Merge pull request #80 from adangel:reuse-pmd-builds

Cache and reuse pmd builds #80

Author: adangel

History.md

@@ -27,6 +27,7 @@ puts summary
 *   [#75](https://github.com/pmd/pmd-regression-tester/pull/75): Add new option "--error-recovery"
 *   [#76](https://github.com/pmd/pmd-regression-tester/pull/76): Speedup XML parsing
 *   [#79](https://github.com/pmd/pmd-regression-tester/pull/79): Add new configuration option "--baseline-download-url"
+*   [#80](https://github.com/pmd/pmd-regression-tester/pull/80): Cache and reuse pmd builds
 
 ## External Contributions
 

README.rdoc

@@ -65,19 +65,21 @@ The tool creates the following folders:
   │   ├── PROJECT_NAME_1
   │   ├── ......
   │   └── PROJECT_NAME_n
-  └── reports
-      ├── BASE_BRANCH_NAME      <- the base baseline is placed here
-      ├── PATCH_BRANCH_NAME      <- the patch baseline is placed here
-      └── diff
-          ├── index.xml <- the summary report of diff reports
-          ├── base_config.xml  <- the resources of the summary report
-          ├── patch_config.xml  <- the resources fo the summary report
-          ├── css  <- css reources are placed here
-          ├── PROJECT_NAME_1
-              └── index.xml   <- the diff report of PROJECT_1
-          ├── .......
-          └── PROJECT_NAME_n
-              └── index.xml   <- the diff report of PROJECT_N
+  ├── reports
+  │   ├── BASE_BRANCH_NAME      <- the base baseline is placed here
+  │   ├── PATCH_BRANCH_NAME      <- the patch baseline is placed here
+  │   └── diff
+  │       ├── index.xml <- the summary report of diff reports
+  │       ├── base_config.xml  <- the resources of the summary report
+  │       ├── patch_config.xml  <- the resources fo the summary report
+  │       ├── css  <- css reources are placed here
+  │       ├── PROJECT_NAME_1
+  │       │   └── index.xml   <- the diff report of PROJECT_1
+  │       ├── .......
+  │       └── PROJECT_NAME_n
+  │           └── index.xml   <- the diff report of PROJECT_N
+  ├── pmd-bin-<version>-<branch_name>-<sha1>  <- cached pmd builds that are reused
+  └── ....
 
 ==== The baseline format
   branch_name

Rakefile

@@ -19,7 +19,7 @@ hoe = Hoe.spec 'pmdtester' do
 
   self.clean_globs = %w[target/reports/**/* target/test/**/* target/dynamic-config.xml Gemfile.lock]
   self.extra_deps += [['nokogiri', '~> 1.8'], ['slop', '~> 4.6'], ['differ', '~> 0.1'],
-                      ['rufus-scheduler', '~> 3.5']]
+                      ['rufus-scheduler', '~> 3.5'], ['logger-colors', '~> 1.0']]
   self.extra_dev_deps += [
     ['hoe-bundler',   '~> 1.5'],
     ['hoe-git',       '~> 1.6'],

lib/pmdtester.rb

@@ -1,6 +1,7 @@
 # frozen_string_literal: true
 
 require 'logger'
+require 'logger/colors'
 
 require_relative 'pmdtester/cmd'
 require_relative 'pmdtester/pmd_branch_detail'

lib/pmdtester/builders/pmd_report_builder.rb

@@ -24,46 +24,52 @@ def initialize(projects, options, branch_config, branch_name)
     def get_pmd_binary_file
       logger.info "#{@pmd_branch_name}: Start packaging PMD"
       Dir.chdir(@local_git_repo) do
-        current_head_sha = Cmd.execute('git rev-parse HEAD')
-        current_branch_sha = Cmd.execute("git rev-parse #{@pmd_branch_name}")
+        build_branch_sha = Cmd.execute("git rev-parse #{@pmd_branch_name}^{commit}")
 
-        @pmd_version = determine_pmd_version
+        checkout_build_branch # needs a clean working tree, otherwise fails
 
-        # in case we are already on the correct branch
-        # and a binary zip already exists...
-        if current_head_sha == current_branch_sha &&
-           File.exist?("pmd-dist/target/pmd-bin-#{@pmd_version}.zip")
-          logger.warn "#{@pmd_branch_name}: Skipping packaging - zip for " \
-                      "#{@pmd_version} already exists"
+        raise "Wrong branch #{get_last_commit_sha}" unless build_branch_sha == get_last_commit_sha
+
+        logger.debug "#{@pmd_branch_name}: PMD Version is #{@pmd_version} " \
+                     "(sha=#{build_branch_sha})"
+        distro_path = saved_distro_path(build_branch_sha)
+        if File.directory?(distro_path)
+          logger.info "#{@pmd_branch_name}: Skipping packaging - saved version exists " \
+                      " in #{distro_path}"
         else
-          build_pmd
+          build_pmd(into_dir: distro_path)
         end
 
-        @pmd_branch_details.branch_last_sha = get_last_commit_sha
+        # we're still on the build branch
+        @pmd_branch_details.branch_last_sha = build_branch_sha
         @pmd_branch_details.branch_last_message = get_last_commit_message
-
-        logger.info "#{@pmd_branch_name}: Extracting the zip"
-        unzip_cmd = "unzip -qo pmd-dist/target/pmd-bin-#{@pmd_version}.zip -d #{@pwd}/target"
-        Cmd.execute(unzip_cmd)
       end
       logger.info "#{@pmd_branch_name}: Packaging PMD completed"
     end
 
-    def build_pmd
-      logger.info "#{@pmd_branch_name}: Checking out the branch"
-      checkout_cmd = "git checkout #{@pmd_branch_name}"
-      Cmd.execute(checkout_cmd)
-
-      # determine the version again - it might be different in the other branch
-      @pmd_version = determine_pmd_version
+    # builds pmd on currently checked out branch
+    def build_pmd(into_dir:)
+      # in CI there might have been a build performed already. In that case
+      # we reuse the build result, otherwise we build PMD freshly
+      pmd_dist_target = "pmd-dist/target/pmd-bin-#{@pmd_version}.zip"
+      if File.exist?("#{@local_git_repo}/#{pmd_dist_target}")
+        # that's a warning, because we don't know, whether this build really
+        # belongs to the current branch or whether it's from a previous branch.
+        # In CI, that's not a problem, because the workspace is always fresh.
+        logger.warn "#{@pmd_branch_name}: Reusing already existing #{pmd_dist_target}"
+      else
+        logger.info "#{@pmd_branch_name}: Building PMD #{@pmd_version}..."
+        package_cmd = './mvnw clean package' \
+                      ' -Dmaven.test.skip=true' \
+                      ' -Dmaven.javadoc.skip=true' \
+                      ' -Dmaven.source.skip=true' \
+                      ' -Dcheckstyle.skip=true'
+        Cmd.execute(package_cmd)
+      end
 
-      logger.info "#{@pmd_branch_name}: Building PMD #{@pmd_version}..."
-      package_cmd = './mvnw clean package' \
-                    ' -Dmaven.test.skip=true' \
-                    ' -Dmaven.javadoc.skip=true' \
-                    ' -Dmaven.source.skip=true' \
-                    ' -Dcheckstyle.skip=true'
-      Cmd.execute(package_cmd)
+      logger.info "#{@pmd_branch_name}: Extracting the zip"
+      Cmd.execute("unzip -qo #{pmd_dist_target} -d pmd-dist/target/exploded")
+      Cmd.execute("mv pmd-dist/target/exploded/pmd-bin-#{@pmd_version} #{into_dir}")
     end
 
     def determine_pmd_version
@@ -73,7 +79,7 @@ def determine_pmd_version
     end
 
     def get_last_commit_sha
-      get_last_commit_sha_cmd = 'git rev-parse HEAD'
+      get_last_commit_sha_cmd = 'git rev-parse HEAD^{commit}'
       Cmd.execute(get_last_commit_sha_cmd)
     end
 
@@ -84,7 +90,7 @@ def get_last_commit_message
 
     def generate_pmd_report(project)
       error_recovery_options = @error_recovery ? 'PMD_JAVA_OPTS="-Dpmd.error_recovery -ea" ' : ''
-      run_path = "target/pmd-bin-#{@pmd_version}/bin/run.sh"
+      run_path = "#{saved_distro_path(@pmd_branch_details.branch_last_sha)}/bin/run.sh"
       pmd_cmd = "#{error_recovery_options}" \
                 "#{run_path} pmd -d #{project.local_source_path} -f xml " \
                 "-R #{project.get_config_path(@pmd_branch_name)} " \
@@ -149,5 +155,39 @@ def build
       get_pmd_binary_file
       generate_pmd_reports
     end
+
+    private
+
+    def checkout_build_branch
+      logger.info "#{@pmd_branch_name}: Checking out the branch"
+      # note that this would fail if the tree is dirty
+      Cmd.execute("git checkout #{@pmd_branch_name}")
+
+      # determine the version
+      @pmd_version = determine_pmd_version
+
+      return unless wd_has_dirty_git_changes
+
+      # working dir is dirty....
+      # we don't allow this because we need the SHA to address the zip file
+      logger.error "#{@pmd_branch_name}: Won\'t build without a clean working tree, " \
+                   'commit your changes'
+    end
+
+    def work_dir
+      "#{@pwd}/target"
+    end
+
+    # path to the unzipped distribution
+    # e.g. <cwd>/pmd-bin-<version>-<branch>-<sha>
+    def saved_distro_path(build_sha)
+      "#{work_dir}/pmd-bin-#{@pmd_version}" \
+      "-#{PmdBranchDetail.branch_filename(@pmd_branch_name)}" \
+      "-#{build_sha}"
+    end
+
+    def wd_has_dirty_git_changes
+      !Cmd.execute('git status --porcelain').empty?
+    end
   end
 end

pmdtester.gemspec

@@ -11,7 +11,7 @@ Gem::Specification.new do |s|
   s.metadata = { "bug_tracker_uri" => "https://github.com/pmd/pmd-regression-tester/issues", "homepage_uri" => "https://pmd.github.io", "source_code_uri" => "https://github.com/pmd/pmd-regression-tester" } if s.respond_to? :metadata=
   s.require_paths = ["lib".freeze]
   s.authors = ["Andreas Dangel".freeze, "Binguo Bao".freeze]
-  s.date = "2020-11-25"
+  s.date = "2020-11-28"
   s.description = "A regression testing tool ensure that new problems and unexpected behaviors will not be introduced to PMD project after fixing an issue , and new rules can work as expected.".freeze
   s.email = ["[email protected]".freeze, "[email protected]".freeze]
   s.executables = ["pmdtester".freeze]
@@ -33,6 +33,7 @@ Gem::Specification.new do |s|
     s.add_runtime_dependency(%q<slop>.freeze, ["~> 4.6"])
     s.add_runtime_dependency(%q<differ>.freeze, ["~> 0.1"])
     s.add_runtime_dependency(%q<rufus-scheduler>.freeze, ["~> 3.5"])
+    s.add_runtime_dependency(%q<logger-colors>.freeze, ["~> 1.0"])
     s.add_development_dependency(%q<hoe-bundler>.freeze, ["~> 1.5"])
     s.add_development_dependency(%q<hoe-git>.freeze, ["~> 1.6"])
     s.add_development_dependency(%q<minitest>.freeze, ["~> 5.10"])
@@ -46,6 +47,7 @@ Gem::Specification.new do |s|
     s.add_dependency(%q<slop>.freeze, ["~> 4.6"])
     s.add_dependency(%q<differ>.freeze, ["~> 0.1"])
     s.add_dependency(%q<rufus-scheduler>.freeze, ["~> 3.5"])
+    s.add_dependency(%q<logger-colors>.freeze, ["~> 1.0"])
     s.add_dependency(%q<hoe-bundler>.freeze, ["~> 1.5"])
     s.add_dependency(%q<hoe-git>.freeze, ["~> 1.6"])
     s.add_dependency(%q<minitest>.freeze, ["~> 5.10"])

test/integration_test_runner.rb

@@ -56,10 +56,11 @@ def test_single_mode_with_html_flag_option
   end
 
   def test_online_mode
-    # This test depends on the file test_branch_2-baseline.zip being available on sourceforge:
-    # https://sourceforge.net/projects/pmd/files/pmd-regression-tester/test_branch_2-baseline.zip/download
+    # This test depends on the file test_branch_2-baseline.zip being available at:
+    # https://pmd-code.org/pmd-regression-tester/test_branch_2-baseline.zip
     base_branch = 'test_branch_2'
-    argv = "-r target/repositories/pmd -m online -b #{base_branch} -p pmd_releases/6.7.0"
+    argv = "-r target/repositories/pmd -m online -b #{base_branch} -p pmd_releases/6.7.0 " \
+        '--baseline-download-url https://pmd-code.org/pmd-regression-tester/'
 
     system("bundle exec bin/pmdtester #{argv}")
 
@@ -76,15 +77,16 @@ def test_online_mode
   end
 
   def test_online_mode_different_project_list_and_config
-    # This test depends on the file pmd_releases_6.6.0-baseline.zip being available on sourceforge:
-    # https://sourceforge.net/projects/pmd/files/pmd-regression-tester/pmd_releases_6.6.0-baseline.zip/download
+    # This test depends on the file pmd_releases_6.6.0-baseline.zip being available at:
+    # https://pmd-code.org/pmd-regression-tester/pmd_releases_6.6.0-baseline.zip
     argv = '--local-git-repo target/repositories/pmd '\
            '--mode online '\
            '--base-branch pmd_releases/6.6.0 '\
            '--patch-branch pmd_releases/6.7.0 '\
            '--patch-config test/resources/integration_test_runner/patch-config.xml '\
            '--list-of-project test/resources/integration_test_runner/project-list.xml '\
-           '--auto-gen-config'
+           '--auto-gen-config ' \
+           '--baseline-download-url https://pmd-code.org/pmd-regression-tester/'
 
     system("bundle exec bin/pmdtester #{argv}")
 

test/resources/project-test.xml

@@ -8,7 +8,7 @@ xsi:noNamespaceSchemaLocation="projectlist_1_1_0.xsd">
     <name>checkstyle</name>
     <type>git</type>
     <connection>https://github.com/checkstyle/checkstyle</connection>
-    <tag>checkstyle-8.0</tag>
+    <tag>checkstyle-8.10</tag>
     <exclude-pattern>.*/target/test-classes/com/puppycrawl/tools/checkstyle/.*</exclude-pattern>
     <exclude-pattern>.*/target/generated-sources/.*</exclude-pattern>
 
@@ -17,7 +17,9 @@ if test -e classpath.txt; then
   exit
 fi
 
-mvn test-compile
+set -e
+
+mvn clean test-compile
 mvn dependency:build-classpath -DincludeScope=test -Dmdep.outputFile=classpath.txt
 ]]></build-command>
     <auxclasspath-command>echo -n "$(pwd)/target/classes:$(pwd)/target/test-classes:"; cat classpath.txt</auxclasspath-command>