Improve jenkinsfile parse perf, and fix parallel xz test result file compressing

Only read the supported and default jdks once.  Also add debug when starting the pipeline on the parameters that will be used for building and testing.

 ref: https://the-asf.slack.com/archives/CK23JSY2K/p1763714160975359

 patch by Mick Semb Wever; reviewed by Sam Tunnicliffe for CASSANDRA-21044

.jenkins/Jenkinsfile

@@ -35,7 +35,7 @@
 //  - cassandra-amd64-medium : 3 cpu, 5GB ram
 //  - cassandra-medium       : 3 cpu, 5GB ram (alias for above but for any arch)
 //  - cassandra-amd64-large  : 7 cpu, 16GB ram
-//  - cassandra-large  : 7 cpu, 16GB ram
+//  - cassandra-large        : 7 cpu, 16GB ram
 //
 // When running builds parameterised to other architectures the corresponding labels are expected.
 //  For example 'arm64' requires the labels: cassandra-arm64-small, cassandra-arm64-medium, cassandra-arm64-large.
@@ -56,6 +56,12 @@
 // `curl -X POST  -F "jenkinsfile=<.jenkins/Jenkinsfile" https://ci-cassandra.apache.org/pipeline-model-converter/validate`
 //
 
+/** CONSTANTS for both the pipeline and scripting **/
+import groovy.transform.Field
+@Field List<String> archsSupported   = ["amd64", "arm64"]
+@Field List<String> pythonsSupported = ["3.8", "3.11"]
+@Field String       pythonDefault    = "3.8"
+/** CONSTANTS end **********************************/
 
 pipeline {
   agent { label 'cassandra-small' }
@@ -70,13 +76,22 @@ pipeline {
     choice(name: 'profile', choices: pipelineProfileNames(params.profile ?: ''), description: 'Pick a pipeline profile.')
     string(name: 'profile_custom_regexp', defaultValue: params.profile_custom_regexp ?: '', description: 'Regexp for stages when using custom profile. See `testSteps` in Jenkinsfile for list of stages. Example: stress.*|jvm-dtest.*')
 
-    choice(name: 'architecture', choices: archsSupported() + "all", description: 'Pick architecture. The ARM64 is disabled by default at the moment.')
+    choice(name: 'architecture', choices: archsSupported + "all", description: 'Pick architecture. The ARM64 is disabled by default at the moment.')
     string(name: 'jdk', defaultValue: params.jdk ?: '', description: 'Restrict JDK versions. (e.g. "11", "17", etc)')
 
     string(name: 'dtest_repository', defaultValue: params.dtest_repository ?: 'https://github.com/apache/cassandra-dtest', description: 'Cassandra DTest Repository')
     string(name: 'dtest_branch', defaultValue: params.dtest_branch ?: 'trunk', description: 'DTest Branch')
   }
   stages {
+    stage('init') {
+      steps {
+        script {
+          // this helps assure folk their parameters are correct and will be used (despite the earlier output about the configured job coordinates)
+          echo "Printing parameters used for this build"
+          ["Repository: ${params.repository}", "Branch: ${params.branch}", "Profile: ${params.profile}", "Custom Profile Regexp: ${params.profile_custom_regexp}", "Architecture: ${params.architecture}", "JDK: ${params.jdk}", "DTest Repository: ${params.dtest_repository}", "DTest Branch: ${params.dtest_branch}"].each { println it }
+        }
+      }
+    }
     stage('jar') {
       // the jar stage executes only the 'jar' build step, via the build(…) function
       // the results of these (per jdk, per arch) are then stashed and used for every other build and test step
@@ -120,10 +135,7 @@ pipeline {
 //// scripting support ////
 ///////////////////////////
 
-def archsSupported() { return ["amd64", "arm64"] }
+@NonCPS
-def pythonsSupported() { return ["3.8", "3.11"] }
-def pythonDefault() { return "3.8" }
-
 def pipelineProfiles() {
   return [
     'packaging': ['artifacts', 'lint', 'debian', 'redhat'],
@@ -135,6 +147,7 @@ def pipelineProfiles() {
   ]
 }
 
+@NonCPS
 def pipelineProfileNames(putFirst) {
     set = pipelineProfiles().keySet() as List
     set = set - putFirst
@@ -142,7 +155,11 @@ def pipelineProfileNames(putFirst) {
     return set
 }
 
+@Field Map cachedTasks = null
+
 def tasks() {
+  if (null != cachedTasks) return cachedTasks
+
   // Steps config
   def buildSteps = [
     'jar': [script: 'build-jars.sh', toCopy: null],
@@ -200,23 +217,30 @@ def tasks() {
 
   def stepsMap = buildSteps + testSteps
 
+  // find the default JDK and the supported JDKs defined in the build.xml
+  def build_xml = readFile(file: 'build.xml')
+  def javaVersionDefaultMatch = (build_xml =~ /property\s*name="java\.default"\s*value="([^"]*)"/)
+  assert javaVersionDefaultMatch, "java.default property not found in build.xml"
+  def javaVersionDefault = javaVersionDefaultMatch[0][1]
+  def javaVersionsSupportedMatch = (build_xml =~ /property\s*name="java\.supported"\s*value="([^"]*)"/)
+  assert javaVersionsSupportedMatch, "java.supported property not found in build.xml"
+  def javaVersionsSupported = javaVersionsSupportedMatch[0][1].split(',') as List
+
   // define matrix axes
   def Map matrix_axes = [
-    arch: archsSupported(),
+    arch: archsSupported,
-    jdk: javaVersionsSupported(),
+    jdk: javaVersionsSupported,
-    python: pythonsSupported(),
+    python: pythonsSupported,
     cython: ['yes', 'no'],
     step: stepsMap.keySet(),
     split: (1..testSteps.values().splits.max()).toList()
   ]
 
-  def javaVersionDefault = javaVersionDefault()
-
   def List _axes = getMatrixAxes(matrix_axes).findAll { axis ->
     (isArchEnabled(axis['arch'])) && // skip disabled archs
     (isJdkEnabled(axis['jdk'])) && // skip disabled jdks
     (isStageEnabled(axis['step'])) && // skip disabled steps
-    !(axis['python'] != pythonDefault() && 'cqlsh-test' != axis['step']) && // Use only python 3.8 for all tests but cqlsh-test
+    !(axis['python'] != pythonDefault && 'cqlsh-test' != axis['step']) && // Use only python 3.8 for all tests but cqlsh-test
     !(axis['cython'] != 'no' && 'cqlsh-test' != axis['step']) && // cython only for cqlsh-test, disable for others
     !(axis['jdk'] != javaVersionDefault && ('cqlsh-test' == axis['step'] || 'simulator-dtest' == axis['step'] || axis['step'].contains('dtest-upgrade'))) && // run cqlsh-test, simulator-dtest, *dtest-upgrade only with jdk11
     // Disable splits for all but proper stages
@@ -238,7 +262,7 @@ def tasks() {
     }
   }
 
-  return tasks
+  return cachedTasks = tasks
 }
 
 @NonCPS
@@ -264,34 +288,14 @@ def getStepName(cell, command) {
 
 def getJarTasks() {
     Map jars = tasks()['jars']
-    assertJarTasks(jars)
+    assert jars.size() > 1, "Nothing to build. Check parameters: jdk ${params.jdk}, arch ${params.architecture}"
     return jars
 }
 
-def assertJarTasks(jars) {
-  if (jars.size() < 2) {
-    error("Nothing to build. Check parameters: jdk ${params.jdk} (${javaVersionsSupported()}), arch ${params.architecture} (${archsSupported()})")
-  }
-}
-
 def hasNonJarTasks() {
   return tasks()['tests'].size() > 1
 }
 
-/**
- * Return the default JDK defined by build.xml
- **/
-def javaVersionDefault() {
-  sh (returnStdout: true, script: 'grep \'property\\s*name=\"java.default\"\' build.xml | sed -ne \'s/.*value=\"\\([^\"]*\\)\".*/\\1/p\'').trim()
-}
-
-/**
- * Return the supported JDKs defined by build.xml
- **/
-def javaVersionsSupported() {
-  sh (returnStdout: true, script: 'grep \'property\\s*name=\"java.supported\"\' build.xml | sed -ne \'s/.*value=\"\\([^\"]*\\)\".*/\\1/p\'').trim().split(',')
-}
-
 /**
  * Is this a post-commit build (or a pre-commit build)
  **/
@@ -412,12 +416,15 @@ def test(command, cell) {
           dir("build") {
             sh """
                 mkdir -p test/output/${cell.step}
-                find test/output -type f -name TEST*.xml -execdir mkdir -p jdk_${cell.jdk}/${cell.arch} ';' -execdir mv {} jdk_${cell.jdk}/${cell.arch}/{} ';'
+                find test/output -type f -name "TEST*.xml" -execdir mkdir -p jdk_${cell.jdk}/${cell.arch} ';' -execdir mv {} jdk_${cell.jdk}/${cell.arch}/{} ';'
                 find test/output -name cqlshlib.xml -execdir mv cqlshlib.xml ${cell.step}/cqlshlib${cell_suffix}.xml ';'
                 find test/output -name nosetests.xml -execdir mv nosetests.xml ${cell.step}/nosetests${cell_suffix}.xml ';'
               """
             junit testResults: "test/**/TEST-*.xml,test/**/cqlshlib*.xml,test/**/nosetests*.xml", testDataPublishers: [[$class: 'StabilityTestDataPublisher']]
-            sh "find test/output -type f -name *.xml -exec sh -c 'xz -f {} &' ';' ; wait ; find test/output -type f -name *.xml.xz | wc -l"
+            sh """
+                find test/output -type f -name "*.xml" -print0 | xargs -0 -r -n1 -P"\$(nproc)" xz -f
+                echo "test result files compressed"; find test/output -type f -name "*.xml.xz" | wc -l
+              """
             archiveArtifacts artifacts: "test/logs/**,test/**/TEST-*.xml.xz,test/**/cqlshlib*.xml.xz,test/**/nosetests*.xml.xz", fingerprint: true
             copyToNightlies("${logfile}, test/logs/**", "${cell.step}/${cell.arch}/jdk${cell.jdk}/python${cell.python}/cython_${cell.cython}/" + "split_${cell.split}_${splits}".replace("/", "_"))
           }
@@ -557,8 +564,8 @@ def generateTestReports() {
       //   TODO parallelised for loop
       //   TODO results_details.tar.xz needs to include all logs for failed tests
       sh """${script_vars} (
-          find build/test/output -type f -name *.xml.xz | wc -l
+          echo "test result files to decompress"; find build/test/output -type f -name "*.xml.xz" | wc -l
-          find build/test/output -name *.xml.xz -exec sh -c 'xz -f --decompress {} &' ';'  ; wait
+          find build/test/output -type f -name "*.xml.xz" -print0 | xargs -0 -r -n1 -P"\$(nproc)" xz -f --decompress
 
           for target in \$(ls build/test/output/) ; do
             if test -d build/test/output/\${target} ; then