GP-4294 - Fixed exception looking for extensions when running headlessly in fat jar mode

Ghidra/Features/Base/ghidra_scripts/BuildGhidraJarScript.java

@@ -62,9 +62,12 @@ public class BuildGhidraJarScript extends GhidraScript {
 		builder.addExcludedFileExtension(".pdf");
 
 		File installDir = Application.getInstallationDirectory().getFile(true);
-		builder.buildJar(new File(installDir, "ghidra.jar"), null, monitor);
+		File file = new File(installDir, "ghidra.jar");
+		builder.buildJar(file, null, monitor);
 
 		// uncomment the following line to create a src zip for debugging.
 		// builder.buildSrcZip(new File(installDir, "GhidraSrc.zip"), monitor);
+
+		println("Finsished writing jar: " + file);
 	}
 }

Ghidra/Features/Base/src/main/java/ghidra/app/util/headless/HeadlessAnalyzer.java

@@ -59,16 +59,16 @@ import ghidra.util.task.TaskMonitor;
 import utilities.util.FileUtilities;
 
 /**
- * The class used kick-off and interact with headless processing.  All headless options have been 
+ * The class used kick-off and interact with headless processing.  All headless options have been
- * broken out into their own class: {@link HeadlessOptions}.  This class is intended to be used 
+ * broken out into their own class: {@link HeadlessOptions}.  This class is intended to be used
  * one of two ways:
  * <ul>
- *   <li>Used by {@link AnalyzeHeadless} to perform headless analysis based on arguments specified 
+ *   <li>Used by {@link AnalyzeHeadless} to perform headless analysis based on arguments specified
  *   on the command line.</li>
  *   <li>Used by another tool as a library to perform headless analysis.</li>
  * </ul>
  * <p>
- * Note: This class is not thread safe.  
+ * Note: This class is not thread safe.
  */
 public class HeadlessAnalyzer {
 
@@ -84,16 +84,16 @@ public class HeadlessAnalyzer {
 	private FileSystemService fsService;
 
 	/**
-	 * Gets a headless analyzer, initializing the application if necessary with the specified 
+	 * Gets a headless analyzer, initializing the application if necessary with the specified
-	 * logging parameters.  An {@link IllegalStateException} will be thrown if the application has 
+	 * logging parameters.  An {@link IllegalStateException} will be thrown if the application has
 	 * already been initialized or a headless analyzer has already been retrieved.  In these cases,
 	 * the headless analyzer should be gotten with {@link HeadlessAnalyzer#getInstance()}.
 	 * 
-	 * @param logFile The desired application log file.  If null, the default application log file 
+	 * @param logFile The desired application log file.  If null, the default application log file
 	 *   will be used (see {@link Application#initializeLogging}).
 	 * @param scriptLogFile The desired scripting log file.  If null, the default scripting log file
 	 *   will be used (see {@link Application#initializeLogging}).
-	 * @param useLog4j true if log4j is to be used; otherwise, false.  If this class is being used by 
+	 * @param useLog4j true if log4j is to be used; otherwise, false.  If this class is being used by
 	 *     another tool as a library, using log4j might interfere with that tool.
 	 * @return An instance of a new headless analyzer.
 	 * @throws IllegalStateException if an application or headless analyzer instance has already been initialized.
@@ -103,7 +103,7 @@ public class HeadlessAnalyzer {
 			boolean useLog4j) throws IllegalStateException, IOException {
 
 		// Prevent more than one headless analyzer from being instantiated.  Too much about it
-		// messes with global system settings, so under the current design of Ghidra, allowing 
+		// messes with global system settings, so under the current design of Ghidra, allowing
 		// more than one to exist could result in unpredictable behavior.
 		if (instance != null) {
 			throw new IllegalStateException(
@@ -141,7 +141,7 @@ public class HeadlessAnalyzer {
 
 	/**
 	 * Gets a headless analyzer instance, with the assumption that the application has already been
-	 * initialized.  If this is called before the application has been initialized, it will 
+	 * initialized.  If this is called before the application has been initialized, it will
 	 * initialize the application with no logging.
 	 * 
 	 * @return An instance of a new headless analyzer.
@@ -151,7 +151,7 @@ public class HeadlessAnalyzer {
 	public static HeadlessAnalyzer getInstance() throws IOException {
 
 		// Prevent more than one headless analyzer from being instantiated.  Too much about it
-		// messes with global system settings, so under the current design of Ghidra, allowing 
+		// messes with global system settings, so under the current design of Ghidra, allowing
 		// more than one to exist could result in unpredictable behavior.
 		if (instance != null) {
 			return instance;
@@ -185,8 +185,10 @@ public class HeadlessAnalyzer {
 			layout = new GhidraApplicationLayout();
 		}
 		catch (IOException e) {
+			Msg.debug(HeadlessAnalyzer.class,
+				"Unable to load the standard Ghidra application layout. " + e.getMessage() +
+					".  Attempting to load the Ghidra Jar application layout.");
 			layout = new GhidraJarApplicationLayout();
-
 		}
 		return layout;
 	}
@@ -201,7 +203,7 @@ public class HeadlessAnalyzer {
 		// Ghidra URL handler registration.  There's no harm in doing this more than once.
 		Handler.registerHandler();
 
-		// Ensure that we are running in "headless mode",  preventing Swing-based methods from 
+		// Ensure that we are running in "headless mode",  preventing Swing-based methods from
 		// running (causing headless operation to lose focus).
 		System.setProperty("java.awt.headless", "true");
 		System.setProperty(SystemUtilities.HEADLESS_PROPERTY, Boolean.TRUE.toString());
@@ -244,12 +246,12 @@ public class HeadlessAnalyzer {
 	 * <li>perform auto-analysis if not disabled</li>
 	 * <li>execute ordered list of post-scripts</li>
 	 * </ol>
-	 * If no import files or directories have been specified the ordered list 
+	 * If no import files or directories have been specified the ordered list
 	 * of pre/post scripts will be executed once.
 	 * 
 	 * @param ghidraURL ghidra URL for existing server repository and optional
 	 *                  folder path
-	 * @param filesToImport directories and files to be imported (null or empty 
+	 * @param filesToImport directories and files to be imported (null or empty
 	 *                      is acceptable if we are in -process mode)
 	 * @throws IOException if there was an IO-related problem
 	 * @throws MalformedURLException specified URL is invalid
@@ -370,16 +372,16 @@ public class HeadlessAnalyzer {
 	 * <li>perform auto-analysis if not disabled</li>
 	 * <li>execute ordered list of post-scripts</li>
 	 * </ol>
-	 * If no import files or directories have been specified the ordered list 
+	 * If no import files or directories have been specified the ordered list
 	 * of pre/post scripts will be executed once.
 	 * 
-	 * @param projectLocation directory path of project 
+	 * @param projectLocation directory path of project
 	 * 						  If project exists it will be opened, otherwise it will be created.
 	 * @param projectName project name
 	 * @param rootFolderPath root folder for imports
 	 * @param filesToImport directories and files to be imported (null or empty is acceptable if
 	 *        				we are in -process mode)
-	 * @throws IOException if there was an IO-related problem.  If caused by a failure to obtain a 
+	 * @throws IOException if there was an IO-related problem.  If caused by a failure to obtain a
 	 * write-lock on the project the exception cause will a {@code LockException}.
 	 */
 	public void processLocal(String projectLocation, String projectName, String rootFolderPath,
@@ -475,7 +477,7 @@ public class HeadlessAnalyzer {
 	/**
 	 * Checks to see if the most recent analysis timed out.
 	 * 
-	 * @return true if the most recent analysis timed out; otherwise, false. 
+	 * @return true if the most recent analysis timed out; otherwise, false.
 	 */
 	public boolean checkAnalysisTimedOut() {
 		return analysisTimedOut;
@@ -766,7 +768,7 @@ public class HeadlessAnalyzer {
 				Class<?> c = Class.forName(className, true, classLoaderForDotClassScripts);
 
 				if (GhidraScript.class.isAssignableFrom(c)) {
-					// No issues, but return null, which signifies we don't actually have a 
+					// No issues, but return null, which signifies we don't actually have a
 					// ResourceFile to associate with the script name
 					return null;
 				}
@@ -962,9 +964,9 @@ public class HeadlessAnalyzer {
 	 * @param fileAbsolutePath Path of the file to analyze.
 	 * @param program The program to analyze.
 	 * @return true if the program file should be kept.  If analysis or scripts have marked
-	 * 			the program as temporary changes should not be saved.  Returns false in 
+	 * 			the program as temporary changes should not be saved.  Returns false in
 	 * 			these cases:
-	 * 		- One of the scripts sets the Headless Continuation Option to "ABORT_AND_DELETE" or 
+	 * 		- One of the scripts sets the Headless Continuation Option to "ABORT_AND_DELETE" or
 	 * 			"CONTINUE_THEN_DELETE".
 	 */
 	private boolean analyzeProgram(String fileAbsolutePath, Program program) {
@@ -1154,7 +1156,7 @@ public class HeadlessAnalyzer {
 
 			Msg.info(this, "REPORT: Processing project file: " + domFile.getPathname());
 
-			// This method already takes into account whether the user has set the "noanalysis" 
+			// This method already takes into account whether the user has set the "noanalysis"
 			// flag or not
 			keepFile = analyzeProgram(domFile.getPathname(), program) || readOnlyFile;
 
@@ -1237,7 +1239,7 @@ public class HeadlessAnalyzer {
 
 			if (!readOnlyFile) { // can't change anything if read-only file
 
-				// Undo checkout of it is still checked-out and either the file is to be 
+				// Undo checkout of it is still checked-out and either the file is to be
 				// deleted, or we just checked it out and file changes have been committed
 				if (domFile.isCheckedOut()) {
 					if (!keepFile ||
@@ -1521,14 +1523,14 @@ public class HeadlessAnalyzer {
 		try {
 
 			// Perform the load.  Note that loading 1 file may result in more than 1 thing getting
-			// loaded. 
+			// loaded.
 			loadResults = loadPrograms(fsrl, folderPath);
 			Msg.info(this, "IMPORTING: Loaded " + (loadResults.size() - 1) + " additional files");
 
 			primary = loadResults.getPrimary();
 			Program primaryProgram = primary.getDomainObject();
 
-			// Make sure we are allowed to save ALL programs to the project.  If not, save none and 
+			// Make sure we are allowed to save ALL programs to the project.  If not, save none and
 			// fail.
 			if (!options.readOnly) {
 				for (Loaded<Program> loaded : loadResults) {
@@ -1549,7 +1551,7 @@ public class HeadlessAnalyzer {
 			// TODO: Analyze non-primary programs (GP-2965).
 			boolean doSave = analyzeProgram(fsrl.toString(), primaryProgram) && !options.readOnly;
 
-			// The act of marking the program as temporary by a script will signal 
+			// The act of marking the program as temporary by a script will signal
 			// us to discard any changes
 			if (!doSave) {
 				loadResults.forEach(e -> e.getDomainObject().setTemporary(true));
@@ -1776,7 +1778,7 @@ public class HeadlessAnalyzer {
 					return;
 
 				default:
-					// Just continue					
+					// Just continue
 			}
 
 			runScriptsList(options.postScripts, options.postScriptFileMap, scriptState,
@@ -1834,7 +1836,7 @@ public class HeadlessAnalyzer {
 	}
 
 	/**
-	 * Ghidra project class required to gain access to specialized project constructor 
+	 * Ghidra project class required to gain access to specialized project constructor
 	 * for URL connection.
 	 */
 	private static class HeadlessProject extends DefaultProject {

Ghidra/Framework/Utility/src/main/java/ghidra/GhidraJarApplicationLayout.java

@@ -23,11 +23,11 @@ import java.util.*;
 import generic.jar.ResourceFile;
 import ghidra.framework.ApplicationProperties;
 import ghidra.framework.GModule;
-import utility.application.ApplicationLayout;
+import ghidra.util.Msg;
 import utility.module.ModuleUtilities;
 
 /**
- * The Ghidra jar application layout defines the customizable elements of the Ghidra application's 
+ * The Ghidra jar application layout defines the customizable elements of the Ghidra application's
  * directory structure when running in "single jar mode."
  */
 public class GhidraJarApplicationLayout extends GhidraApplicationLayout {
@@ -51,7 +51,11 @@ public class GhidraJarApplicationLayout extends GhidraApplicationLayout {
 	protected Collection<ResourceFile> findGhidraApplicationRootDirs() {
 		List<ResourceFile> dirs = new ArrayList<>();
 		String appPropPath = "/_Root/Ghidra/" + ApplicationProperties.PROPERTY_FILE;
-		URL appPropUrl = ApplicationLayout.class.getResource(appPropPath);
+		URL appPropUrl = getClass().getResource(appPropPath);
+		if (appPropUrl == null) {
+			throw new IllegalStateException(
+				"The Ghidra Jar must have an application.properties file at " + appPropPath);
+		}
 		ResourceFile rootDir = fromUrl(appPropUrl).getParentFile();
 		dirs.add(rootDir);
 		return dirs;
@@ -79,7 +83,12 @@ public class GhidraJarApplicationLayout extends GhidraApplicationLayout {
 
 	@Override
 	protected List<ResourceFile> findExtensionInstallationDirectories() {
-		URL extensionInstallUrl = ApplicationLayout.class.getResource("/_Root/Ghidra/Extensions");
+		String path = "/_Root/Ghidra/Extensions";
+		URL extensionInstallUrl = getClass().getResource(path);
+		if (extensionInstallUrl == null) {
+			Msg.debug(this, "No Extensions dir found at " + path);
+			return List.of();
+		}
 		ResourceFile extensionInstallDir = fromUrl(extensionInstallUrl);
 		return Collections.singletonList(extensionInstallDir);
 	}
@@ -94,7 +103,7 @@ public class GhidraJarApplicationLayout extends GhidraApplicationLayout {
 		String urlString = url.toExternalForm();
 		try {
 			// Decode the URL to replace things like %20 with real spaces.
-			// Note: can't use URLDecoder.decode(String, Charset) because Utility must be 
+			// Note: can't use URLDecoder.decode(String, Charset) because Utility must be
 			// Java 1.8 compatible.
 			urlString = URLDecoder.decode(urlString, "UTF-8");
 		}