1 files changed, 139 insertions, 29 deletions
diff --git a/BJC-Utils2/src/main/java/bjc/utils/funcutils/TokenSplitter.java b/BJC-Utils2/src/main/java/bjc/utils/funcutils/TokenSplitter.java
index 84f5270..084bdae 100644
--- a/BJC-Utils2/src/main/java/bjc/utils/funcutils/TokenSplitter.java
+++ b/BJC-Utils2/src/main/java/bjc/utils/funcutils/TokenSplitter.java
@@ -1,47 +1,157 @@
 package bjc.utils.funcutils;
 
-import bjc.utils.funcdata.FunctionalList;
-import bjc.utils.funcdata.IList;
+import java.util.regex.Pattern;
 
-import java.util.Iterator;
-import java.util.function.BiFunction;
+/**
+ * Split a string and keep given delimiters.
+ *
+ * @author Ben Culkin
+ */
+public class TokenSplitter {
+	/*
+	 * This string is a format template for the delimiter matching regex
+	 *
+	 * It does two things:
+	 * 
+	 * <ol>
+	 * <li> Match to the left of the provided delimiter by positive lookahead </li>
+	 * <li> Match to the right of the provided delimiter by positive lookbehind </li>
+	 * </ol>
+	 *
+	 * Thus, it will only match in places where the delimiter is, but won't
+	 * actually match the delimiter, leaving split to put it into the stream
+	 */
+	private static String WITH_DELIM = "((?<=%1$s)|(?=%1$s))";
 
-final class TokenSplitter implements BiFunction<String, String, IList<String>> {
-	private String tokenToSplit;
+	/*
+	 * This string is a format template for the multi-delimiter matching
+	 * regex.
+	 *
+	 * It does the same thing as the single delimiter regex, but has to have
+	 * some negative lookahead/lookbehind assertions to avoid splitting a
+	 * delimiter into pieces.
+	 */
+	private static String WITH_MULTI_DELIM = "((?<=%1$s+)(?!%1$s)|(?<!%1$s)(?=%1$s+))";
 
-	public TokenSplitter(String tok) {
-		this.tokenToSplit = tok;
+	/*
+	 * These represent the internal state of the splitter.
+	 */
+	private StringBuilder	currPatt;
+	private StringBuilder	currExclusionPatt;
+
+	/*
+	 * These represent the external state of the splitter.
+	 *
+	 * Compilation causes internal to become external.
+	 */
+	private Pattern	compPatt;
+	private Pattern	exclusionPatt;
+
+	/**
+	 * Create a new token splitter.
+	 */
+	public TokenSplitter() {
 	}
 
-	@Override
-	public IList<String> apply(String operatorName, String operatorRegex) {
-		if(operatorName == null)
-			throw new NullPointerException("Operator name must not be null");
-		else if(operatorRegex == null) throw new NullPointerException("Operator regex must not be null");
+	/**
+	 * Split a provided string using configured delimiters, and keeping the
+	 * delimiters.
+	 *
+	 * The splitter must be compiled first.
+	 *
+	 * @param inp
+	 *                The string to split.
+	 *
+	 * @return The split string, including delimiters.
+	 *
+	 * @throws IllegalStateException
+	 *                 If the splitter isn't compiled.
+	 */
+	public String[] split(String inp) {
+		if(compPatt == null) throw new IllegalStateException("Token splitter has not been compiled yet");
+
+		/*
+		 * Don't split something that we should exclude from being split.
+		 */
+		if(exclusionPatt.matcher(inp).matches()) return new String[] { inp };
 
-		if(tokenToSplit.contains(operatorName)) {
-			if(StringUtils.containsOnly(tokenToSplit, operatorRegex))
-				return new FunctionalList<>(tokenToSplit);
+		return compPatt.split(inp);
+	}
 
-			IList<String> splitTokens = new FunctionalList<>(tokenToSplit.split(operatorRegex));
-			IList<String> result = new FunctionalList<>();
+	/**
+	 * Adds a string as a matched delimiter to split on.
+	 *
+	 * Only works for fixed length delimiters.
+	 *
+	 * The provided string is regex-escaped before being used.
+	 *
+	 * @param delim
+	 *                The delimiter to match on.
+	 */
+	public void addDelimiter(String delim) {
+		String quoteDelim = Pattern.quote(delim);
+		String delimPat   = String.format(WITH_DELIM, quoteDelim);
 
-			Iterator<String> itr = splitTokens.toIterable().iterator();
-			int tokenExpansionSize = splitTokens.getSize();
+		if(currPatt == null) {
+			currPatt          = new StringBuilder();
+			currExclusionPatt = new StringBuilder();
 
-			String elm = itr.next();
+			currPatt.append("(?:" + delimPat + ")");
+			currExclusionPatt.append("(?:" + quoteDelim + ")");
+		} else {
+			currPatt.append("|(?:" + delimPat + ")");
+			currExclusionPatt.append("|(?:" + quoteDelim + ")");
+		}
+	}
 
-			for(int i = 0; itr.hasNext(); elm = itr.next()) {
-				result.add(elm);
+	/**
+	 * Adds a character class as a matched delimiter to split on.
+	 *
+	 * The provided string should be a pattern to match one or more
+	 * occurances of.
+	 *
+	 * @param delim
+	 *                The delimiter to split on.
+	 */
+	public void addMultiDelimiter(String delim) {
+		String delimPat = String.format(WITH_MULTI_DELIM, "(?:" + delim + ")");
 
-				if(i != tokenExpansionSize) {
-					result.add(operatorName);
-				}
-			}
+		if(currPatt == null) {
+			currPatt          = new StringBuilder();
+			currExclusionPatt = new StringBuilder();
 
-			return result;
+			currPatt.append("(?:" + delimPat + ")");
+			currExclusionPatt.append("(?:(?:" + delim + ")+)");
+
+		} else {
+			currPatt.append("|(?:" + delimPat + ")");
+			currExclusionPatt.append("|(?:(?:" + delim + ")+)");
 		}
+	}
+
+	/**
+	 * Marks strings matching the pattern delim as non-splittable.
+	 *
+	 * @param delim
+	 * 		The regex to not splitting matching strings.
+	 */
+	public void addNonMatcher(String delim) {
+		if(currPatt == null) {
+			currPatt          = new StringBuilder();
+			currExclusionPatt = new StringBuilder();
 
-		return new FunctionalList<>(tokenToSplit);
+			currExclusionPatt.append("(?:" + delim + ")");
+		} else {
+			currExclusionPatt.append("|(?:" + delim + ")");
+		}
+	}
+	/**
+	 * Compiles the current set of delimiters to a pattern.
+	 *
+	 * Makes this splitter ready to use.
+	 */
+	public void compile() {
+		compPatt      = Pattern.compile(currPatt.toString());
+		exclusionPatt = Pattern.compile(currExclusionPatt.toString());
 	}
 }