The permissions section can include one or more of the following features:

  • Printing – Top Quality Printing
  • DegradedPrinting – Lower Quality Printing
  • ModifyContents – Also allows Assembly
  • Assembly
  • CopyContents – Also allows ScreenReaders
  • ScreenReaders
  • ModifyAnnotations – Also allows FillIn
  • FillIn
  • AllFeatures – Allows the user to perform all of the above, and top quality printing.

[ owner_pw < owner password | PROMPT > ]

[ user_pw < user password | PROMPT > ]

If an encryption strength is given but no passwords are supplied, then the owner and user passwords remain empty. This means that the resulting PDF can be opened and its security parameters altered by anybody.

[ flatten ]

Use this option to merge an input PDF’s interactive form fields (and their data) with the PDF’s pages. Only one input PDF can be given. Sometimes used with the fill_form operation.

[ compress | uncompress ]

These are only useful when you want to edit PDF page code in a text editor like vim or emacs. Remove PDF page stream compression by applying the uncompress filter. Use the compress filter to restore page stream compression.

[ keep_first_id | keep_final_id ]

When combining pages from multiple PDFs, use one of these options to copy the document ID from either the first or final input document into the new output PDF. Otherwise pdftk creates a new document ID for the output PDF. When no operation is given, pdftk always uses the ID from the (single) input PDF.

[ drop_xfa ]

If your input PDF is a form created using Acrobat 7 or Adobe Designer, then it probably has XFA data. Filling such a form using pdftk yields a PDF with data that fails to display in Acrobat 7. The workaround solution is to remove the form’s XFA data, either before you fill the form using pdftk or at the time you fill the form. Using this option causes pdftk to omit the XFA data from the output PDF form. This option is only needed when running pdftk on a single input PDF. When assembling a PDF from multiple inputs using pdftk, any XFA data in the input is automatically omitted. [ verbose ] By default, pdftk runs quietly. Append verbose to the end and it will speak up.

[ dont_ask | do_ask ]

Depending on the compile-time settings (see ASK_ABOUT_WARNINGS), pdftk might prompt you for further input when it encounters a problem, such as a bad password. Override this default behavior by adding dont_ask (so pdftk won’t ask you what to do) or do_ask (so pdftk will ask you what to do). When running in dont_ask mode, pdftk will over-write files with its output without notice.

[ decrypt ]

pdftk secured.pdf input_pw foopass output unsecured.pdf