ohair@286: /*
mkos@397:  * Copyright (c) 2007, 2013, Oracle and/or its affiliates. All rights reserved.
ohair@286:  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
ohair@286:  *
ohair@286:  * This code is free software; you can redistribute it and/or modify it
ohair@286:  * under the terms of the GNU General Public License version 2 only, as
ohair@286:  * published by the Free Software Foundation.  Oracle designates this
ohair@286:  * particular file as subject to the "Classpath" exception as provided
ohair@286:  * by Oracle in the LICENSE file that accompanied this code.
ohair@286:  *
ohair@286:  * This code is distributed in the hope that it will be useful, but WITHOUT
ohair@286:  * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
ohair@286:  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
ohair@286:  * version 2 for more details (a copy is included in the LICENSE file that
ohair@286:  * accompanied this code).
ohair@286:  *
ohair@286:  * You should have received a copy of the GNU General Public License version
ohair@286:  * 2 along with this work; if not, write to the Free Software Foundation,
ohair@286:  * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
ohair@286:  *
ohair@286:  * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
ohair@286:  * or visit www.oracle.com if you need additional information or have any
ohair@286:  * questions.
ohair@286:  */
ohair@286: 
ohair@286: package javax.xml.bind;
ohair@286: 
ohair@286: /**
ohair@286:  * Processes white space normalization.
ohair@286:  *
ohair@286:  * @since 1.0
ohair@286:  */
ohair@286: abstract class WhiteSpaceProcessor {
ohair@286: 
ohair@286: // benchmarking (see test/src/ReplaceTest.java in the CVS Attic)
ohair@286: // showed that this code is slower than the current code.
ohair@286: //
ohair@286: //    public static String replace(String text) {
ohair@286: //        final int len = text.length();
ohair@286: //        StringBuffer result = new StringBuffer(len);
ohair@286: //
ohair@286: //        for (int i = 0; i < len; i++) {
ohair@286: //            char ch = text.charAt(i);
ohair@286: //            if (isWhiteSpace(ch))
ohair@286: //                result.append(' ');
ohair@286: //            else
ohair@286: //                result.append(ch);
ohair@286: //        }
ohair@286: //
ohair@286: //        return result.toString();
ohair@286: //    }
ohair@286: 
ohair@286:     public static String replace(String text) {
ohair@286:         return replace( (CharSequence)text ).toString();
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * @since 2.0
ohair@286:      */
ohair@286:     public static CharSequence replace(CharSequence text) {
ohair@286:         int i=text.length()-1;
ohair@286: 
ohair@286:         // look for the first whitespace char.
ohair@286:         while( i>=0 && !isWhiteSpaceExceptSpace(text.charAt(i)) )
ohair@286:             i--;
ohair@286: 
ohair@286:         if( i<0 )
ohair@286:             // no such whitespace. replace(text)==text.
ohair@286:             return text;
ohair@286: 
ohair@286:         // we now know that we need to modify the text.
ohair@286:         // allocate a char array to do it.
ohair@286:         StringBuilder buf = new StringBuilder(text);
ohair@286: 
ohair@286:         buf.setCharAt(i--,' ');
ohair@286:         for( ; i>=0; i-- )
ohair@286:             if( isWhiteSpaceExceptSpace(buf.charAt(i)))
ohair@286:                 buf.setCharAt(i,' ');
ohair@286: 
ohair@286:         return new String(buf);
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * Equivalent of {@link String#trim()}.
ohair@286:      * @since 2.0
ohair@286:      */
ohair@286:     public static CharSequence trim(CharSequence text) {
ohair@286:         int len = text.length();
ohair@286:         int start = 0;
ohair@286: 
ohair@286:         while( start<len && isWhiteSpace(text.charAt(start)) )
ohair@286:             start++;
ohair@286: 
ohair@286:         int end = len-1;
ohair@286: 
ohair@286:         while( end>start && isWhiteSpace(text.charAt(end)) )
ohair@286:             end--;
ohair@286: 
ohair@286:         if(start==0 && end==len-1)
ohair@286:             return text;    // no change
ohair@286:         else
ohair@286:             return text.subSequence(start,end+1);
ohair@286:     }
ohair@286: 
ohair@286:     public static String collapse(String text) {
ohair@286:         return collapse( (CharSequence)text ).toString();
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * This is usually the biggest processing bottleneck.
ohair@286:      *
ohair@286:      * @since 2.0
ohair@286:      */
ohair@286:     public static CharSequence collapse(CharSequence text) {
ohair@286:         int len = text.length();
ohair@286: 
ohair@286:         // most of the texts are already in the collapsed form.
ohair@286:         // so look for the first whitespace in the hope that we will
ohair@286:         // never see it.
ohair@286:         int s=0;
ohair@286:         while(s<len) {
ohair@286:             if(isWhiteSpace(text.charAt(s)))
ohair@286:                 break;
ohair@286:             s++;
ohair@286:         }
ohair@286:         if(s==len)
ohair@286:             // the input happens to be already collapsed.
ohair@286:             return text;
ohair@286: 
ohair@286:         // we now know that the input contains spaces.
ohair@286:         // let's sit down and do the collapsing normally.
ohair@286: 
ohair@286:         StringBuilder result = new StringBuilder(len /*allocate enough size to avoid re-allocation*/ );
ohair@286: 
ohair@286:         if(s!=0) {
ohair@286:             for( int i=0; i<s; i++ )
ohair@286:                 result.append(text.charAt(i));
ohair@286:             result.append(' ');
ohair@286:         }
ohair@286: 
ohair@286:         boolean inStripMode = true;
ohair@286:         for (int i = s+1; i < len; i++) {
ohair@286:             char ch = text.charAt(i);
ohair@286:             boolean b = isWhiteSpace(ch);
ohair@286:             if (inStripMode && b)
ohair@286:                 continue; // skip this character
ohair@286: 
ohair@286:             inStripMode = b;
ohair@286:             if (inStripMode)
ohair@286:                 result.append(' ');
ohair@286:             else
ohair@286:                 result.append(ch);
ohair@286:         }
ohair@286: 
ohair@286:         // remove trailing whitespaces
ohair@286:         len = result.length();
ohair@286:         if (len > 0 && result.charAt(len - 1) == ' ')
ohair@286:             result.setLength(len - 1);
ohair@286:         // whitespaces are already collapsed,
ohair@286:         // so all we have to do is to remove the last one character
ohair@286:         // if it's a whitespace.
ohair@286: 
ohair@286:         return result;
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * Returns true if the specified string is all whitespace.
ohair@286:      */
ohair@286:     public static final boolean isWhiteSpace(CharSequence s) {
ohair@286:         for( int i=s.length()-1; i>=0; i-- )
ohair@286:             if(!isWhiteSpace(s.charAt(i)))
ohair@286:                 return false;
ohair@286:         return true;
ohair@286:     }
ohair@286: 
ohair@286:     /** returns true if the specified char is a white space character. */
ohair@286:     public static final boolean isWhiteSpace(char ch) {
ohair@286:         // most of the characters are non-control characters.
ohair@286:         // so check that first to quickly return false for most of the cases.
ohair@286:         if( ch>0x20 )   return false;
ohair@286: 
ohair@286:         // other than we have to do four comparisons.
ohair@286:         return ch == 0x9 || ch == 0xA || ch == 0xD || ch == 0x20;
ohair@286:     }
ohair@286: 
ohair@286:     /**
ohair@286:      * Returns true if the specified char is a white space character
ohair@286:      * but not 0x20.
ohair@286:      */
ohair@286:     protected static final boolean isWhiteSpaceExceptSpace(char ch) {
ohair@286:         // most of the characters are non-control characters.
ohair@286:         // so check that first to quickly return false for most of the cases.
ohair@286:         if( ch>=0x20 )   return false;
ohair@286: 
ohair@286:         // other than we have to do four comparisons.
ohair@286:         return ch == 0x9 || ch == 0xA || ch == 0xD;
ohair@286:     }
ohair@286: }