From the Rebol Console:

do http://reb4.me/r/altxml
? load-xml

For Rebol 3:

do http://reb4.me/r3/altxml

For Red:

do https://raw.githubusercontent.com/rgchris/Scripts/master/experimental/altxml.red

For Ren-C:

do https://raw.githubusercontent.com/rgchris/Scripts/master/experimental/altxml.reb

Sample Usage

To get all the links from an xhtml page:

html: load-xml/dom http://w3.org
links: html/get-by-tag <a>
foreach link links [print ["(" link/get #href ")" link/text]]

To get entry titles from an RSS feed:

rss: load-xml/dom http://www.rebol.com/article/carl-rss.xml
entries: rss/get-by-tag <item>
foreach entry entries [probe entry/get <title>]

Today’s Weather

weather: http://weather.yahooapis.com/forecastrss?p=35205
weather: load-xml/dom weather
weather: pick weather/get-by-tag <condition> 1
probe weather: context [
temp: rejoin [weather/get #temp "°F"]
sky: weather/get #text


This function parses an XML document provided by string!, url! or file! producing a representative block. Tags are represented by the tag! type, attributes by the issue! type and text prefixed by a /text refinement.

>> load-xml "<a b='c'>D</a>"
== [
<a> [
#b "c"
%.txt "D"

Tags with only text as children have a single value:

>> load-xml "<a>B</a>"
== [
<a> "B"

/dom refinement

Wraps the above document in an object! with accessor functions.  Values from the accessor functions are similarly wrapped in objects.

get-by-tag ‘tag

Returns a block of child nodes where the tag matches the ‘tag value.

body: doc/get-by-tag <body>
body/get-by-tag <p>

get-by-id ‘id

Returns the first matching child node that has an attribute id matching the the ‘id value.

header: doc/get-by-id "header"


Returns a block of child nodes.


sibling /before /after

Returns an adjacent node or none!


get ‘name

Returns the value of an immediate child node of name ‘name (attribute or child tag)

doc/get #b
doc/get <a>


Returns the textual value of a tag node.



The value of a node.



Serializes the document back to an XML string.


(only in the experimental version for now)

Select nodes and values using a path notation:

; all titles in an RSS feed
rss/path [<rss> <channel> <item> <title> ?]

; all item nodes in an XML document
rss/path [* <item>]

; first header in an HTML document
html/path [<html> * <h1> 1]

Notation includes:

<rss> - select tags at the current depth
#version - select attributes at the current depth
* <item> - select any descendant with this tag
2 - selects the second result
? - converts the selection(s) to that node's value

Not implemented:


To follow.  Will return a node’s parent node.