001: /*
002: * Copyright 1999-2004 The Apache Software Foundation.
003: *
004: * Licensed under the Apache License, Version 2.0 (the "License");
005: * you may not use this file except in compliance with the License.
006: * You may obtain a copy of the License at
007: *
008: * http://www.apache.org/licenses/LICENSE-2.0
009: *
010: * Unless required by applicable law or agreed to in writing, software
011: * distributed under the License is distributed on an "AS IS" BASIS,
012: * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
013: * See the License for the specific language governing permissions and
014: * limitations under the License.
015: */
016: /*
017: * $Id: XStringForFSB.java,v 1.24 2005/01/23 01:08:21 mcnamara Exp $
018: */
019: package org.apache.xpath.objects;
020:
021: import org.apache.xalan.res.XSLMessages;
022: import org.apache.xml.utils.FastStringBuffer;
023: import org.apache.xml.utils.XMLCharacterRecognizer;
024: import org.apache.xml.utils.XMLString;
025: import org.apache.xml.utils.XMLStringFactory;
026: import org.apache.xpath.res.XPATHErrorResources;
027:
028: /**
029: * This class will wrap a FastStringBuffer and allow for
030: */
031: public class XStringForFSB extends XString {
032: static final long serialVersionUID = -1533039186550674548L;
033:
034: /** The start position in the fsb. */
035: int m_start;
036:
037: /** The length of the string. */
038: int m_length;
039:
040: /** If the str() function is called, the string will be cached here. */
041: protected String m_strCache = null;
042:
043: /** cached hash code */
044: protected int m_hash = 0;
045:
046: /**
047: * Construct a XNodeSet object.
048: *
049: * @param val FastStringBuffer object this will wrap, must be non-null.
050: * @param start The start position in the array.
051: * @param length The number of characters to read from the array.
052: */
053: public XStringForFSB(FastStringBuffer val, int start, int length) {
054:
055: super (val);
056:
057: m_start = start;
058: m_length = length;
059:
060: if (null == val)
061: throw new IllegalArgumentException(
062: XSLMessages
063: .createXPATHMessage(
064: XPATHErrorResources.ER_FASTSTRINGBUFFER_CANNOT_BE_NULL,
065: null));
066: }
067:
068: /**
069: * Construct a XNodeSet object.
070: *
071: * @param val String object this will wrap.
072: */
073: private XStringForFSB(String val) {
074:
075: super (val);
076:
077: throw new IllegalArgumentException(XSLMessages
078: .createXPATHMessage(
079: XPATHErrorResources.ER_FSB_CANNOT_TAKE_STRING,
080: null)); // "XStringForFSB can not take a string for an argument!");
081: }
082:
083: /**
084: * Cast result object to a string.
085: *
086: * @return The string this wraps or the empty string if null
087: */
088: public FastStringBuffer fsb() {
089: return ((FastStringBuffer) m_obj);
090: }
091:
092: /**
093: * Cast result object to a string.
094: *
095: * @return The string this wraps or the empty string if null
096: */
097: public void appendToFsb(org.apache.xml.utils.FastStringBuffer fsb) {
098: // %OPT% !!! FSB has to be updated to take partial fsb's for append.
099: fsb.append(str());
100: }
101:
102: /**
103: * Tell if this object contains a java String object.
104: *
105: * @return true if this XMLString can return a string without creating one.
106: */
107: public boolean hasString() {
108: return (null != m_strCache);
109: }
110:
111: // /** NEEDSDOC Field strCount */
112: // public static int strCount = 0;
113: //
114: // /** NEEDSDOC Field xtable */
115: // static java.util.Hashtable xtable = new java.util.Hashtable();
116:
117: /**
118: * Since this object is incomplete without the length and the offset, we
119: * have to convert to a string when this function is called.
120: *
121: * @return The java String representation of this object.
122: */
123: public Object object() {
124: return str();
125: }
126:
127: /**
128: * Cast result object to a string.
129: *
130: * @return The string this wraps or the empty string if null
131: */
132: public String str() {
133:
134: if (null == m_strCache) {
135: m_strCache = fsb().getString(m_start, m_length);
136:
137: // strCount++;
138: //
139: // RuntimeException e = new RuntimeException("Bad! Bad!");
140: // java.io.CharArrayWriter writer = new java.io.CharArrayWriter();
141: // java.io.PrintWriter pw = new java.io.PrintWriter(writer);
142: //
143: // e.printStackTrace(pw);
144: //
145: // String str = writer.toString();
146: //
147: // str = str.substring(0, 600);
148: //
149: // if (null == xtable.get(str))
150: // {
151: // xtable.put(str, str);
152: // System.out.println(str);
153: // }
154: // System.out.println("strCount: " + strCount);
155:
156: // throw e;
157: // e.printStackTrace();
158: // System.exit(-1);
159: }
160:
161: return m_strCache;
162: }
163:
164: /**
165: * Directly call the
166: * characters method on the passed ContentHandler for the
167: * string-value. Multiple calls to the
168: * ContentHandler's characters methods may well occur for a single call to
169: * this method.
170: *
171: * @param ch A non-null reference to a ContentHandler.
172: *
173: * @throws org.xml.sax.SAXException
174: */
175: public void dispatchCharactersEvents(org.xml.sax.ContentHandler ch)
176: throws org.xml.sax.SAXException {
177: fsb().sendSAXcharacters(ch, m_start, m_length);
178: }
179:
180: /**
181: * Directly call the
182: * comment method on the passed LexicalHandler for the
183: * string-value.
184: *
185: * @param lh A non-null reference to a LexicalHandler.
186: *
187: * @throws org.xml.sax.SAXException
188: */
189: public void dispatchAsComment(org.xml.sax.ext.LexicalHandler lh)
190: throws org.xml.sax.SAXException {
191: fsb().sendSAXComment(lh, m_start, m_length);
192: }
193:
194: /**
195: * Returns the length of this string.
196: *
197: * @return the length of the sequence of characters represented by this
198: * object.
199: */
200: public int length() {
201: return m_length;
202: }
203:
204: /**
205: * Returns the character at the specified index. An index ranges
206: * from <code>0</code> to <code>length() - 1</code>. The first character
207: * of the sequence is at index <code>0</code>, the next at index
208: * <code>1</code>, and so on, as for array indexing.
209: *
210: * @param index the index of the character.
211: * @return the character at the specified index of this string.
212: * The first character is at index <code>0</code>.
213: * @exception IndexOutOfBoundsException if the <code>index</code>
214: * argument is negative or not less than the length of this
215: * string.
216: */
217: public char charAt(int index) {
218: return fsb().charAt(m_start + index);
219: }
220:
221: /**
222: * Copies characters from this string into the destination character
223: * array.
224: *
225: * @param srcBegin index of the first character in the string
226: * to copy.
227: * @param srcEnd index after the last character in the string
228: * to copy.
229: * @param dst the destination array.
230: * @param dstBegin the start offset in the destination array.
231: * @exception IndexOutOfBoundsException If any of the following
232: * is true:
233: * <ul><li><code>srcBegin</code> is negative.
234: * <li><code>srcBegin</code> is greater than <code>srcEnd</code>
235: * <li><code>srcEnd</code> is greater than the length of this
236: * string
237: * <li><code>dstBegin</code> is negative
238: * <li><code>dstBegin+(srcEnd-srcBegin)</code> is larger than
239: * <code>dst.length</code></ul>
240: * @exception NullPointerException if <code>dst</code> is <code>null</code>
241: */
242: public void getChars(int srcBegin, int srcEnd, char dst[],
243: int dstBegin) {
244:
245: // %OPT% Need to call this on FSB when it is implemented.
246: // %UNTESTED% (I don't think anyone calls this yet?)
247: int n = srcEnd - srcBegin;
248:
249: if (n > m_length)
250: n = m_length;
251:
252: if (n > (dst.length - dstBegin))
253: n = (dst.length - dstBegin);
254:
255: int end = srcBegin + m_start + n;
256: int d = dstBegin;
257: FastStringBuffer fsb = fsb();
258:
259: for (int i = srcBegin + m_start; i < end; i++) {
260: dst[d++] = fsb.charAt(i);
261: }
262: }
263:
264: /**
265: * Compares this string to the specified object.
266: * The result is <code>true</code> if and only if the argument is not
267: * <code>null</code> and is a <code>String</code> object that represents
268: * the same sequence of characters as this object.
269: *
270: * @param obj2 the object to compare this <code>String</code>
271: * against.
272: *
273: * @return <code>true</code> if the <code>String </code>are equal;
274: * <code>false</code> otherwise.
275: * @see java.lang.String#compareTo(java.lang.String)
276: * @see java.lang.String#equalsIgnoreCase(java.lang.String)
277: */
278: public boolean equals(XMLString obj2) {
279:
280: if (this == obj2) {
281: return true;
282: }
283:
284: int n = m_length;
285:
286: if (n == obj2.length()) {
287: FastStringBuffer fsb = fsb();
288: int i = m_start;
289: int j = 0;
290:
291: while (n-- != 0) {
292: if (fsb.charAt(i) != obj2.charAt(j)) {
293: return false;
294: }
295:
296: i++;
297: j++;
298: }
299:
300: return true;
301: }
302:
303: return false;
304: }
305:
306: /**
307: * Tell if two objects are functionally equal.
308: *
309: * @param obj2 Object to compare this to
310: *
311: * @return true if the two objects are equal
312: *
313: * @throws javax.xml.transform.TransformerException
314: */
315: public boolean equals(XObject obj2) {
316:
317: if (this == obj2) {
318: return true;
319: }
320: if (obj2.getType() == XObject.CLASS_NUMBER)
321: return obj2.equals(this );
322:
323: String str = obj2.str();
324: int n = m_length;
325:
326: if (n == str.length()) {
327: FastStringBuffer fsb = fsb();
328: int i = m_start;
329: int j = 0;
330:
331: while (n-- != 0) {
332: if (fsb.charAt(i) != str.charAt(j)) {
333: return false;
334: }
335:
336: i++;
337: j++;
338: }
339:
340: return true;
341: }
342:
343: return false;
344: }
345:
346: /**
347: * Tell if two objects are functionally equal.
348: *
349: * @param anotherString Object to compare this to
350: *
351: * @return true if the two objects are equal
352: *
353: * @throws javax.xml.transform.TransformerException
354: */
355: public boolean equals(String anotherString) {
356:
357: int n = m_length;
358:
359: if (n == anotherString.length()) {
360: FastStringBuffer fsb = fsb();
361: int i = m_start;
362: int j = 0;
363:
364: while (n-- != 0) {
365: if (fsb.charAt(i) != anotherString.charAt(j)) {
366: return false;
367: }
368:
369: i++;
370: j++;
371: }
372:
373: return true;
374: }
375:
376: return false;
377: }
378:
379: /**
380: * Compares this string to the specified object.
381: * The result is <code>true</code> if and only if the argument is not
382: * <code>null</code> and is a <code>String</code> object that represents
383: * the same sequence of characters as this object.
384: *
385: * @param obj2 the object to compare this <code>String</code>
386: * against.
387: *
388: * @return <code>true</code> if the <code>String </code>are equal;
389: * <code>false</code> otherwise.
390: * @see java.lang.String#compareTo(java.lang.String)
391: * @see java.lang.String#equalsIgnoreCase(java.lang.String)
392: */
393: public boolean equals(Object obj2) {
394:
395: if (null == obj2)
396: return false;
397:
398: if (obj2 instanceof XNumber)
399: return obj2.equals(this );
400:
401: // In order to handle the 'all' semantics of
402: // nodeset comparisons, we always call the
403: // nodeset function.
404: else if (obj2 instanceof XNodeSet)
405: return obj2.equals(this );
406: else if (obj2 instanceof XStringForFSB)
407: return equals((XMLString) obj2);
408: else
409: return equals(obj2.toString());
410: }
411:
412: /**
413: * Compares this <code>String</code> to another <code>String</code>,
414: * ignoring case considerations. Two strings are considered equal
415: * ignoring case if they are of the same length, and corresponding
416: * characters in the two strings are equal ignoring case.
417: *
418: * @param anotherString the <code>String</code> to compare this
419: * <code>String</code> against.
420: * @return <code>true</code> if the argument is not <code>null</code>
421: * and the <code>String</code>s are equal,
422: * ignoring case; <code>false</code> otherwise.
423: * @see #equals(Object)
424: * @see java.lang.Character#toLowerCase(char)
425: * @see java.lang.Character#toUpperCase(char)
426: */
427: public boolean equalsIgnoreCase(String anotherString) {
428: return (m_length == anotherString.length()) ? str()
429: .equalsIgnoreCase(anotherString) : false;
430: }
431:
432: /**
433: * Compares two strings lexicographically.
434: *
435: * @param xstr the <code>String</code> to be compared.
436: *
437: * @return the value <code>0</code> if the argument string is equal to
438: * this string; a value less than <code>0</code> if this string
439: * is lexicographically less than the string argument; and a
440: * value greater than <code>0</code> if this string is
441: * lexicographically greater than the string argument.
442: * @exception java.lang.NullPointerException if <code>anotherString</code>
443: * is <code>null</code>.
444: */
445: public int compareTo(XMLString xstr) {
446:
447: int len1 = m_length;
448: int len2 = xstr.length();
449: int n = Math.min(len1, len2);
450: FastStringBuffer fsb = fsb();
451: int i = m_start;
452: int j = 0;
453:
454: while (n-- != 0) {
455: char c1 = fsb.charAt(i);
456: char c2 = xstr.charAt(j);
457:
458: if (c1 != c2) {
459: return c1 - c2;
460: }
461:
462: i++;
463: j++;
464: }
465:
466: return len1 - len2;
467: }
468:
469: /**
470: * Compares two strings lexicographically, ignoring case considerations.
471: * This method returns an integer whose sign is that of
472: * <code>this.toUpperCase().toLowerCase().compareTo(
473: * str.toUpperCase().toLowerCase())</code>.
474: * <p>
475: * Note that this method does <em>not</em> take locale into account,
476: * and will result in an unsatisfactory ordering for certain locales.
477: * The java.text package provides <em>collators</em> to allow
478: * locale-sensitive ordering.
479: *
480: * @param xstr the <code>String</code> to be compared.
481: *
482: * @return a negative integer, zero, or a positive integer as the
483: * the specified String is greater than, equal to, or less
484: * than this String, ignoring case considerations.
485: * @see java.text.Collator#compare(String, String)
486: * @since 1.2
487: */
488: public int compareToIgnoreCase(XMLString xstr) {
489:
490: int len1 = m_length;
491: int len2 = xstr.length();
492: int n = Math.min(len1, len2);
493: FastStringBuffer fsb = fsb();
494: int i = m_start;
495: int j = 0;
496:
497: while (n-- != 0) {
498: char c1 = Character.toLowerCase(fsb.charAt(i));
499: char c2 = Character.toLowerCase(xstr.charAt(j));
500:
501: if (c1 != c2) {
502: return c1 - c2;
503: }
504:
505: i++;
506: j++;
507: }
508:
509: return len1 - len2;
510: }
511:
512: /**
513: * Returns a hashcode for this string. The hashcode for a
514: * <code>String</code> object is computed as
515: * <blockquote><pre>
516: * s[0]*31^(n-1) + s[1]*31^(n-2) + ... + s[n-1]
517: * </pre></blockquote>
518: * using <code>int</code> arithmetic, where <code>s[i]</code> is the
519: * <i>i</i>th character of the string, <code>n</code> is the length of
520: * the string, and <code>^</code> indicates exponentiation.
521: * (The hash value of the empty string is zero.)
522: *
523: * @return a hash code value for this object.
524: */
525: public int hashCode() {
526: // Commenting this out because in JDK1.1.8 and VJ++
527: // we don't match XMLStrings. Defaulting to the super
528: // causes us to create a string, but at this point
529: // this only seems to get called in key processing.
530: // Maybe we can live with it?
531:
532: /*
533: int h = m_hash;
534:
535: if (h == 0)
536: {
537: int off = m_start;
538: int len = m_length;
539: FastStringBuffer fsb = fsb();
540:
541: for (int i = 0; i < len; i++)
542: {
543: h = 31 * h + fsb.charAt(off);
544:
545: off++;
546: }
547:
548: m_hash = h;
549: }
550: */
551:
552: return super .hashCode(); // h;
553: }
554:
555: /**
556: * Tests if this string starts with the specified prefix beginning
557: * a specified index.
558: *
559: * @param prefix the prefix.
560: * @param toffset where to begin looking in the string.
561: * @return <code>true</code> if the character sequence represented by the
562: * argument is a prefix of the substring of this object starting
563: * at index <code>toffset</code>; <code>false</code> otherwise.
564: * The result is <code>false</code> if <code>toffset</code> is
565: * negative or greater than the length of this
566: * <code>String</code> object; otherwise the result is the same
567: * as the result of the expression
568: * <pre>
569: * this.subString(toffset).startsWith(prefix)
570: * </pre>
571: * @exception java.lang.NullPointerException if <code>prefix</code> is
572: * <code>null</code>.
573: */
574: public boolean startsWith(XMLString prefix, int toffset) {
575:
576: FastStringBuffer fsb = fsb();
577: int to = m_start + toffset;
578: int tlim = m_start + m_length;
579: int po = 0;
580: int pc = prefix.length();
581:
582: // Note: toffset might be near -1>>>1.
583: if ((toffset < 0) || (toffset > m_length - pc)) {
584: return false;
585: }
586:
587: while (--pc >= 0) {
588: if (fsb.charAt(to) != prefix.charAt(po)) {
589: return false;
590: }
591:
592: to++;
593: po++;
594: }
595:
596: return true;
597: }
598:
599: /**
600: * Tests if this string starts with the specified prefix.
601: *
602: * @param prefix the prefix.
603: * @return <code>true</code> if the character sequence represented by the
604: * argument is a prefix of the character sequence represented by
605: * this string; <code>false</code> otherwise.
606: * Note also that <code>true</code> will be returned if the
607: * argument is an empty string or is equal to this
608: * <code>String</code> object as determined by the
609: * {@link #equals(Object)} method.
610: * @exception java.lang.NullPointerException if <code>prefix</code> is
611: * <code>null</code>.
612: * @since JDK1. 0
613: */
614: public boolean startsWith(XMLString prefix) {
615: return startsWith(prefix, 0);
616: }
617:
618: /**
619: * Returns the index within this string of the first occurrence of the
620: * specified character. If a character with value <code>ch</code> occurs
621: * in the character sequence represented by this <code>String</code>
622: * object, then the index of the first such occurrence is returned --
623: * that is, the smallest value <i>k</i> such that:
624: * <blockquote><pre>
625: * this.charAt(<i>k</i>) == ch
626: * </pre></blockquote>
627: * is <code>true</code>. If no such character occurs in this string,
628: * then <code>-1</code> is returned.
629: *
630: * @param ch a character.
631: * @return the index of the first occurrence of the character in the
632: * character sequence represented by this object, or
633: * <code>-1</code> if the character does not occur.
634: */
635: public int indexOf(int ch) {
636: return indexOf(ch, 0);
637: }
638:
639: /**
640: * Returns the index within this string of the first occurrence of the
641: * specified character, starting the search at the specified index.
642: * <p>
643: * If a character with value <code>ch</code> occurs in the character
644: * sequence represented by this <code>String</code> object at an index
645: * no smaller than <code>fromIndex</code>, then the index of the first
646: * such occurrence is returned--that is, the smallest value <i>k</i>
647: * such that:
648: * <blockquote><pre>
649: * (this.charAt(<i>k</i>) == ch) && (<i>k</i> >= fromIndex)
650: * </pre></blockquote>
651: * is true. If no such character occurs in this string at or after
652: * position <code>fromIndex</code>, then <code>-1</code> is returned.
653: * <p>
654: * There is no restriction on the value of <code>fromIndex</code>. If it
655: * is negative, it has the same effect as if it were zero: this entire
656: * string may be searched. If it is greater than the length of this
657: * string, it has the same effect as if it were equal to the length of
658: * this string: <code>-1</code> is returned.
659: *
660: * @param ch a character.
661: * @param fromIndex the index to start the search from.
662: * @return the index of the first occurrence of the character in the
663: * character sequence represented by this object that is greater
664: * than or equal to <code>fromIndex</code>, or <code>-1</code>
665: * if the character does not occur.
666: */
667: public int indexOf(int ch, int fromIndex) {
668:
669: int max = m_start + m_length;
670: FastStringBuffer fsb = fsb();
671:
672: if (fromIndex < 0) {
673: fromIndex = 0;
674: } else if (fromIndex >= m_length) {
675:
676: // Note: fromIndex might be near -1>>>1.
677: return -1;
678: }
679:
680: for (int i = m_start + fromIndex; i < max; i++) {
681: if (fsb.charAt(i) == ch) {
682: return i - m_start;
683: }
684: }
685:
686: return -1;
687: }
688:
689: /**
690: * Returns a new string that is a substring of this string. The
691: * substring begins with the character at the specified index and
692: * extends to the end of this string. <p>
693: * Examples:
694: * <blockquote><pre>
695: * "unhappy".substring(2) returns "happy"
696: * "Harbison".substring(3) returns "bison"
697: * "emptiness".substring(9) returns "" (an empty string)
698: * </pre></blockquote>
699: *
700: * @param beginIndex the beginning index, inclusive.
701: * @return the specified substring.
702: * @exception IndexOutOfBoundsException if
703: * <code>beginIndex</code> is negative or larger than the
704: * length of this <code>String</code> object.
705: */
706: public XMLString substring(int beginIndex) {
707:
708: int len = m_length - beginIndex;
709:
710: if (len <= 0)
711: return XString.EMPTYSTRING;
712: else {
713: int start = m_start + beginIndex;
714:
715: return new XStringForFSB(fsb(), start, len);
716: }
717: }
718:
719: /**
720: * Returns a new string that is a substring of this string. The
721: * substring begins at the specified <code>beginIndex</code> and
722: * extends to the character at index <code>endIndex - 1</code>.
723: * Thus the length of the substring is <code>endIndex-beginIndex</code>.
724: *
725: * @param beginIndex the beginning index, inclusive.
726: * @param endIndex the ending index, exclusive.
727: * @return the specified substring.
728: * @exception IndexOutOfBoundsException if the
729: * <code>beginIndex</code> is negative, or
730: * <code>endIndex</code> is larger than the length of
731: * this <code>String</code> object, or
732: * <code>beginIndex</code> is larger than
733: * <code>endIndex</code>.
734: */
735: public XMLString substring(int beginIndex, int endIndex) {
736:
737: int len = endIndex - beginIndex;
738:
739: if (len > m_length)
740: len = m_length;
741:
742: if (len <= 0)
743: return XString.EMPTYSTRING;
744: else {
745: int start = m_start + beginIndex;
746:
747: return new XStringForFSB(fsb(), start, len);
748: }
749: }
750:
751: /**
752: * Concatenates the specified string to the end of this string.
753: *
754: * @param str the <code>String</code> that is concatenated to the end
755: * of this <code>String</code>.
756: * @return a string that represents the concatenation of this object's
757: * characters followed by the string argument's characters.
758: * @exception java.lang.NullPointerException if <code>str</code> is
759: * <code>null</code>.
760: */
761: public XMLString concat(String str) {
762:
763: // %OPT% Make an FSB here?
764: return new XString(str().concat(str));
765: }
766:
767: /**
768: * Removes white space from both ends of this string.
769: *
770: * @return this string, with white space removed from the front and end.
771: */
772: public XMLString trim() {
773: return fixWhiteSpace(true, true, false);
774: }
775:
776: /**
777: * Returns whether the specified <var>ch</var> conforms to the XML 1.0 definition
778: * of whitespace. Refer to <A href="http://www.w3.org/TR/1998/REC-xml-19980210#NT-S">
779: * the definition of <CODE>S</CODE></A> for details.
780: * @param ch Character to check as XML whitespace.
781: * @return =true if <var>ch</var> is XML whitespace; otherwise =false.
782: */
783: private static boolean isSpace(char ch) {
784: return XMLCharacterRecognizer.isWhiteSpace(ch); // Take the easy way out for now.
785: }
786:
787: /**
788: * Conditionally trim all leading and trailing whitespace in the specified String.
789: * All strings of white space are
790: * replaced by a single space character (#x20), except spaces after punctuation which
791: * receive double spaces if doublePunctuationSpaces is true.
792: * This function may be useful to a formatter, but to get first class
793: * results, the formatter should probably do it's own white space handling
794: * based on the semantics of the formatting object.
795: *
796: * @param trimHead Trim leading whitespace?
797: * @param trimTail Trim trailing whitespace?
798: * @param doublePunctuationSpaces Use double spaces for punctuation?
799: * @return The trimmed string.
800: */
801: public XMLString fixWhiteSpace(boolean trimHead, boolean trimTail,
802: boolean doublePunctuationSpaces) {
803:
804: int end = m_length + m_start;
805: char[] buf = new char[m_length];
806: FastStringBuffer fsb = fsb();
807: boolean edit = false;
808:
809: /* replace S to ' '. and ' '+ -> single ' '. */
810: int d = 0;
811: boolean pres = false;
812:
813: for (int s = m_start; s < end; s++) {
814: char c = fsb.charAt(s);
815:
816: if (isSpace(c)) {
817: if (!pres) {
818: if (' ' != c) {
819: edit = true;
820: }
821:
822: buf[d++] = ' ';
823:
824: if (doublePunctuationSpaces && (d != 0)) {
825: char prevChar = buf[d - 1];
826:
827: if (!((prevChar == '.') || (prevChar == '!') || (prevChar == '?'))) {
828: pres = true;
829: }
830: } else {
831: pres = true;
832: }
833: } else {
834: edit = true;
835: pres = true;
836: }
837: } else {
838: buf[d++] = c;
839: pres = false;
840: }
841: }
842:
843: if (trimTail && 1 <= d && ' ' == buf[d - 1]) {
844: edit = true;
845:
846: d--;
847: }
848:
849: int start = 0;
850:
851: if (trimHead && 0 < d && ' ' == buf[0]) {
852: edit = true;
853:
854: start++;
855: }
856:
857: XMLStringFactory xsf = XMLStringFactoryImpl.getFactory();
858:
859: return edit ? xsf.newstr(buf, start, d - start) : this ;
860: }
861:
862: /**
863: * Convert a string to a double -- Allowed input is in fixed
864: * notation ddd.fff.
865: *
866: * %OPT% CHECK PERFORMANCE against generating a Java String and
867: * converting it to double. The advantage of running in native
868: * machine code -- perhaps even microcode, on some systems -- may
869: * more than make up for the cost of allocating and discarding the
870: * additional object. We need to benchmark this.
871: *
872: * %OPT% More importantly, we need to decide whether we _care_ about
873: * the performance of this operation. Does XString.toDouble constitute
874: * any measurable percentage of our typical runtime? I suspect not!
875: *
876: * @return A double value representation of the string, or return Double.NaN
877: * if the string can not be converted. */
878: public double toDouble() {
879: if (m_length == 0)
880: return Double.NaN;
881: int i;
882: char c;
883: String valueString = fsb().getString(m_start, m_length);
884:
885: // The following are permitted in the Double.valueOf, but not by the XPath spec:
886: // - a plus sign
887: // - The use of e or E to indicate exponents
888: // - trailing f, F, d, or D
889: // See function comments; not sure if this is slower than actually doing the
890: // conversion ourselves (as was before).
891:
892: for (i = 0; i < m_length; i++)
893: if (!XMLCharacterRecognizer.isWhiteSpace(valueString
894: .charAt(i)))
895: break;
896: if (i == m_length)
897: return Double.NaN;
898: if (valueString.charAt(i) == '-')
899: i++;
900: for (; i < m_length; i++) {
901: c = valueString.charAt(i);
902: if (c != '.' && (c < '0' || c > '9'))
903: break;
904: }
905: for (; i < m_length; i++)
906: if (!XMLCharacterRecognizer.isWhiteSpace(valueString
907: .charAt(i)))
908: break;
909: if (i != m_length)
910: return Double.NaN;
911:
912: try {
913: return new Double(valueString).doubleValue();
914: } catch (NumberFormatException nfe) {
915: // This should catch double periods, empty strings.
916: return Double.NaN;
917: }
918: }
919: }
|