🚀 go-pugleaf

I think a lexical parse to an attribute tree of att-value pairs consed
to the element would probably far easier to implement here. Accounting for
implied values, like

<element att1_with_implied_val att2="value2">

makes for an extremely complicated regular expression, probably with
miserable performance (wonder if it breaks on multi-byte characters).

Use one regex to find an element, parse to the first non-word-constituent,
non tag-close character, and then a start a new context on the next
word-constituent character, which must be the start of an attribute name.

From there until you get to tag-close, you can do a character-by-character
parse, building an attribute list of name-value pairs, which simplifies
omitting " characters and makes the attribute-attribute_value parsing
routine reusable. 

The sensible thing, it seems to me, is to parse the whole document and
build a nested tree of elements-with-attributes-values and their offsets
into the document. That could be reused for dsssl.

Regards, Clayton Weaver  cgweav@eskimo.com  (Seattle)

-- 

Regards, Clayton Weaver  cgweav@eskimo.com  (Seattle)

"Since this is a technical newsgroup and has mostly done a pretty good job