Positional and Named Attributes
This page breaks down the difference between positional and named attributes on an element and the rules for parsing an attribute list.
Entries in an attribute list that only consist of a value are referred to as positional attributes. The position is the 1-based index of the entry once all named attributes have been removed (so they may be interspersed).
The positional attribute may be dually assigned to an implicit attribute name if the block or macro defines a mapping for positional attributes. Here are some examples of those mappings:
icon:1 ⇒ size
image::1 ⇒ alt (text), 2 ⇒ width, 3 ⇒ height
Delimited blocks: 1 ⇒ block style and attribute shorthand
Other inline quoted text: 1 ⇒ attribute shorthand
xref:1 ⇒ text
Custom blocks and macros can also specify positional attributes
For example, the following two image macros are equivalent.
The second macro is the same as the first, but written out in longhand form.
The first positional attribute on all blocks (including sections) is special. It’s used to define the block style. It also supports a shorthand syntax for defining the ID, role, and options attributes. This shorthand syntax can also be used on formatted text, even though formatted text doesn’t technically support attributes.
The attribute shorthand is inspired by the HAML and Slim template languages as a way of saving the author some typing. Instead of having to use the longhand form of a name attribute, it’s possible to compress the assignment to a value prefixed by a special marker. The markers are mapped as follows:
Each shorthand entry is placed directly adjacent to previous one, starting immediately after the optional block style. The order of the entries does not matter, except for the style, which must come first.
Here’s an example that shows how to set an ID on a section using this shorthand syntax:
[#custom-id] == Section with Custom ID
The shorthand entry must follow the block style, if present. Here’s an example that shows how to set an ID on an appendix section using this shorthand syntax:
[appendix#custom-id] == Appendix with Custom ID
Here’s an example of a block that uses the shorthand syntax to set the ID, a role, and an option for a list.
Specifically, this syntax sets the ID to
rules, adds the role
prominent, and sets the option
[#rules.prominent%incremental] * Work hard * Play hard * Be happy
A block can have multiple roles and options, so these shorthand entries may be repeated.
Here’s an example that shows how to set several options on a table.
Specifically, this syntax sets the
[%header%footer%autowidth] |=== |Header A |Header B |Footer A |Footer B |===
This shorthand syntax also appears on formatted text.
Here’s an example that shows how to set the ID and add a role to a strong phrase.
Specifically, this syntax sets the ID to
free-world and adds the
[#free-world.goals]*free the world*
Formatted text does not support a style, so the first and only positional attribute is always the shorthand syntax.
A named attribute consists of a name and a value separated by an
= character (e.g.,
If the value contains a space, comma, or quote character, it must be enclosed in double or single quotes (e.g.,
name="value with space").
In all other cases, the surrounding quotes are optional.
If the value contains the same quote character used to enclose the value, the quote character in the value must be escaped by prefixing it with a backslash (e.g.,
value="the song \"Dark Horse\"").
If enclosing quotes are used, they are dropped from the parsed value and the preceding backslash is dropped from any escaped quotes.
The source text that’s used to define attributes for an element is referred to as an attrlist. An attrlist is always enclosed in a pair of square brackets. This applies for block attributes as well as attributes on a block or inline macro. The processor splits the attrlist into individual attribute entries, determines whether each entry is a positional or named attribute, parses the entry accordingly, and assigns the result as an attribute on the node.
The rules for what defines the boundaries of an individual attribute, and whether the attribute is positional or named, are defined below.
In these rules,
name consists of a word character (letter or numeral) followed by any number of word or
- characters (e.g.,
Attribute references are expanded before the attrlist is parsed (i.e., the attributes substitution is applied).
Parsing an attribute proceeds from the beginning of the attribute list string or after a previously identified delimiter (
The first character of an attribute list cannot be a tab or space. For subsequent attributes, any leading space or tab characters are skipped.
If a valid attribute name is found, and it is followed by an equals sign (=), then the parser recognizes this as a named attribute. The text after the equals sign (=) and up to the next comma or end of list is taken as the attribute value. Space and tab characters around the equals sign (=) and at the end of the value are ignored.
Otherwise, this is a positional attribute with a value that ends at the next delimiter or end of list. Any space or tab characters at the boundaries of the value are ignored.
To parse the attribute value:
If the first character is not a quote, the string is read until the next delimiter or end of string.
If the first character is a double quote (i.e.,
"), then the string is read until the next unescaped double quote or, if there is no closing double quote, the next delimiter. If there is a closing double quote, the enclosing double quote characters are removed and escaped double quote characters are unescaped; if not, the initial double quote is retained.
If the next character is a single quote (i.e.,
'), then the string is read until the next unescaped single quote or, if there is no closing single quote, the next delimiter. If there is a closing single quote, the enclosing single quote characters are removed and escaped single quote characters are unescaped; if not, the initial single quote is retained. If there is a closing single quote, and the first character is not an escaped single quote, substitutions are performed on the value as described in Substitutions.
Recall that attribute references are expanded before the attrlist is parsed. Therefore, it’s not necessary to force substitutions to be applied to a value if you’re only interested in applying the attributes substitution. The attributes substitution has already been applied at this point.
If the attribute name (in the case of a positional attribute) or value (in the case of a named attribute) is enclosed in single quotes (e.g.,
title='Processed by https://asciidoctor.org'), and the attribute is defined in an attrlist on a block, normal substitutions are applied to the value at assignment time.
No special processing is performed, aside from the expansion of attribute references, if the value is not enclosed in quotes or is enclosed in double quotes.
If the value contains the same quote character used to enclose the value, escape the quote character in the value by prefixing it with a backslash (e.g.,
title='A \'use case\' diagram, generated by https://plantuml.com').