PDF/A-3: the edition an electronic invoice uses

In plain words

Written for anybody, with nothing taken for granted. It says what the rule is for and what it changes on the document you end up with.

The second edition let a document carry another file inside it only if that carried file was itself archival. The third lifts the condition: the carried file may be of any kind at all.

That one change is what makes an electronic invoice possible. The page is the invoice a person reads; tucked inside it is the same invoice as a data file that accounting software reads without anybody retyping a figure. One document, two readers, and it is archival for both of them.

Everything else is the second edition word for word: the same promise, the same three levels, the same doors closed.

For a developer

The same thing without the detour: the objects, the constraints, and where the line falls between what is produced and what is declared.

The part rests on PDF 1.7. The only substantive difference from the second edition is the rule on carried files: any format is allowed, and each carried file states its relationship to the document — the source it was made from, the data behind it, an alternative form of it, or a supplement.

Three levels, as in the second edition. Level A asks for a structure hierarchy and the engine refuses to write the file without one; level B asks for the page alone.

Through the rendering service this part is asked for as archive with the value pdfa3b. The document header says PDF 1.7, and a document whose version and claim disagree is refused before a byte is written.

Clause by clause

The published text behind each requirement, what the engine measures the document against, and what it leaves to a validator.

Two of the claims of this part are written by the engine: level B, which a request can ask for by name today, and level A, which the engine writes when the document carries a structure hierarchy. Level U is planned for a future version.

Both are checked from outside before they are published here: veraPDF reads the file against the text of the standard, and Ghostscript reads it a second time. A document is not called conforming because it was written here.

The levels of this part

Each level of this part, and how it is asked for
Level	What it adds	Asked for over the network as
`PDF/A-3A`	The page, its text mapped back to Unicode, and the structure of the document, so it can be read aloud and searched by paragraph.	Written by the engine; planned for a future version of a request
`PDF/A-3B`	The page alone. This is the level an electronic invoice is written at.	`pdfa3b`
`PDF/A-3U`	The page, and every character drawn mapped back to the letter it stands for.	Planned for a future version

What this part asks for

Everything the second edition asks for, unchanged.
Every carried file states what it is to the document carrying it, and carries its own name, its kind and the moment it was written.
A structure hierarchy, at level A.
The document header names PDF 1.7.

What this part rules out

A rule that closes a door is what makes the promise keepable: each of these is a way a document could stop being readable later.

Encryption, programs inside the document, and film or sound.
Anything fetched from outside the file.
A typeface named but not carried, whether it is one of the fourteen every reader owns or one of your own.
An annotation that has no drawing of how it is to appear.

Where this part stands here today

Written today, and the part every electronic invoice produced here is written under. A request asks for it by name; the engine writes level A of its own accord when the document carries a structure hierarchy.

What every archival document here carries

This holds whichever part is claimed. Nothing on it has to be asked for a second time.

Every letter the document draws travels inside it, cut down to the shapes the document actually uses, so the page reads the same on a machine where that typeface was never installed.
A colour statement is written whether or not one was asked for, so the greys and the reds mean the same thing on the next screen as on this one.
The block that says what the file is — its title, the moment it was written, the part it claims and the year of that edition — is written into the document, and the claim is stated where the part being claimed expects to find it.
A report can be asked for before anything is written: it comes back with every finding at once, each naming the published text it answers to and the font, layer, page, link, field or cell it is about.
A document that claimed a part and breaks one of its rules is refused rather than written wrong, and the refusal names the rule.
Everything the document leads to is inside the document. Every part of the standard allows an instruction opening a file that sits beside it instead; the team declines that instruction under an archival claim, so the document stays readable on its own the day the folder around it is gone. It is the same line the engine already draws on forms.
Every archival document published on this site is put through two checkers that had no hand in writing it, veraPDF and Ghostscript, and it is their verdict that is shown — not ours.

The requirements, family by family

The eight families below are the same set of requirements for every part, and each of their pages says what comes out on its own and what only you can supply.

Which part of the archiving standard to claim The archiving standard is a family, not a single rule. Four of its members can be claimed here, and choosing between them takes two questions.
Every letter the document draws travels inside it A document that carries its own lettering opens the same way on a machine that has never had that typeface installed. That is the first thing the standard asks for, and the one that changes the file most.
Colours that mean the same thing on every device Numbers on a page are not colours until something says what device they were chosen for. An archival document carries that statement inside it.
A page that comes out the same for everyone The promise of an archival document is that two readers, twenty years apart, see the same page. That holds only where nothing on the page is left for the reader to settle.
The document says what its marks stand for A page of ink says where the marks went and nothing about what they mean. The level that is read aloud, and the accessibility standard, ask the document to say it.
A document that carries other files inside it Part 3 exists because a document sometimes has to carry another file inside it, and that file has to come back out years later, whole.
What the file says about itself A document kept for thirty years is found again by what it says about itself, not by the folder it happened to be in.
Checking the claim, before and after Claiming a standard is one line inside the file. What makes the claim worth anything is that something independent agrees with it.

The published texts behind it

Each published text, and what it asks for
Published text	What it asks for
`ISO 19005-3 6.1.2`	The version the file header names is the one the part rests on.
`ISO 19005-3 6.2.11.4.1`	Every typeface the document draws with is carried inside the file.
`ISO 19005-3 6.2.8`	What a reader is to do between an image's samples is stated, not left to it.
`ISO 19005-3 6.3.2`	What an annotation's flags may and may not say.
`ISO 19005-3 6.5.1`	Which kinds of action a document of this part may hold.
`ISO 19005-3 6.7.3.3`	A document at level A carries a structure hierarchy.
`ISO 19005-3 6.9`	What a set of layers a reader can switch off has to declare.

The other editions of the standard

PDF/A-1 PDF/A-2 PDF/A-4

Back to what the archiving standard asks for

See the prices See the examples

Glossary

byte: The unit a file's weight is counted in, the way a parcel is counted in grams. A thousand of them make a kilobyte, and a million make a megabyte — the size of one photograph taken with a telephone. A one-page PDF here weighs between six thousand and a hundred and twenty thousand, so a hundred of them fit in the space of five photographs.
standard: A rule argued out in committee, published under a number anybody may buy and read, and identical for every firm claiming it. A claim to follow one can therefore be checked against the text. A way of working that merely spread because it worked is a habit of the trade: useful, widespread, and answerable to no text at all.
veraPDF: A free program that checks whether a file really is what it claims to be, against the published rules. It is the tool archives use, so it is the one used here: a document is not called compliant because we say so, but because veraPDF passed it.
Factur-X: An invoice that is a page for a person and a data file for a machine, in one document. The page looks like any invoice; inside, the same amounts are attached in a form accounting software reads without anybody retyping them. French law requires this exchange between companies.
attachment: A whole file tucked inside a PDF and carried along with it, the way a paper clip holds a sheet to a folder. Opening the document is enough to get it back out. An electronic invoice works this way: the page a person reads carries the same amounts as a file accounting software reads.
font: The drawing of every letter, digit and mark a document writes, held in a file of its own. A PDF carries the fonts it is set in inside itself, which is why it opens looking the same on a machine that has never had them. Leave them out and a reader puts another font in their place, and the layout moves.
structure tree: The list, held inside a tagged PDF, that says what each part of the page is and what it belongs to: this block is a section, this line its heading, this run of cells one row of that table. It is the shape of the document, kept apart from the ink.
request: One call to the service: you send what the document should say, and get the document back. Your bill counts these calls, one for each document. The number of pages inside a document is never counted.
layer: A named group of things drawn on a page that a reader can switch on and off, like tracing paper laid over a plan. A drawing can hold its measurements on one and its notes on another, and a layer can be set to show on screen and stay off the printer.
annotation: Something laid over a page rather than drawn into it: a clickable rectangle, a box to type in, a button. It sits at its own place on the sheet and is what makes a PDF answer to a click at all.

Every word the site explains