SWASH (Draft)

A specification for a Simple Web Address ScHeme.

Summary

This is a document describing SWASH, a specification for a simple address notation for web applications and internet interfaces.

The goals of this proposal is to define a minimalist notation for web application resource address; an alternative to URI. The notation maintains general semantics which can then be used to build higher-order semantics. swash serves as a foundational grammar of simple address schemes for networking applications alike.

Conventions

SWASH (all caps) refers to the specification. A swash (plural swashes) is a encoding scheme defined within the specification.

Grammar

The notation for a swash is a chain of address elements delimited by the / (forward-slash) character:

<swash> = ('/' <address-element>)+

<address-element> = <level> | <parameter>

<level> = <value>

<parameter> = <parameter-key> '=' <value>

<parameter-key> = (<valid-char>)*

<value> = (<valid-char>)*

<valid-char> = 'a' ... 'z' | 'A' .. 'Z' | '0' ... '9' | '-' | '.' | '_' | '~' | '?' | '#' | '[' | ']' | '@' | '!' | '$' | '&' | "'" | '(' | ')' | '*' | '+' | ',' | ';' | ':' | '`'

As it stands, <valid-char> is open for comments.

Semantics

A swash is a series of address-elements prefixed with a forward-slash character. Each address element is assigned a zero-based index called the element-index. An address-element is either a parameter or a level.

Every level is assigned a zero-based index called the level-index.

Every parameter is assigned a zero-based index called the parameter-index.

In addition, a table called the parameter-table maps parameter keys to one or more values. Each row contains a unique parameter-key column and a column containing an array called parameter-values. Each element in parameter-values is a value from a parameter with the same parameter-key of that of the row's parameter-key.

In conclusion, a swash is a general, unopinionated scheme for encoding web addresses. This lays the groundwork for further use-case-specific semantics to be built onto the swash specification.

Swash Semantics to URI (SS2URI)

SS2URI is a rough specification for translation of a swash to a URI.

Semantics for a URI can be specified such that:

The level's value at index 0 is the URI scheme.
The level's value at index 1 is the URI authority.
All values for levels >2 compose the URI path.
The parameter with # as its key composes the URI fragment by joining it's parameter-values.
All other parameters are formatted into the URI query.

/http/domain.tld/path/index.html/x=1/y=2/#=fragment
/http/domain.tld/path/index.html/y=2/x=1/#=fragment
/http/domain.tld/path/index.html/y=2/x=1/#=frag/#=ment

The above examples are all equivalent encodings for the following URI:

http://domain.tld/path/index.html?x=1&y=2#fragment

Note that each of the above swashes are semantically different by the SWASH specification, however SS2URI has no use for all semantics defined within SWASH. Some information is lost in translation, therefore this can be referred to as a "lossy translation".

Further, certain parts of the swash become case-insensitive, even though a swash is a case-sensitive encoding.

SS2URI implementation may include defaults for certain level values in a swash, for example:

Without URI scheme, path, query, nor fragment

//domain.tld

URI Translation

http://domain.tld/

Conclusion

The goal for SWASH is to create a general encoding syntax for web addresses that is minimal and flexible for future iterations of web technology. The design goals are to be humanized while fully capable.

This specification is currently in a working draft stage. All comments and feedback are appreciated and help to move the specification forward and towards a finalized standard.

samholmes / SWASH