SPECjEnterprise2010の結果一覧をTSVファイルに書き出すGroovyスクリプト

specj2010dump.groovy

@Grab(group='net.sourceforge.nekohtml', module='nekohtml', version='1.9.14')
import org.cyberneko.html.parsers.SAXParser

def parser = new XmlParser(new SAXParser())
def html = parser.parse('http://www.spec.org/jEnterprise2010/results/jEnterprise2010.html')

//new XmlNodePrinter().print(html)

def result = []
def baseUrl = 'http://www.spec.org/jEnterprise2010/results/'

html.BODY.DIV.TABLE.TBODY.TR.each {
	def r = []
	def submitter_name = it.TD.find{it.@class == 'submitter_name'}?.text()
	if(!submitter_name) return
	r << submitter_name
	r << it.TD.find{it.@class == 'system_title'}?.text()
	r << it.TD.SPAN.find{it.@class == 'disclosures'}?.text().replaceAll(/\|/, '')
	r << baseUrl + it.TD.SPAN.A.@href[0]
	def metric = it.TD.find{it.@class == 'result_metric'}?.text()
	r << metric
	r << it.TD.find{it.@class == 'system_sw_JEE_nNode'}?.text()
	def cores = it.TD.find{it.@class == 'system_sw_JEE_nCPU'}?.text()
	r << cores
	r << it.TD.find{it.@class == 'system_sw_DB_nNode'}?.text()
	r << it.TD.find{it.@class == 'system_sw_DB_nCPU'}?.text()
	def coresNum = (cores =~ /(\d+) cores,/)[0][1]
	def resultPerCore = metric.replaceAll(/,/,'').toDouble() / coresNum.toInteger()
	r << coresNum
	r << resultPerCore
	result << r
}
//println result.size()
println( ['Tested By','System Name','Date','URL','Result','JEE Nodes','JEE CPU','DB Nodes','DB CPU','Cores','jops/Core'].join('\t') )
result.each{ println it.join('\t') }