Adjusted task names and error logging

Author: AlexanderSchultheiss

README.md

@@ -43,28 +43,32 @@ _Extraction/extraction-results_ once the extraction is complete, or has been sto
 Start the extraction by executing the `start-extraction` script (see examples further below).
 The basic syntax is the following:
 
-- `start-extraction.(sh|bat) (linux|busybox) [commit-id/tag] [commit-id/tag]`
+- `start-extraction.(sh|bat) (linux|busybox|GIT_HTTPS_LINK) [commit-id/tag] [commit-id/tag]`
 - `(option-1|option-2)` -> You *must* provide either a value for `option-1` or `option-2`.
 - `[option]` -> You *may* provide a value
 
-The script must be provided with `busybox` or `linux` as first argument, in order to specify which SPL should be considered.
+The script must be provided with `busybox`, `linux`, or the https clone link of a public git repository as first argument, in order to specify which SPL should be considered.
 In addition, you can optionally provide either one or two more arguments specifying a commit-id or git-tag.
 
-If you specify __no__ id or tag, the entire history is considered.
+> If you specify `linux` or `busybox`, a full extraction, including the analysis of the build system, will be performed for either Linux or BusyBox. If you specify the clone link to a git repository, a partial extraction, without considering the build system, will be performed. This is due to technicalities of the build system analysis, which require a project specific setup, which we only performed for Linux and BusyBox. 
 
-If you specify __exactly one__ id or tag, the extraction will only consider the one commit that is found under the id/tag.
+> If you specify __no__ id or tag, the entire history is considered.
+
+> If you specify __exactly one__ id or tag, the extraction will only consider the one commit that is found under the id/tag.
 This can be used to quickly test whether everything is working as intended or to run the extraction for one commit only (e.g., when no evolution information is necessary).
 
-If you specify __two__ ids or tags, the extraction will consider the range of commits that lies between the first and the second
+> If you specify __two__ ids or tags, the extraction will consider the range of commits that lies between the first and the second
 commit. The commit retrieval follows the same logic as [git log](https://git-scm.com/docs/git-log), i.e., it will retrieve
 all commits that are ancestors of the second commit, but __not__ ancestors of the first commit.
 
 - Windows CMD:
   - `start-extraction.bat busybox [id/tag] [id/tag]` 
   - `start-extraction.bat linux [id/tag] [id/tag]`
+  - `start-extraction.bat https://github.com/MarlinFirmware/Marlin.git [id/tag] [id/tag]`
 - Linux terminal:
   - `./start-extraction.sh busybox [id/tag] [id/tag]`
   - `./start-extraction.sh linux [id/tag] [id/tag]`
+  - `./start-extraction.sh https://github.com/MarlinFirmware/Marlin.git [id/tag] [id/tag]`
 
 #### Runtime
 The entire history of BusyBox can be extracted in about one day.
@@ -77,21 +81,24 @@ Therefore, errors that appear in the log do not necessarily indicate a problem w
 
 #### Examples:
 ```
-Extract the ground truth for all commits of BusyBox
+# Extract the ground truth for all commits of BusyBox
 start-extraction.bat busybox
 ./start-extraction.sh busybox
 
-Extract the ground truth between two specific commits of Busybox
+# Extract the ground truth between two specific commits of Busybox
 start-extraction.bat busybox b35eef5383a4e7a6fb60fcf3833654a0bb2245e0 7de0ab21d939a5a304157f75918d0318a95261a3
 ./start-extraction.sh busybox b35eef5383a4e7a6fb60fcf3833654a0bb2245e0 7de0ab21d939a5a304157f75918d0318a95261a3
 
-Extract the ground truth for the commit under revision tag v4.1 of Linux
+# Extract the ground truth for the commit under revision tag v4.1 of Linux
 start-extraction.bat linux v4.1
 ./start-extraction.sh linux v4.1
 
-Extract the ground truth for all commits between two minor revisions of Linux
+# Extract the ground truth for all commits between two minor revisions of Linux
 start-extraction.bat linux v4.3 v4.4
 ./start-extraction.sh linux v4.3 v4.4
+
+# Extract a partial ground truth (no feature mode, no file conditions) for the entire history of Marlin
+./start-extraction.sh https://github.com/MarlinFirmware/Marlin.git
 ```
 
 ### Stopping the Ground Truth Extraction

docker-resources/extraction_busybox.properties

@@ -47,7 +47,7 @@ source_tree = ./busybox
 source_repo_url = https://git.busybox.net/busybox/
 
 # Do not change this
-analysis.class = org.variantsync.vevos.extraction.kh.FullAnalysis
+analysis.class = org.variantsync.vevos.extraction.kh.FullExtraction
 preparation.class.0 = net.ssehub.kernel_haven.busyboot.PrepareBusybox
 analysis.output.type = csv
 

docker-resources/extraction_generic.properties

@@ -17,7 +17,7 @@ extraction.timeout = 60
 ###  Analysis Parameters ###
 # How many commits should be analyzed in parallel? Warning: each task requires a considerable amount
 # of resources
-analysis.number_of_tasks = 5
+analysis.number_of_tasks = 3
 
 # Whether the file condition (aka. presence condition of source file) should be treated as 'true' or 'false' (default is 'true'),
 # in case of missing build model information (i.e., no feature model or file condition)
@@ -40,7 +40,7 @@ source_tree = TBD
 source_repo_url = TBD
 
 # Do not change this
-analysis.class = org.variantsync.vevos.extraction.kh.PartialAnalysis
+analysis.class = org.variantsync.vevos.extraction.kh.PartialExtraction
 analysis.output.type = csv
 
 ######################################

docker-resources/extraction_linux.properties

@@ -46,7 +46,7 @@ source_tree = ./linux
 source_repo_url = https://github.com/torvalds/linux.git
 
 # Do not change this
-analysis.class = org.variantsync.vevos.extraction.kh.FullAnalysis
+analysis.class = org.variantsync.vevos.extraction.kh.FullExtraction
 analysis.output.type = csv
 
 ######################################

pom.xml

@@ -130,7 +130,7 @@
         <dependency>
             <groupId>net.lingala.zip4j</groupId>
             <artifactId>zip4j</artifactId>
-            <version>2.9.0</version>
+            <version>2.10.0</version>
         </dependency>
     </dependencies>
 </project>

src/main/java/org/variantsync/vevos/extraction/AnalysisTask.java

@@ -43,14 +43,16 @@ public class AnalysisTask implements Runnable {
     private final File parentPropertiesFile;
     private final String splName;
     private final long timeout;
+    private final boolean fullExtraction;
 
-    public AnalysisTask(List<RevCommit> commits, File parentDir, File propertiesFile, String splName, long timeout) {
+    public AnalysisTask(List<RevCommit> commits, File parentDir, File propertiesFile, String splName, long timeout, boolean fullExtraction) {
         this.commits = commits;
         this.parentDir = parentDir;
         this.parentPropertiesFile = propertiesFile;
         this.splName = splName;
         this.taskNumber = existingTasksCount++;
         this.timeout = timeout;
+        this.fullExtraction = fullExtraction;
     }
 
     @Override
@@ -168,7 +170,7 @@ private Set<String> determineProcessedCommits(Path pathToTargetDir) {
         return processedCommits;
     }
 
-    private static void moveResultsToDirectory(File workDir, Path pathToTargetDir, RevCommit commit, File prepareFail) {
+    private void moveResultsToDirectory(File workDir, Path pathToTargetDir, RevCommit commit, File prepareFail) {
         String commitId = commit.getName();
         LOGGER.logStatus("Moving result to common output directory.");
         File data_collection_dir = pathToTargetDir.resolve("data").resolve(commitId).toFile();
@@ -186,8 +188,11 @@ private static void moveResultsToDirectory(File workDir, Path pathToTargetDir, R
         // Move the results of the analysis to the collected output directory according to the current commit
         LOGGER.logStatus("Moving presence conditions to common output directory.");
         boolean hasError = movePresenceConditions(outputDir, data_collection_dir);
+
         LOGGER.logStatus("Moving DIMACS feature model to common output directory.");
-        hasError = hasError | moveDimacsModel(outputDir, data_collection_dir);
+        if (fullExtraction) {
+            hasError = hasError | moveDimacsModel(outputDir, data_collection_dir);
+        }
 
         LOGGER.logStatus("Moving FILTERED file to common output directory.");
         if(moveFilterCount(outputDir, data_collection_dir)) {
@@ -201,7 +206,9 @@ private static void moveResultsToDirectory(File workDir, Path pathToTargetDir, R
 
         // Move the cache of the extractors to the collected output directory
         LOGGER.logStatus("Moving extractor cache to common output directory.");
-        hasError = hasError | moveFeatureModel(workDir, data_collection_dir);
+        if (fullExtraction) {
+            hasError = hasError | moveFeatureModel(workDir, data_collection_dir);
+        }
 
         // Move the log to the common output directory
         LOGGER.logStatus("Moving KernelHaven log to common output directory");
@@ -219,7 +226,7 @@ private static void moveResultsToDirectory(File workDir, Path pathToTargetDir, R
         } else {
             writeParents(commit, data_collection_dir);
             writeToFile(data_collection_dir, COMMIT_MESSAGE_FILE, commit.getFullMessage());
-            if (prepareFail.exists()) {
+            if (prepareFail.exists() && this.fullExtraction) {
                 LOGGER.logWarning("KernelHaven was not able to correctly load the build model, the extracted file presence conditions are incomplete!");
                 EXECUTOR.execute("echo \"" + commitId + "\" >> " + INCOMPLETE_PC_COMMIT_FILE, pathToTargetDir.toFile());
             } else {
@@ -287,11 +294,13 @@ private static boolean moveFeatureModel(File workDir, File targetDir) {
         return hasError;
     }
 
-    private static boolean moveOutputFile(File outputDir, File targetDir, String sourceName, String targetName) {
+    private static boolean moveOutputFile(File outputDir, File targetDir, String sourceName, String targetName, boolean errorExpected) {
         boolean hasError = false;
         File[] resultFiles = outputDir.listFiles((dir, name) -> name.contains(sourceName));
         if (resultFiles == null || resultFiles.length == 0) {
-            LOGGER.logError("NO RESULT FILE IN " + outputDir.getAbsolutePath());
+            if (!errorExpected) {
+                LOGGER.logError("NO RESULT FILE IN " + outputDir.getAbsolutePath());
+            }
             hasError = true;
         } else if (resultFiles.length == 1) {
             try {
@@ -313,19 +322,19 @@ private static boolean moveOutputFile(File outputDir, File targetDir, String sou
     }
 
     private static boolean movePresenceConditions(File outputDir, File targetDir) {
-        return moveOutputFile(outputDir, targetDir, "Blocks.csv", "code-variability.spl.csv");
+        return moveOutputFile(outputDir, targetDir, "Blocks.csv", "code-variability.spl.csv", false);
     }
 
     private static boolean moveDimacsModel(File outputDir, File targetDir) {
-        return moveOutputFile(outputDir, targetDir, "feature-model.dimacs", "feature-model.dimacs");
+        return moveOutputFile(outputDir, targetDir, "feature-model.dimacs", "feature-model.dimacs", false);
     }
 
     private static boolean moveFilterCount(File outputDir, File targetDir) {
-        return moveOutputFile(outputDir, targetDir, "FILTERED.txt", "FILTERED.txt");
+        return moveOutputFile(outputDir, targetDir, "FILTERED.txt", "FILTERED.txt", true);
     }
 
     private static boolean moveVariablesFile(File outputDir, File targetDir) {
-        return moveOutputFile(outputDir, targetDir, "VARIABLES.txt", "VARIABLES.txt");
+        return moveOutputFile(outputDir, targetDir, "VARIABLES.txt", "VARIABLES.txt", true);
     }
 
     private void createBlocker(File dir) {

src/main/java/org/variantsync/vevos/extraction/Extraction.java

@@ -45,6 +45,8 @@ public class Extraction {
     public static final @NonNull Setting<@Nullable Integer> EXTRACTION_TIMEOUT
             = new Setting<>("extraction.timeout", Setting.Type.INTEGER, false, "0", "" +
             "The timeout for the KernelHaven execution in seconds.");
+    public static final @NonNull Setting<@Nullable String> ANALYSIS_CLASS
+            = new Setting<>("analysis.class", Setting.Type.STRING, true, null, "Class of the pipeline that is used for the analysis");
     private static final Logger LOGGER = Logger.get();
     private static final ShellExecutor EXECUTOR = new ShellExecutor(LOGGER);
 
@@ -104,9 +106,10 @@ public static void main(String... args) throws IOException, GitAPIException {
         // Create a task for each commit subset and submit it to the thread pool
         int count = 0;
         LOGGER.logStatus("Scheduling tasks...");
+        boolean fullExtraction = config.getValue(ANALYSIS_CLASS).endsWith("FullAnalysis");
         for (List<RevCommit> commitSubset : commitSubsets) {
             count += commitSubset.size();
-            threadPool.submit(new AnalysisTask(commitSubset, workingDirectory, propertiesFile, splDir.getName(), config.getValue(EXTRACTION_TIMEOUT)));
+            threadPool.submit(new AnalysisTask(commitSubset, workingDirectory, propertiesFile, splDir.getName(), config.getValue(EXTRACTION_TIMEOUT), fullExtraction));
         }
         LOGGER.logStatus("all " + commitSubsets.size() + " tasks scheduled.");
         threadPool.shutdown();
@@ -157,6 +160,7 @@ private static Configuration getConfiguration(File propertiesFile) {
             config.registerSetting(RESULT_REPO_COMMITTER_NAME);
             config.registerSetting(RESULT_REPO_COMMITTER_EMAIL);
             config.registerSetting(EXTRACTION_TIMEOUT);
+            config.registerSetting(ANALYSIS_CLASS);
         } catch (SetUpException e) {
             LOGGER.logError("Invalid configuration detected:", e.getMessage());
             quitOnError();

…nc/vevos/extraction/kh/FullAnalysis.java …/vevos/extraction/kh/FullExtraction.javasrc/main/java/org/variantsync/vevos/extraction/kh/FullAnalysis.java renamed to src/main/java/org/variantsync/vevos/extraction/kh/FullExtraction.java src/main/java/org/variantsync/vevos/extraction/kh/FullAnalysis.java renamed to src/main/java/org/variantsync/vevos/extraction/kh/FullExtraction.java

@@ -7,14 +7,14 @@
 import net.ssehub.kernel_haven.util.null_checks.NonNull;
 import net.ssehub.kernel_haven.fe_analysis.pcs.CodeBlockAnalysis;
 
-public class FullAnalysis extends PipelineAnalysis {
+public class FullExtraction extends PipelineAnalysis {
 
     /**
-     * Creates a new {@link FullAnalysis}.
+     * Creates a new {@link FullExtraction}.
      *
      * @param config The global configuration.
      */
-    public FullAnalysis(@NonNull Configuration config) {
+    public FullExtraction(@NonNull Configuration config) {
         super(config);
     }
 

…vevos/extraction/kh/PartialAnalysis.java …vos/extraction/kh/PartialExtraction.javasrc/main/java/org/variantsync/vevos/extraction/kh/PartialAnalysis.java renamed to src/main/java/org/variantsync/vevos/extraction/kh/PartialExtraction.java src/main/java/org/variantsync/vevos/extraction/kh/PartialAnalysis.java renamed to src/main/java/org/variantsync/vevos/extraction/kh/PartialExtraction.java

@@ -7,14 +7,14 @@
 import net.ssehub.kernel_haven.fe_analysis.pcs.CodeBlockAnalysis;
 import net.ssehub.kernel_haven.util.null_checks.NonNull;
 
-public class PartialAnalysis extends PipelineAnalysis {
+public class PartialExtraction extends PipelineAnalysis {
 
     /**
-     * Creates a new {@link PartialAnalysis}.
+     * Creates a new {@link PartialExtraction}.
      *
      * @param config The global configuration.
      */
-    public PartialAnalysis(@NonNull Configuration config) {
+    public PartialExtraction(@NonNull Configuration config) {
         super(config);
     }
 

src/main/resources/extraction_busybox.properties

@@ -49,7 +49,7 @@ analysis.code_block.consider_missing_bm_infos = true
 #######################################
 
 # Do not change this
-analysis.class = org.variantsync.vevos.extraction.kh.FullAnalysis
+analysis.class = org.variantsync.vevos.extraction.kh.FullExtraction
 preparation.class.0 = net.ssehub.kernel_haven.busyboot.PrepareBusybox
 analysis.output.type = csv