csstoxpath

CSS to XPath

Converts most CSS 2.1/3 selectors (see exclusions below) to equivalent XPath 1.0 expression.

See test.js for examples.

Custom pseudos

To take advantage of the different capabilities of XPath several custom pseudo selectors have been implemented:

:text

All text matching pseudo classes normalize whitespace and ignore tags. E.g. " my <i>string</i> " is treated as "my string".

• :text("foo") Case-insensitive matching of element text
• :text-case("foo") Case-sensitive :text-case
• :text-contains("foo") Case-insensitive substring matching of element text
• :text-contains-case("foo") Case-sensitive :text-contains
• :text-start("foo") Case-insensitive matching of element starting text
• :text-start-case("foo") Case-sensitive :text-start
• :text-end("foo") Case-insensitive matching of element ending text
• :text-end-case("foo") Case-sensitive :text-end

:comment

• :comment Select comment nodes
• :comment(n) Select comment nodes at child position n

Note: Can be combined with :text to match based on comment text content. E.g. p > :comment:text("foo")

:childless

• :childless As :empty but ignoring whitespace

Unsupported psuedos

The following may be supported at a later date:

• :nth-last-child

Part-dynamic pseudos are excluded as they can only be partially supported by attribute matching:

• :checked
• :disabled
• :enabled
• :required
• :lang

The following cannot be implemented in XPath 1.0:

• :*-of-type
• States: :hover, :focus, :active, :visited, :target etc.
• Elements: ::before, ::after, ::first-letter etc.