refactor: Whitespace tokens can no longer be multi-line

Multiple whitespace tokens will be created once a newline
is encountered in the input.

Author: quintesse

src/main/java/org/codejive/properties/PropertiesParser.java

@@ -4,6 +4,7 @@
 import java.io.Reader;
 import java.util.Objects;
 import java.util.Spliterators;
+import java.util.function.BiFunction;
 import java.util.function.Consumer;
 import java.util.function.Function;
 import java.util.stream.Stream;
@@ -88,29 +89,31 @@ public Token nextToken() throws IOException {
         if (isEof(ch)) {
             return null;
         }
-        Function<Integer, Boolean> isValid = (c) -> false;
+        int oldch = -1;
+        BiFunction<Integer, Integer, Boolean> isValid = (c, oldc) -> false;
         Type nextState = null;
         if (state == null) {
             if (isCommentChar(ch)) {
                 state = Type.COMMENT;
-                isValid = this::isNotEol;
+                isValid = (c, oldc) -> !isEol(c) && !isEof(c);
             } else if (isWhitespaceChar(ch)) {
                 state = Type.WHITESPACE;
-                isValid = this::isWhitespaceChar;
+                isValid = (c, oldc) -> isWhitespaceChar(c) && !isEol(oldc);
             } else {
                 state = Type.KEY;
-                isValid = (c) -> !isSeparatorChar(c);
+                isValid = (c, oldc) -> !isSeparatorChar(c);
                 nextState = Type.SEPARATOR;
             }
         } else if (state == Type.SEPARATOR) {
-            isValid = this::isSeparatorChar;
+            isValid = (c, oldc) -> isSeparatorChar(c);
             nextState = Type.VALUE;
         } else if (state == Type.VALUE) {
-            isValid = this::isNotEol;
+            isValid = (c, oldc) -> !isEol(c) && !isEof(c);
         }
         while (true) {
-            if (isValid.apply(ch)) {
+            if (isValid.apply(ch, oldch)) {
                 addChar(readChar());
+                oldch = ch;
                 ch = peekChar();
             } else {
                 String text = (state == Type.VALUE || state == Type.COMMENT) ? trimmedString() : string();
@@ -221,7 +224,7 @@ private boolean isSeparatorChar(int ch) {
     }
 
     private boolean isWhitespaceChar(int ch) {
-        return ch == ' ' || ch == '\t' || ch == '\f' || ch == '\n' || ch == '\r';
+        return ch == ' ' || ch == '\t' || ch == '\f' || isEol(ch);
     }
 
     private boolean isCommentChar(int ch) {
@@ -233,8 +236,8 @@ private boolean isHexDigitChar(int ch) {
         return Character.isDigit(ch) || (uch >= 'A' && uch <= 'F');
     }
 
-    private boolean isNotEol(int ch) {
-        return ch != '\n' && ch != '\r' && !isEof(ch);
+    private boolean isEol(int ch) {
+        return ch == '\n' || ch == '\r';
     }
 
     private boolean isEof(int ch) {

src/test/java/org/codejive/properties/TestPropertiesParser.java

@@ -31,8 +31,8 @@ public class TestPropertiesParser {
                     + "multiline = one \\\n"
                     + "    two  \\\n\r"
                     + "\tthree\n"
-                    + "key.4 = \\u1234\n"
-                    + "# final comment";
+                    + "key.4 = \\u1234\n\r"
+                    + "  # final comment";
 
     @Test
     void testTokens() throws IOException {
@@ -45,7 +45,8 @@ void testTokens() throws IOException {
                                 new Token(Type.COMMENT, "#comment1"),
                                 new Token(Type.WHITESPACE, "\n"),
                                 new Token(Type.COMMENT, "#  comment2"),
-                                new Token(Type.WHITESPACE, "   \n\n"),
+                                new Token(Type.WHITESPACE, "   \n"),
+                                new Token(Type.WHITESPACE, "\n"),
                                 new Token(Type.COMMENT, "! comment3"),
                                 new Token(Type.WHITESPACE, "\n"),
                                 new Token(Type.KEY, "one"),
@@ -61,7 +62,8 @@ void testTokens() throws IOException {
                                 new Token(Type.KEY, "three"),
                                 new Token(Type.SEPARATOR, "="),
                                 new Token(Type.VALUE, "and escapes\\n\\t\\r\\f", "and escapes\n\t\r\f"),
-                                new Token(Type.WHITESPACE, "\n  "),
+                                new Token(Type.WHITESPACE, "\n"),
+                                new Token(Type.WHITESPACE, "  "),
                                 new Token(Type.KEY, "\\ with\\ spaces", " with spaces"),
                                 new Token(Type.SEPARATOR, "   =    "),
                                 new Token(Type.VALUE, "everywhere"),
@@ -77,7 +79,8 @@ void testTokens() throws IOException {
                                 new Token(Type.KEY, "key.4"),
                                 new Token(Type.SEPARATOR, " = "),
                                 new Token(Type.VALUE, "\\u1234", "\u1234"),
-                                new Token(Type.WHITESPACE, "\n"),
+                                new Token(Type.WHITESPACE, "\n\r"),
+                                new Token(Type.WHITESPACE, "  "),
                                 new Token(Type.COMMENT, "# final comment"))));
     }