replace XmlSlurper tag with arbitrary XML
- by Misha Koshelev
Dear All:
I am trying to replace specific XmlSlurper tags with arbitrary XML strings. The best way I have managed to come up with to do this is:
#!/usr/bin/env groovy
import groovy.xml.StreamingMarkupBuilder
def page=new XmlSlurper(new org.cyberneko.html.parsers.SAXParser()).parseText("""
<html>
<head></head>
<body>
<one attr1='val1'>asdf</one>
<two />
<replacemewithxml />
</body>
</html>
""".trim())
import groovy.xml.XmlUtil
def closure
closure={ bind,node->
if (node.name()=="REPLACEMEWITHXML") {
bind.mkp.yieldUnescaped "<replacementxml>sometext</replacementxml>"
} else {
bind."${node.name()}"(node.attributes()) {
mkp.yield node.text()
node.children().each { child->
closure(bind,child)
}
}
}
}
println XmlUtil.serialize(
new StreamingMarkupBuilder().bind { bind->
closure(bind,page)
}
)
However, the only problem is the text() element seems to capture all child text nodes, and thus I get:
<?xml version="1.0" encoding="UTF-8"?>
<HTML>asdf<HEAD/>
<BODY>asdf<ONE attr1="val1">asdf</ONE>
<TWO/>
<replacementxml>sometext</replacementxml>
</BODY>
</HTML>
Any ideas/help much appreciated.
Thank you!
Misha
p.s. Also, out of curiosity, if I change the above to the "Groovier" notation as follows, the groovy compiler thinks I am trying to access the ${node.name()} member of my test class. Is there a way to specify this is not the case while still not passing the actual builder object? Thank you! :)
def closure
closure={ node->
if (node.name()=="REPLACEMEWITHXML") {
mkp.yieldUnescaped "<replacementxml>sometext</replacementxml>"
} else {
"${node.name()}"(node.attributes()) {
mkp.yield node.text()
node.children().each { child->
closure(child)
}
}
}
}
println XmlUtil.serialize(
new StreamingMarkupBuilder().bind {
closure(page)
}
)