Updated annotations to support escaping some characters inside of the annotation text

Ghidra/Features/Base/src/main/java/ghidra/app/util/viewer/field/Annotation.java

@@ -16,8 +16,6 @@
 package ghidra.app.util.viewer.field;
 
 import java.util.*;
-import java.util.regex.Matcher;
-import java.util.regex.Pattern;
 
 import docking.widgets.fieldpanel.field.AttributedString;
 import ghidra.app.nav.Navigatable;
@@ -26,12 +24,8 @@ import ghidra.program.model.listing.Program;
 import ghidra.util.classfinder.ClassSearcher;
 
 public class Annotation {
-	/**
-	 * A pattern to match text between two quote characters and to capture that text.  This
-	 * pattern does not match quote characters that are escaped with a '\' character.
-	 */
-	private static final Pattern QUOTATION_PATTERN =
-		Pattern.compile("(?<!\\\\)[\"](.*?)(?<!\\\\)[\"]");
+
+	public static final String ESCAPABLE_CHARS = "{}\"\\";
 
 	private static List<AnnotatedStringHandler> ANNOTATED_STRING_HANDLERS;
 	private static Map<String, AnnotatedStringHandler> ANNOTATED_STRING_MAP;
@@ -149,49 +143,6 @@ public class Annotation {
 			serviceProvider);
 	}
 
-	private String[] parseAnnotationText(String theAnnotationText) {
-		StringBuilder buffer = new StringBuilder(theAnnotationText);
-
-		// strip off the brackets
-		buffer.delete(0, 2); // remove '{' and '@'
-		buffer.deleteCharAt(buffer.length() - 1);
-
-		// first split out the tokens on '"' so that annotations can have groupings with
-		// whitespace
-		int unqouotedOffset = 0;
-		List<String> tokens = new ArrayList<>();
-		Matcher matcher = QUOTATION_PATTERN.matcher(buffer.toString());
-		while (matcher.find()) {
-			// put all text in the buffer,
-			int quoteStart = matcher.start();
-			String contentBeforeQuote = buffer.substring(unqouotedOffset, quoteStart);
-			grabTokens(tokens, contentBeforeQuote);
-			unqouotedOffset = matcher.end();
-
-			String quotedContent = matcher.group(1); // group 0 is the entire string
-			tokens.add(quotedContent);
-		}
-
-		// handle any remaining part of the text after quoted sections
-		if (unqouotedOffset < buffer.length()) {
-			String remainingString = buffer.substring(unqouotedOffset);
-			grabTokens(tokens, remainingString);
-		}
-
-		// split on whitespace
-		return tokens.toArray(new String[tokens.size()]);
-	}
-
-	private void grabTokens(List<String> tokenContainer, String content) {
-		String[] strings = content.split("\\s");
-		for (String string : strings) {
-			// 0 length strings can happen when 'content' begins with a space
-			if (string.length() > 0) {
-				tokenContainer.add(string);
-			}
-		}
-	}
-
 	public String getAnnotationText() {
 		return annotationText;
 	}
@@ -204,4 +155,137 @@ public class Annotation {
 	/*package*/ static Set<String> getAnnotationNames() {
 		return Collections.unmodifiableSet(getAnnotatedStringHandlerMap().keySet());
 	}
+
+	private String[] parseAnnotationText(String text) {
+
+		String trimmed = text.substring(2, text.length() - 1); // remove "{@" and '}' 
+		List<String> tokens = new ArrayList<>();
+		List<TextPart> parts = parseText(trimmed);
+		for (TextPart part : parts) {
+			part.grabTokens(tokens);
+		}
+
+		return tokens.toArray(new String[tokens.size()]);
+	}
+
+	private List<TextPart> parseText(String text) {
+
+		List<TextPart> textParts = new ArrayList<>();
+		boolean escaped = false;
+		boolean inQuote = false;
+		int partStart = 0;
+		int n = text.length();
+		for (int i = 0; i < n; i++) {
+
+			boolean wasEscaped = escaped;
+			escaped = false;
+			char prev = '\0';
+			if (i != 0 && !wasEscaped) {
+				prev = text.charAt(i - 1);
+			}
+
+			char c = text.charAt(i);
+			if (prev == '\\') {
+				if (Annotation.ESCAPABLE_CHARS.indexOf(c) != -1) {
+					escaped = true;
+					continue;
+				}
+			}
+
+			if (c == '"') {
+				if (inQuote) {
+					// end quote
+					String s = text.substring(partStart, i + 1); // keep the quote
+					textParts.add(new QuotedTextPart(s));
+					partStart = i + 1;
+				}
+				else {
+					// end previous word; start quote
+					if (i != 0) {
+						String s = text.substring(partStart, i);
+						textParts.add(new TextPart(s));
+						partStart = i;
+					}
+				}
+				inQuote = !inQuote;
+			}
+		}
+
+		if (partStart < n) { // grab trailing text
+			String s = text.substring(partStart, n);
+			textParts.add(new TextPart(s));
+		}
+
+		return textParts;
+	}
+
+	// remove any backslashes that escape special annotation characters, like '{' and '}'
+	private static String removeEscapeChars(String text) {
+		boolean escaped = false;
+		StringBuilder buffy = new StringBuilder();
+		for (int i = 0; i < text.length(); i++) {
+			char c = text.charAt(i);
+			boolean wasEscaped = escaped;
+			escaped = false;
+			if (c != '\\') {
+				buffy.append(c);
+				continue;
+			}
+
+			char next = '\0';
+			if (i != text.length() - 1 && !wasEscaped) {
+				next = text.charAt(i + 1);
+			}
+
+			if (ESCAPABLE_CHARS.indexOf(next) != -1) {
+				escaped = true;
+				continue;
+			}
+			buffy.append(c);
+		}
+
+		return buffy.toString();
+	}
+
+	/**
+	 * A simple class to hold text and extract tokens 
+	 */
+	private class TextPart {
+
+		protected String text;
+
+		TextPart(String text) {
+			this.text = text;
+		}
+
+		public void grabTokens(List<String> tokens) {
+			String escaped = removeEscapeChars(text);
+			String[] strings = escaped.split("\\s");
+			for (String string : strings) {
+				// 0 length strings can happen when 'content' begins with a space
+				if (string.length() > 0) {
+					tokens.add(string);
+				}
+			}
+		}
+
+		@Override
+		public String toString() {
+			return text;
+		}
+	}
+
+	private class QuotedTextPart extends TextPart {
+		QuotedTextPart(String text) {
+			super(text);
+		}
+
+		@Override
+		public void grabTokens(List<String> tokens) {
+			String unquoted = text.substring(1, text.length() - 1);
+			String escaped = removeEscapeChars(unquoted);
+			tokens.add(escaped); // all quoted text is a 'token'
+		}
+	}
+
 }

Ghidra/Features/Base/src/main/java/ghidra/app/util/viewer/field/CommentUtils.java

@@ -191,8 +191,7 @@ public class CommentUtils {
 
 		List<WordLocation> annotations = getCommentAnnotations(text);
 		if (annotations.isEmpty()) {
-			String updatedText = removeEscapeChars(text);
-			results.add(new StringCommentPart(text, updatedText, prototype));
+			results.add(new StringCommentPart(text, prototype));
 			return results;
 		}
 
@@ -204,23 +203,20 @@ public class CommentUtils {
 			if (offset != start) {
 				// text between annotations
 				String preceeding = text.substring(offset, start);
-				String updatedText = removeEscapeChars(preceeding);
-				results.add(new StringCommentPart(preceeding, updatedText, prototype));
+				results.add(new StringCommentPart(preceeding, prototype));
 			}
 
 			String annotationText = word.getWord();
-			String updatedText = removeEscapeChars(annotationText);
-			Annotation annotation = new Annotation(updatedText, prototype, program);
+			Annotation annotation = new Annotation(annotationText, prototype, program);
 			annotation = fixerUpper.apply(annotation);
-			results.add(new AnnotationCommentPart(updatedText, annotation));
+			results.add(new AnnotationCommentPart(annotationText, annotation));
 
 			offset = start + annotationText.length();
 		}
 
 		if (offset != text.length()) { // trailing text
 			String trailing = text.substring(offset);
-			String updatedText = removeEscapeChars(trailing);
-			results.add(new StringCommentPart(trailing, updatedText, prototype));
+			results.add(new StringCommentPart(trailing, prototype));
 		}
 
 		return results;
@@ -269,23 +265,6 @@ public class CommentUtils {
 		//@formatter:on
 	}
 
-	// remove any backslashes that escape special annotation characters, like '{' and '}'
-	private static String removeEscapeChars(String text) {
-		StringBuilder buffy = new StringBuilder();
-		for (int i = 0; i < text.length(); i++) {
-			char c = text.charAt(i);
-			if (c == '\\') {
-				char next = i == text.length() ? '\0' : text.charAt(i + 1);
-				if (next == '{' || next == '}') {
-					continue;
-				}
-			}
-			buffy.append(c);
-		}
-
-		return buffy.toString();
-	}
-
 	/*
 	 * Starts at the given index and looks for the end an annotation, ignoring quoted text 
 	 * and escaped characters along the way.   The value returned is the index after the last 
@@ -294,26 +273,30 @@ public class CommentUtils {
 	 */
 	private static int findAnnotationEnd(String comment, int start) {
 
-		boolean startQuote = false;
-		int count = 0;
+		boolean escaped = false;
+		boolean inQuote = false;
 		for (int i = start; i < comment.length(); i++) {
-			char prev = i == 0 ? '\0' : comment.charAt(i - 1);
-			if (prev == '\\') {
-				continue; // escaped
+
+			boolean wasEscaped = escaped;
+			escaped = false;
+			char prev = '\0';
+			if (i != 0 && !wasEscaped) {
+				prev = comment.charAt(i - 1);
 			}
 
 			char c = comment.charAt(i);
+			if (prev == '\\') {
+				if (Annotation.ESCAPABLE_CHARS.indexOf(c) != -1) {
+					escaped = true;
+					continue;
+				}
+			}
+
 			if (c == '"') {
-				if (startQuote) {
-					--count;
-				}
-				else {
-					++count;
-				}
-				startQuote = !startQuote;
+				inQuote = !inQuote;
 			}
 			else if (c == '}') {
-				if (count == 0) {
+				if (!inQuote) {
 					return i + 1;
 				}
 			}

Ghidra/Features/Base/src/main/java/ghidra/app/util/viewer/field/StringCommentPart.java

@@ -20,21 +20,15 @@ import docking.widgets.fieldpanel.field.*;
 public class StringCommentPart extends CommentPart {
 
 	private AttributedString prototype;
-	private String rawText;
 
-	StringCommentPart(String rawText, AttributedString prototype) {
-		this(rawText, rawText, prototype);
-	}
-
-	StringCommentPart(String rawText, String displayText, AttributedString prototype) {
-		super(displayText);
-		this.rawText = rawText;
+	StringCommentPart(String text, AttributedString prototype) {
+		super(text);
 		this.prototype = prototype;
 	}
 
 	@Override
 	String getRawText() {
-		return rawText;
+		return getDisplayText();
 	}
 
 	@Override
@@ -45,6 +39,6 @@ public class StringCommentPart extends CommentPart {
 
 	@Override
 	public String toString() {
-		return rawText;
+		return getDisplayText();
 	}
 }

Ghidra/Features/Base/src/test.slow/java/ghidra/app/util/viewer/field/AnnotationTest.java

@@ -206,16 +206,64 @@ public class AnnotationTest extends AbstractGhidraHeadedIntegrationTest {
 
 	@Test
 	public void testSymbolAnnotation_WithEscapedItemsOutsideOfAnnotation() {
-		String rawComment = "This is a foo\\} symbol {@sym mySym\\{0\\}} annotation \\{bar";
+		String rawComment = "This is a foo} symbol {@sym mySym\\{0\\}} annotation {bar";
 		String display = CommentUtils.getDisplayString(rawComment, program);
 		assertEquals("This is a foo} symbol mySym{0} annotation {bar", display);
 	}
 
+	@Test
+	public void testAddressAnnotation_QuotedQuote() {
+		String rawComment = "Test {@address 0 \"quote\\\"\"} extra}";
+		String display = CommentUtils.getDisplayString(rawComment, program);
+		assertEquals("Test quote\" extra}", display);
+	}
+
+	@Test
+	public void testAddressAnnotation_EscapedBrace() {
+		String rawComment = "Test {@address 0 \"quote\\}\"} blah";
+		String display = CommentUtils.getDisplayString(rawComment, program);
+		assertEquals("Test quote} blah", display);
+	}
+
+	@Test
+	public void testAddressAnnotation_BackslashAndEscapedBrace() {
+		String rawComment = "Test {@address 0 \"quote\\\\}\"} blah";
+		String display = CommentUtils.getDisplayString(rawComment, program);
+		assertEquals("Test quote\\} blah", display);
+	}
+
+	@Test
+	public void testAddressAnnotation_BackslashBackslash() {
+		String rawComment = "Test {@address 0 \"quote\\\\\"} blah";
+		String display = CommentUtils.getDisplayString(rawComment, program);
+		assertEquals("Test quote\\ blah", display);
+	}
+
+	@Test
+	public void testAddressAnnotation_LonelyBackslash() {
+		String rawComment = "Test {@address 0 bo\\b} some text";
+		String display = CommentUtils.getDisplayString(rawComment, program);
+		assertEquals("Test bo\\b some text", display);
+	}
+
 	@Test
 	public void testSymbolAnnotation_FullyEscaped() {
+		// We currently don't support rendering escaped annotation characters unless they are 
+		// inside of an annotation.
 		String rawComment = "This is a symbol \\{@sym bob\\} annotation";
 		String display = CommentUtils.getDisplayString(rawComment, program);
-		assertEquals("This is a symbol {@sym bob} annotation", display);
+		assertEquals("This is a symbol \\{@sym bob\\} annotation", display);
+	}
+
+	@Test
+	public void testSymbolAnnotation_LonelyBackslash() {
+		// We currently don't support rendering escaped annotation characters unless they are 
+		// inside of an annotation.
+		String rawComment = "This is a symbol {@sym bob jo\\e} annotation";
+		String display = CommentUtils.getDisplayString(rawComment, program);
+
+		// Note: Symbol Annotations ignore display text which means that the symbol name 
+		assertEquals("This is a symbol bob annotation", display);
 	}
 
 	@Test

Ghidra/Features/Base/src/test.slow/java/ghidra/app/util/viewer/field/CommentUtilsTest.java

@@ -37,7 +37,7 @@ public class CommentUtilsTest extends AbstractGhidraHeadlessIntegrationTest {
 	@Test
 	public void testGetCommentAnnotations_PlainAnnotation() {
 
-		String comment = "This is an {@symbol symbolName}";
+		String comment = "This is a {@symbol symbolName}";
 		List<WordLocation> annotations = CommentUtils.getCommentAnnotations(comment);
 		assertEquals(1, annotations.size());
 		WordLocation word = annotations.get(0);
@@ -47,7 +47,7 @@ public class CommentUtilsTest extends AbstractGhidraHeadlessIntegrationTest {
 	@Test
 	public void testGetCommentAnnotations_QuotedAnnotation() {
 
-		String comment = "This is an {@symbol \"symbolName\"}";
+		String comment = "This is a {@symbol \"symbolName\"}";
 		List<WordLocation> annotations = CommentUtils.getCommentAnnotations(comment);
 		assertEquals(1, annotations.size());
 		WordLocation word = annotations.get(0);
@@ -57,7 +57,7 @@ public class CommentUtilsTest extends AbstractGhidraHeadlessIntegrationTest {
 	@Test
 	public void testGetCommentAnnotations_QuotedAnnotation_WithEscapedQuotes() {
 
-		String comment = "This is an {@symbol \"symbol\\\"Name\\\"\"}";
+		String comment = "This is a {@symbol \"symbol\\\"Name\\\"\"}";
 		List<WordLocation> annotations = CommentUtils.getCommentAnnotations(comment);
 		assertEquals(1, annotations.size());
 		WordLocation word = annotations.get(0);
@@ -67,7 +67,7 @@ public class CommentUtilsTest extends AbstractGhidraHeadlessIntegrationTest {
 	@Test
 	public void testGetCommentAnnotations_QuotedAnnotationWithBraces() {
 
-		String comment = "This is an {@symbol \"symbol{Name}\"}";
+		String comment = "This is a {@symbol \"symbol{Name}\"}";
 		List<WordLocation> annotations = CommentUtils.getCommentAnnotations(comment);
 		assertEquals(1, annotations.size());
 		WordLocation word = annotations.get(0);
@@ -79,7 +79,7 @@ public class CommentUtilsTest extends AbstractGhidraHeadlessIntegrationTest {
 
 		// the second brace is ignored (if the first brace is part of the symbol name, then it
 		// needs to be escaped or quoted
-		String comment = "This is an {@symbol symbol{Name}}";
+		String comment = "This is a {@symbol symbol{Name}}";
 		List<WordLocation> annotations = CommentUtils.getCommentAnnotations(comment);
 		assertEquals(1, annotations.size());
 		WordLocation word = annotations.get(0);
@@ -90,7 +90,7 @@ public class CommentUtilsTest extends AbstractGhidraHeadlessIntegrationTest {
 	public void testGetCommentAnnotations_UnquotedAnnotation_WithUnbalancedBraces() {
 
 		// the second brace is ignored
-		String comment = "This is an {@symbol symbolName}}";
+		String comment = "This is a {@symbol symbolName}}";
 		List<WordLocation> annotations = CommentUtils.getCommentAnnotations(comment);
 		assertEquals(1, annotations.size());
 		WordLocation word = annotations.get(0);
@@ -101,7 +101,7 @@ public class CommentUtilsTest extends AbstractGhidraHeadlessIntegrationTest {
 	public void testGetCommentAnnotations_UnquotedAnnotation_WithEscapedBraces() {
 
 		// escaped braces get ignored
-		String comment = "This is an {@symbol symbol\\{Name\\}}";
+		String comment = "This is a {@symbol symbol\\{Name\\}}";
 		List<WordLocation> annotations = CommentUtils.getCommentAnnotations(comment);
 		assertEquals(1, annotations.size());
 		WordLocation word = annotations.get(0);

Ghidra/Features/GraphFunctionCalls/src/main/java/functioncalls/graph/view/FcgComponent.java

@@ -53,12 +53,6 @@ public class FcgComponent extends GraphComponent<FcgVertex, FcgEdge, FunctionCal
 		build();
 	}
 
-	@Override
-	protected void setGraph(FunctionCallGraph g) {
-		this.fcGraph = g;
-		super.setGraph(g);
-	}
-
 	@Override
 	protected FcgVertex getInitialVertex() {
 		return fcGraph.getSource();