Adobe Systems Incorporated

345 Park Ave San Jose CA 95110 USA mahardy@adobe.com

Adobe Systems Incorporated

345 Park Ave San Jose CA 95110 USA masinter@adobe.com http://larry.masinter.net

Adobe Systems Incorporated

345 Park Ave San Jose CA 95110 USA dmarkovi@adobe.com

PDF Association

Neue Kantstrasse 14 Berlin 14057 Germany duff.johnson@pdfa.org

Global Graphics

2030 Cambourne Business Park Cambridge CB23 6DW UK martin.bailey@globalgraphics.com http://www.globalgraphics.com

Requests for Comment The Portable Document Format (PDF) is an ISO standard (ISO 32000-1:2008) defining a final-form document representation language in use for document exchange, including on the Internet, since 1993. This document provides an overview of the PDF format and updates the media type registration of application/pdf. It obsoletes RFC 3778.

This document is intended to provide updated information on the registration of the MIME Media Type application/pdf for documents defined in the PDF , "Portable Document Format", syntax. It obsoletes . PDF was originally envisioned as a way to reliably communicate and view printed information electronically across a wide variety of machine configurations, operating systems, and communication networks. PDF is used to represent "final form" formatted documents. PDF pages may include text, images, graphics and multimedia content such as video and audio. PDF is also capable of containing auxiliary structures including annotations, bookmarks, file attachments, hyperlinks, logical structure and metadata. These features are useful for navigation, building collections of related documents and for reviewing and commenting on documents. A rich JavaScript model has been defined for interacting with PDF documents. PDF used the imaging model of the PostScript page description language to render complex text, images, and graphics in a device and resolution-independent manner. PDF supports encryption and digital signatures. The encryption capability is combined with access control information to facilitate management of the functionality available to the recipient. PDF supports the inclusion of document and object-level metadata through the eXtensible Metadata Platform.

PDF is used widely in the Internet community. The first version of PDF, 1.0, was published in 1993 by Adobe Systems Incorporated. Since then PDF has grown to be a widely-used format for capturing and exchanging formatted documents electronically across the Web, via e-mail and virtually every other document exchange mechanism. In 2008, PDF 1.7 was published as an ISO standard , ISO 32000-1:2008. It was adopted using ISO Fast-Track process and is technically identical to Adobe Portable Document Format version 1.7 referenced by . The ISO TC-171 committee is presently working on a refresh of PDF, known as ISO 32000-2, with a version of PDF 2.0, expected to be published in 2017. In addition to ISO 32000-1:2008 and 32000-2, several subset standards have been defined to address specific use cases and standardized by the ISO. These standards include PDF for Archival (PDF/A) , PDF for Engineering (PDF/E) , PDF for Universal Accessibility (PDF/UA) , PDF for Variable Data and Transactional Printing (PDF/VT) , and PDF for Prepress Digital Data Exchange (PDF/X) . The subset standards are fully compliant PDF files capable of being displayed in a general PDF viewer.

Fragment identifiers appear at the end of a URI, and provide a way to reference an anchor to subordinate content within the target of the URI, or additional parameters to the process of opening the identified content. The syntax and semantics of fragment identifiers is referenced in the media type definition. The specification of fragment identifiers for PDF appeared originally in , but now will be included in ISO 32000-2 . This section is a summary of that material. Any disagreements between that document and this should be resolved in favor of the ISO 32000-2 definition, once that has been approved. A fragment identifier for PDF has one or more parameters, separated by the ampersand (&) or pound (#) character. Each parameter consists of the parameter name, "=" (equal), and the parameter value; lists of values are comma-separated, and parameter value strings may be URI-encoded (). Parameters are processed left to right. Coordinate values (such as <left>, <right>, <width>) are expressed in the default user space coordinate system of the document: 1/72 of an inch measured down and to the right from the upper-left corner of the (current) page. ( 8.3.2.3 "User Space") The following parameters identify subordinate content of a PDF file, but also may be used to set the document view to make the (start of) the identified content visible: Identifies a specified (physical) page; the first page in the document has a pageNum value of 1. Identifies a named destination ( 12.3.2.4 "Named destinations"). structID is a byte string with URI encoding; identifies the structure element with ID key within a StructElem dictionary of the document. The commentID is the value of an annotation name, which is defined by the NM key in the corresponding annotation dictionary (of the selected page. ( 12.5.2 "Annotation dictionaries") Identifies the embedded file where the parameter string <name> matches a file specification dictionary in the EmbeddedFiles name tree. If the "ef" parameter is not at the end of the fragment identifier, then the rest of the fragment identifier (after the ampersand or hash delimiter) is applied to the embedded file according to its own media type. This allows identification of content within the embedded file (which itself might be a PDF file). NOTE: When opening a PDF file that is not from a trusted source, processor may choose to prompt the user or even prevent opening of the file. These parameters also operate on the view of the PDF document when it is opened. <scale> is the percentage to which the document should be zoomed, where a value of 100 correspond to a zoom of 100%. <left> and <top> are optional, but both must be specified if either is included. The arguments correspond to those found in 12.3.2.2 "Explicit destinations". keyword is one of the keywords defined in "Table 149: Destination syntax" with appropriate position values. Set the view rectangle. Highlight the specified rectangle. Open the document and search for one or more words, selecting the first matching word in the document. wordList is a string enclosed in quotation marks where individual words are separated by the space character (or %20). Imports data into PDF form fields. The URI is either a relative or absolute URI to an FDF or XFDF file. The fdf parameter should be specified as the last parameter to a given URI.

Several subsets of PDF have been published as distinct ISO standards: PDF/X, initially released in 2001 as PDF/X-1a , specifies how to use PDF for graphics exchange, with the aim to fascilitate correct and predictable printing by print service providers. The standard has gone through multiple revisions over the years and has several published parts, the most recently released being part 8, specifying different levels of conformance: PDF/X-1a:2001, PDF/X-3:2002, PDF/X-1a:2003, PDF/X-3:2003, PDF/X-4, PDF/X-4p, PDF/X-5g, PDF/X-5pg and PDF/X-5n. PDF/A, initially released in 2005, specifies how to use PDF for long-term preservation (archiving) of electronic documents. It prohibits PDF features which are not well suited to long term archiving of documents, including JavaScript or executable file launches. Its requirements for PDF/A viewers include color management guidelines and support for embedded fonts. There are three parts of this standard and a total of eight conformance levels: PDF/A-1a, PDF/A-1b, PDF/A-2a, PDF/A-2b, PDF/A-2u, PDF/A-3a, PDF/A-3b and PDF/A-3u. PDF/E, initially released in 2008 as PDF/E-1 , specifies how to use PDF in engineering workflows, such as manufacturing, construction and geospatial analysis. Future revisions of PDF/E are supposed to include support for 3D PDF workflows. PDF/VT, initially released in 2010, specifies how to use PDF in variable and transactional printing. It is based on PDF/X, and adds adidtional restrictions on PDF content elements and supporting metadata. It specifies three conformance levels: PDF/VT-1, PDF/VT-2 and PDF/VT-2s . PDF/UA, initially released in 2012 as PDF/UA-1 , specifies how to create accessible electonic documents. It requires use of ISO 32000's Tagged PDF feature, and adds many requirements regarding semantic correctness in applying logical structures to content in PDF documents. All of these subset standards use application/pdf media type. The subset standards are generally not exclusive, so it is possible to construct a PDF file which conforms to, for example, both PDF/A-2b and PDF/X-4 subset standards. PDF documents claiming conformance to one or more of the subset standards use XMP metadata to identify levels of conformance. PDF processors should examine document metadata streams for such subset standards identifiers and, if apropriate, label documents as such when presenting them to the user.

PDF format has gone through several revisions, primarily for the addition of features. PDF features have generally been added in a way that older viewers "fail gracefully", because they can just ignore features they do not recognize. Even so, the older the PDF version produced, the more legacy viewers will support that version, but the fewer features will be enabled. See Annex I, "PDF Versions and Compatibility".

PDF files are experienced through a reader or viewer of PDF files. For most of the common platforms in use (iOS, OS X, Windows, Android, ChromeOS, Kindle) and for most browsers (Edge, Safari, Chrome, Firefox), PDF viewing is built-in. In addition, there are many PDF viewers available for download and install. The PDF specification was published and freely available since the format was introduced in 1993, so hundreds of companies and organizations make tools for PDF creation, viewing, and manipulation.

PDF is certainly a complex media type as per Section 4.6 of , which sets requirements for security analysis of media type registrations. (which this document obsoletes) contained a detailed analysis of some of the security issues for PDF implementations known at the time. While the analysis isn't necessarily wrong, the threat analysis is much too limited, and the mitigations somewhat out of date. There is now extensive literature on security threats involving PDF implementations and how to avoid them, consistent with broad implementation over decades. We are not registering a new media type but rather making a primarily administrative update. With those caveats: The PDF file format allows several constructs which may compromise security if handled inadequately by PDF processors. For example: PDF may contain scripts to customize the displaying and processing of PDF files. These scripts are expressed in a version of JavaScript and are intended for execution by the PDF processor. PDF file may refer to other PDF files for portions of content. PDF processors are expected to find these external files and load them in order to display the document. PDF may act as a container for various files embedded in it (for example, as attached files). PDF processors may offer functionality to open and display such files or store them on the system, such as with the "ef" open action. THe PDF specification places no restrictions on types of files which may be embedded, so PDF processors should be extremely careful to prevent unwanted execution of attached executables or decompression of attached archives which may store dangerous files in the host file system. PDF files may contain links to content on the internet. PDF processors may offer functionality to show such content upon following the link. The fragment identifier syntax () contains directives for opening ("ef") or inluding ("fdf") additional material. PDF interpreters executing any scripts or programs related to these constructs must be extremely careful to insure that untrusted software is executed in a protected environment. In addition, the PDF processor itself, as well as its plugins, scripts etc. may be a source of insecurity, by either obvious or subtle means.

This document updates the registration of application/pdf, a media type registration as defined in : Type name: application Subtype name: pdf Required parameters: none Optional parameter: none Encoding considerations: binary Security considerations: See of this document. Interoperability considerations: See of this document. Published specification: ISO 32000-1:2008 (PDF 1.7) . ISO 32000-2 (PDF 2.0) is currently under development. Applications which use this media type: See of this document. Fragment identifier considerations: See of this document. Additional information: Deprecated alias names for this type: none Magic number(s): All PDF files start with the characters '%PDF-' followed by the PDF version number, e.g., "%PDF-1.7". These characters are in US-ASCII encoding. File extension(s): .pdf Macintosh file type code(s): "PDF " Person & email address to contact for further information: Duff Johnson <duff@duff-johnson.com>, Peter Wyatt <Peter.wyatt@cisra.canon.com.au>, ISO 32000 Project Leaders Intended usage: COMMON Restrictions on usage: none Author: Authors of this document Change controller: ISO; in particular, ISO 32000 is by ISO/TC 171/SC 02/WG 08, "PDF specification". Duff Johnson <duff@duff-johnson.com> and Peter Wyatt <Peter.wyatt@cisra.canon.com.au are current ISO 32000 Project Leaders.

ISO Also available free from Adobe. ISO Currently under development - publication expected in 2017. This becomes a Normative Reference on approval. ISO ISO ISO ISO ISO ISO Not available for free, but there are a number of descriptive resources, e.g., Adobe Systems Incorporated Adobe Systems Incorporated

This specification replaces RFC 3778, which previously defined the application/pdf Media Type. Differences include: To reflect the transition from a proprietary specification by Adobe to an open ISO Standard, the Change Controller has changed from Adobe to ISO, and references updated. The overview of PDF capabilitiies, the history of PDF, and the descriptions of PDF subsets were updated to reflect more recent relevant history. The section on Fragment identifiers was updated to closely reflect the material which has been added to ISO-32000-2. The status of popular PDF implementations was updated. The Security Considerations were updated to match the current understanding of PDF vulnerabilities. The registration template was updated to match RFC 6838.