Revert "experiment: avoid string creation from char[] for tags and attribute names"

This reverts commit d8f44bc.

Author: rbri

src/main/java/org/htmlunit/cyberneko/HTMLScanner.java

@@ -28,7 +28,6 @@
 import org.htmlunit.cyberneko.HTMLElements.Element;
 import org.htmlunit.cyberneko.io.PlaybackInputStream;
 import org.htmlunit.cyberneko.util.MiniStack;
-import org.htmlunit.cyberneko.util.StringCache;
 import org.htmlunit.cyberneko.xerces.util.EncodingTranslator;
 import org.htmlunit.cyberneko.xerces.util.NamespaceSupport;
 import org.htmlunit.cyberneko.xerces.util.StandardEncodingTranslator;
@@ -524,8 +523,6 @@ public class HTMLScanner implements XMLDocumentSource, XMLLocator, HTMLComponent
     /** Reusable parser for numeric character references (&#x...; and &#...;) */
     private final HTMLUnicodeEntitiesParser fUnicodeEntitiesParser = new HTMLUnicodeEntitiesParser();
 
-    final StringCache fStringCache = new StringCache();
-
     final HTMLConfiguration htmlConfiguration_;
 
     /**
@@ -1352,7 +1349,7 @@ else if (NAMES_LOWERCASE == mode && !Character.isLowerCase(c)) {
         }
 
         final int length = fCurrentEntity.offset_ - offset;
-        final String name = length > 0 ? fStringCache.get(fCurrentEntity.buffer_, offset, length) : null;
+        final String name = length > 0 ? new String(fCurrentEntity.buffer_, offset, length) : null;
         if (DEBUG_BUFFER) {
             fCurrentEntity.debugBufferIfNeeded(")scanName: ", " -> \"" + name + '"');
         }
@@ -1422,7 +1419,7 @@ else if (NAMES_LOWERCASE == fNamesElems && !Character.isLowerCase(c)) {
         }
 
         final int length = fCurrentEntity.offset_ - offset;
-        final String name = length > 0 ? fStringCache.get(fCurrentEntity.buffer_, offset, length) : null;
+        final String name = length > 0 ? new String(fCurrentEntity.buffer_, offset, length) : null;
         if (DEBUG_BUFFER) {
             fCurrentEntity.debugBufferIfNeeded(")scanName: ", " -> \"" + name + '"');
         }
@@ -1840,7 +1837,7 @@ int read() throws IOException {
          * @return the read string (length may be smaller if EOF is encountered)
          * @throws IOException in case of io problems
          */
-        String nextContent(final StringCache strCache, final int len) throws IOException {
+        String nextContent(final int len) throws IOException {
             final int originalOffset = offset_;
             final int originalColumnNumber = getColumnNumber();
             final int originalCharacterOffset = getCharacterOffset();
@@ -1867,7 +1864,7 @@ String nextContent(final StringCache strCache, final int len) throws IOException
             columnNumber_ = originalColumnNumber;
             characterOffset_ = originalCharacterOffset;
 
-            return strCache.get(buff, 0, nbRead);
+            return new String(buff, 0, nbRead);
         }
 
         // Reads a single character, preserving the old buffer content
@@ -2422,7 +2419,7 @@ private void scanUntilEndTag(final String tagNameWithLeadingSlash) throws IOExce
                     break;
                 }
                 if (c == '<') {
-                    final String next = fCurrentEntity.nextContent(fStringCache, lengthToScan) + " ";
+                    final String next = fCurrentEntity.nextContent(lengthToScan) + " ";
                     if (next.length() >= lengthToScan
                             && tagNameWithLeadingSlash.equalsIgnoreCase(
                                     next.substring(0, tagNameWithLeadingSlash.length()))
@@ -3639,7 +3636,7 @@ public int scan(final boolean complete) throws IOException {
                             state = ScanScriptState.ESCAPED;
                         }
                         else if (c == '<') {
-                            final String next = fCurrentEntity.nextContent(fStringCache, 8) + " ";
+                            final String next = fCurrentEntity.nextContent(8) + " ";
                             if (next.length() >= 8 && "/script".equalsIgnoreCase(next.substring(0, 7))
                                     && ('>' == next.charAt(7) || Character.isWhitespace(next.charAt(7)))) {
                                 fCurrentEntity.rewind();
@@ -3658,7 +3655,7 @@ else if (fScanScriptContent.endsWith("--!")) {
                             }
                         }
                         else if (c == '<') {
-                            final String next = fCurrentEntity.nextContent(fStringCache, 8) + " ";
+                            final String next = fCurrentEntity.nextContent(8) + " ";
                             if (next.length() >= 8 && "/script".equalsIgnoreCase(next.substring(0, 7))
                                     && ('>' == next.charAt(7) || Character.isWhitespace(next.charAt(7)))) {
                                 fCurrentEntity.rewind();

src/main/java/org/htmlunit/cyberneko/util/StringCache.java

@@ -209,7 +209,7 @@ class MyContentScanner extends HTMLScanner.ContentScanner {
             @Override
             protected int scanComment() throws IOException {
                 // bug was here: calling nextContent() at the end of the buffer/input
-                fCurrentEntity.nextContent(fStringCache, 30);
+                fCurrentEntity.nextContent(30);
                 return super.scanComment();
             }
         }