Invisible XML: the specification

The author

Steven Pemberton, CWI, Amsterdam


Invisible XML

A method for treating non-XML documents as if they were XML.

Almost everything can be treated as XML!

How it works

Extra information in the grammar controls details of the serialization:


Email type from XForms 2.0:

<xs:simpleType name="email">
  <xs:restriction base="xs:string">
    <xs:pattern value="([A-Za-z0-9!#-'\*\+\-/=\?\^_`\{-~]+)

Email in ixml

email:  user, "@", host. {two parts separated by an @ sign}
user:   atom+".".        {one or more atoms, separated by dots}
atom:   char+.           {a string of one or more 'char'}
host:   domain+".".      {a series of domains, separated by dots}
domain: word+"-".        {may contain a hyphen, but not start or end with one}
word:   letgit+.         {otherwise consists of letters and digits}
-letgit: ["A"-"Z"; "a"-"z"; "0"-"9"].
-char:   letgit; ["!#$%&'*+-/=?^_`{|}~"].
                         {A letter, digit, or punctuation.}

Parsing the following email address with this






The rule for letgit

-letgit: ["A"-"Z"; "a"-"z"; "0"-"9"].

has a dash before it, otherwise instead of


you would have had:




Since the word part of a domain has no semantic meaning, exclude it from the serialisation:

word: letgit+. ⇒ -word: letgit+.



Change the rule for atom and domain:

atom: char+.      ⇒ -atom: char+.
domain: word+"-". ⇒ -domain: word+"-".



Change the rules for user and host:

user: atom+".".   ⇒ @user: atom+".".
host: domain+".". ⇒ @host: domain+".".



To get rid of the left-over "@", we can change the rule for email:

  email: atoms, "@", host.
⇒ email: atoms, -"@", host.



The grammar

Format known as 1VWG:

Expressing this in ixml:

ixml: rule+.
rule: name, ":", alternatives, ".".
alternatives: alternative+";".
alternative: term*",".


A term is a factor, an optional factor, or a factor repeated zero or more, or one or more times:

term: factor;
option:    factor, "?".
repeat0:   factor, "*", separator?.
repeat1:   factor, "+", separator?.
separator: factor.

A factor is a terminal, a nonterminal, or a bracketed series of alternatives:

factor: terminal; 
        "(", alternatives, ")".


Terminals are the elements that actually match characters in the input.

Unicode characters are used.

There are two forms for terminals: literal strings, and character sets.


Strings may be delimited by either double quotes or single: ":" and ':' are equivalent.

If you want to include the delimiting quote in a string, it should be doubled: "don't" and 'don''t' are equivalent.

Encoded characters

For characters with no explicit visible representation or with an ambiguous representation (or not representable with XML).

For instance #a0 represents a non-breaking space.

Encoded characters do not appear within strings, but are free-standing.

This doesn't restrict expressiveness, since a rule like

end: "the", #a0, "end".

represents the seven characters with a non-breaking space in the middle.

Character sets

Character sets match a character from a set of characters. E.g.

["A"-"Z"; "a"-"z"; "0"-"9"]

Elements of a character set can be

Character classes

Unicode defines 30 character classes, such as lower case letter and upper case letter, encoded with two-letter abbreviations, such as Ll and Lu.

These abbreviations save the work of having to define those classes yourself.

And encourage people to use them, hopefully making formats more international, since you no longer need to know which characters are letters.


These allow you to say "any character but these". For instance you can say:

comment: "{", cchar*, "}".
cchar: ~["}"].


A nonterminal is just a name, referring to the rule of that name.

What's in a name?

Is this allowed?


What's in a name?

Is this allowed?


Answer: it depends whether that character is #B5 or #3BC.

This is allowed:


This is not:


The XML definition

NameStartChar   ::=
                ":" |
                [A-Z] | 
                "_" | 
                [a-z] | 
                [#xC0-#xD6] | 
                [#xD8-#xF6] | 
                [#xF8-#x2FF] | 
                [#x370-#x37D] | 
                [#x37F-#x1FFF] | 
                [#x200C-#x200D] | 
                [#x2070-#x218F] | 
                [#x2C00-#x2FEF] | 
                [#x3001-#xD7FF] | 
                [#xF900-#xFDCF] | 
                [#xFDF0-#xFFFD] | 
NameChar   ::=   
                NameStartChar | 
                "-" | 
                "." | 
               [0-9] | 
               #xB7 | 
               [#x0300-#x036F] | 


Unicode has 30 character classes:

Name Description Number Examples
Cc Control 65 Ack, Bell, Backspace, Tab, LF, CR, etc
Cf Format 151 Soft hyphen, Arabic Number Sign, Zero-width space, left-to-right mark, invisible times, etc.
Co Private Use #E000-#F8FF
Cs Surrogate #D800-#DFFF
Ll Lowercase Letter 2,063 a, µ, ß, à, æ, ð, ñ, π, Latin, Greek, Coptic, Cyrillic, Armenian, Georgian, Cherokee, Glagolitic, many more
Lm Modifier Letter 250 letter or symbol typically written next to another letter that it modifies in some way.
Lo Other Letter 121,047 ª, º, ƻ, dental click, glottal stop, etc., and letters from languages that don't have cased letters, such as Hebrew, Arabic, Syriac, ...
Lt Titlecase Letter 31 Mostly ligatures that have to be treated specially when starting a word.
Lu Uppercase Letter 1,702 A, Á, etc
Mc Spacing Mark 401 Spacing combining marks, Devengari, Bengali, etc.
Me Enclosing Mark 13 Combining enclosing characters such as "Enclosing circle"
Mn Nonspacing Mark 1763 Combining marks, such as combining grave accent.
Nd Decimal Number 590 0-9, in many languages, mathematical variants,
Nl Letter Number 236 Ⅰ, Ⅱ, Ⅲ, Ⅳ,...
No Other Number 676 subscripts, superscripts, fractions, circled and bracketed numbers, many languages
Pc Connector Punctuation 10 _, ‿, ⁀, ...
Pd Dash Punctuation 24 -, –, —, ...
Pe Close Punctuation 73 ), ], }, ...
Pf Final Punctuation 10 », ’, ”, ...
Pi Initial Punctuation 12 «, ‘, “, ...
Po Other Punctuation 566 !@#"%&'*,./:;?¶ ...
Ps Open Punctuation 75 (, [, {, ...
Sc Currency Symbol 54 $, £, €, ¢, ¥, ¤, ...
Sk Modifier Symbol 121 ^, ´, `, ¨, ˚, ...
Sm Math Symbol 948 +, <, =, >, |, ~, ±, ×, ÷, ...
So Other Symbol 5855 ©, ®, °, various arrows, much more.
Zl Line Separator 1 (Not cr, lf)
Zp Paragraph Separator 1
Zs Space Separator 17 space, nbsp, en quad, em quad, thin space, etc. (Not tab, cr, lf etc.)

ixml Names

Names in ixml are defined using Unicode character classes, while keeping as close as possible to the spirit of what is allowed in XML:

name: namestart, namefollower*.
namestart: ["_"; Ll; Lu; Lm; Lt; Lo].
namefollower: namestart; ["-.·‿⁀"; Nd; Mn].

Consequently there are small differences in what a name is.

For instance, Unicode classes the characters ª and º as letters, and does class both mu characters as letters, while XML doesn't

It is the responsibility of the ixml author to ensure that names that are serialised adhere to the XML rules.

Spaces and comments

Allowed after any token, and before the very first token. All rules indicate this. For instance:

ixml: S, rule+.
rule: name, S, ":", S, alternatives, ".", S.
alternatives: alternative+(";", S).

Spaces and comments

S defines what a space is:

S: (whitespace; comment)*.
whitespace: [Zs; #9 {tab}; #a {lf}; #d {cr}].
comment: "{", (cchar; comment)*, "}".
cchar: ~["{}"].

The Zs character class is all characters classified in Unicode as a space character.

(Tab, line feed, and carriage return are classified as control characters in Unicode).

Wherever whitespace is permitted in ixml, so is a comment.

Comments may be nested so that you can comment out sections of a grammar.


It is not specified which parse algorithm should be used.

It must however:


By default, a parse tree is serialised as XML elements:

The parse tree is traversed in document order (depth first, left to right)

Default Serialisation

For instance, for this small grammar for simple expressions:

expr: operand+operator.
operand: id; number.
id: letter+.
number: digit+.
letter: ["a"-"z"].
digit: ["0"-"9"].
operator: ["+-×÷"].

parsing the following string:


would produce



To control serialisation, marks are added to grammars.

There are three options for serialising a nonterminal:

  1. "^": as an element (the default, and therefore seldom needed);
  2. "@": as an attribute: all (serialised) terminal descendents of the node become the value of the attribute;
  3. "-": partial: only serialising the children, the same as option 1, but without the surrounding tags.

For serialising a terminal the only option is between serialising it ("^", the default) and not ("-").

Attribute lifting

The only unusual case for serialisation is for attribute children of a partially serialised node:

-number: @value.

There is no element for the value attribute, so it is moved up to the nearest parent element.


expr: operand+operator.
operand: id; number.
-id: @name.
name: letter+.
-number: @value.
value: digit+.
-letter: ["a"-"z"].
-digit: ["0"-"9"].
operator: ["+-×÷"].




   <operand name='pi'/>
   <operand value='10'/>


A grammar may be ambiguous, so that a given input may have more than one possible parse.

In that case:

ixml in ixml

The ixml grammar is itself an application of ixml.

That means that the grammar can be parsed with itself, and then serialised to XML.

This has consequences for the design of the grammar: where to use attributes.

The decision taken was to put all semantic terminals (such as names) in attributes, and otherwise to use elements.


So as an example, the serialisation for the rule for rule,

rule: (mark, S)?, name, S, ["=:"], S, -alts, ".", S.


   <rule name='rule'>:
                  <nonterminal name='mark'/>,
                  <nonterminal name='S'/>
         <nonterminal name='name'/>,
         <nonterminal name='S'/>,
            <literal dstring='=:'/>]</inclusion>,
         <nonterminal name='S'/>,
         <nonterminal mark='-' name='alts'/>,
         <literal dstring='.'/>,
         <nonterminal name='S'/>

Although all terminal symbols are preserved in the serialisation, the only ones of import are in attribute values.


The process


The process


The process


It is the XML serialisation of the grammar that is used as input to the parser.

This means that a different ixml grammar may be used, as long as the serialisation is the same.

So if the grammar looks like this:

<expr> ::= <id> | <number> | <expr> <op> <expr>

that is fine, as long as it produces the same serialisation structure.


A pilot implementation of an ixml processor has been created (and used for the examples here).

The next step is to turn this into a full-strength implementation.

Future work

If you look at an ixml grammar in the right way, you can also see it as a type of schema for an XML format.

Future work will look at the possibilities of using ixml to define XML formats, and methods for round-tripping.


The bind element in XForms 1.1 in XSL:

<element name="bind">
      <sequence minOccurs="0" maxOccurs="unbounded">
         <element ref="xforms:bind"/>
      <attributeGroup ref="xforms:Common.Attributes"/>
      <attribute name="nodeset"    type="xforms:XPathExpression" use="optional"/>
      <attribute name="calculate"  type="xforms:XPathExpression" use="optional"/>
      <attribute name="type"       type="QName"                  use="optional"/>
      <attribute name="required"   type="xforms:XPathExpression" use="optional"/>
      <attribute name="constraint" type="xforms:XPathExpression" use="optional"/>
      <attribute name="relevant"   type="xforms:XPathExpression" use="optional"/>
      <attribute name="readonly"   type="xforms:XPathExpression" use="optional"/>
      <attribute name="p3ptype"    type="xsd:string"             use="optional"/>


You could express this in ixml as follows:

bind: -Common, @nodeset?, -MIP*, bind*.
  MIP: @calculate; @type; @required; @constraint; 
       @relevant; @readonly; @p3ptype.
     nodeset: xpath.
   calculate: xpath.
        type: QName.
  constraint: xpath.
    relevant: xpath.
    readonly: xpath.
     p3ptype: string.

The main hurdle is that a rule name must be unique in a grammar, and in XML attributes and elements with the same name may have different content models.

For instance, there is also a bind attribute on other elements in XForms.


The input element in XForms:

<element name="input">
         <element ref="xforms:label"/>
         <group ref="xforms:UI.Common" minOccurs="0" maxOccurs="unbounded"/>
      <attributeGroup ref="xforms:Common.Attributes"/>
      <attributeGroup ref="xforms:Single.Node.Binding.Attributes"/>
      <attribute      name="inputmode"   type="xsd:string" use="optional"/>
      <attributeGroup ref="xforms:UI.Common.Attrs"/>
      <attribute      name="incremental" type="xsd:boolean" use="optional" default="false"/>

<attributeGroup name="Single.Node.Binding.Attributes">
   <attribute name="model" type="xsd:IDREF" use="optional"/>
   <attribute name="ref"   type="xforms:XPathExpression" use="optional"/>
   <attribute name="bind"  type="xsd:IDREF" use="optional"/>


which becomes:

  input: -Common, -UICommonAtts, -Binding?,
         @inputmode?, @incremental?, label, UICommon*.
Binding: (@model?, @ref; @ref, @model); @bind.
  model: IDREF.
   bind: IDREF.
    ref: xpath.

Compact formats

You could even design 'compact' versions of XML formats, eg for XForms:

bind //@open type boolean
input age "How old are you?"

by altering the rules above to

     bind: -"bind" -Common, @ref?, -MIP*, bind*.
      MIP: @nodeset; @calculate; @type; @required; @constraint; @relevant; @readonly; @p3ptype.
  nodeset: xpath.
calculate: -"calculate", xpath.
     type: -"type", QName.

etc., and

input: -"input", -Common, -UICommonAtts, -Binding?, @inputmode?, @incremental?, label, UICommon*.



A host of new non-XML documents are opened to the XML process pipeline by ixml.

In some senses it also mitigates the loss of SGML minimisation in

By defining ixml in ixml, it becomes the first large application of ixml.

Future work will be to allow designers to create formats in a compact version and an equivalent XML version in parallel.

Please comment on the specification, which you can find on my homepage.


Watch "The 100 year Web" from my home page.