1 // Copyright (C) 2003 Adam Megacz <adam@ibex.org> all rights reserved.
3 // You may modify, copy, and redistribute this code under the terms of
4 // the GNU Library Public License version 2.1, with the exception of
5 // the portion of clause 6a after the semicolon (aka the "obnoxious
10 import java.io.Reader;
11 import java.io.Writer;
12 import java.io.IOException;
13 import java.io.EOFException;
16 * An Event-Driving, Non-Validating XML Parser with Namespace support.
18 * A subclass can implement the abstract functions for receiving details
19 * about an xml file as it is parsed. To initate a parse, use the parse()
22 * <h3>Implementation Notes</h3>
23 * <p>As the parser traverses into an element, it adds it to the linked list
24 * called <tt>elements</tt>. However, <tt>elements</tt> has been pre-filled
25 * with instances of the Element inner class. So in the vast majority of
26 * cases, the pointer current is moved along one, and the values for the
27 * new element are filled into the current object.</p>
29 * <p>This parser supports all the unicode ranges required by the XML
30 * Specification. However, it is optimised for well-formed ASCII documents.
31 * Documents containing unicode Names and Attributes will take much longer
32 * to process, and invalid documents (badly formed Names or invalid attributes)
33 * will be run through a test on every single unicode character range before
34 * being declared invalid.</p>
37 * <li>Each time the buffer offset <tt>off</tt> is moved, the length
38 * <tt>len</tt> must be decreased.</li>
39 * <li>Each time the buffer length is decreased, it must be checked to make
40 * sure it is >0.</li>
41 * <li><i>error</i> is defined as a Validity Constraint Violation and
43 * <li><i>fatal error</i> is defined as a Well-formedness Constraint
44 * Violation and is not recoverable</li>
47 * @author David Crawshaw
48 * @see <a href="http://w3.org/TR/REC-xml">XML Specification</a>
49 * @see <a href="http://w3.org/TR/REC-xml-names">XML Namespaces</a>
51 public abstract class XML {
53 /////////////////////////////////////////////////////////////////////////////////////////////
55 /////////////////////////////////////////////////////////////////////////////////////////////
57 public static final int BUFFER_SIZE = 255;
59 /** static pool of XML.Element instances shared by all XML Parsers. */
60 private static final Queue elements = new Queue(30);
62 private static final char[] single_amp = new char[] { '&' };
63 private static final char[] single_apos = new char[] { '\'' };
64 private static final char[] single_gt = new char[] { '>' };
65 private static final char[] single_lt = new char[] { '<' };
66 private static final char[] single_quot = new char[] { '"' };
74 int base; // base+off == distance into the stream
77 Element current = null;
79 // used in readEntity() to process a single character without creating a new array
80 private char[] singlechar = new char[1];
83 public XML() { this(BUFFER_SIZE); }
85 public XML(int bSize) {
86 buf = new char[bSize];
87 //current = (Element)elements.remove(false);
88 if (current == null) current = newElement();
91 /** Returns the line number at the beginning of the last process call. */
92 public int getLine() { return line; }
94 /** Returns the column number at the beginning of the last process call. */
95 public int getCol() { return col; }
97 /** Returns the global file offset at the beginning of the last process call. */
98 public int getGlobalOffset() { return base + off; }
100 Element newElement() { return new Element(); }
103 * Parse given input and call the abstract event functions.
105 * Careful with threading, as this function is not synchronized.
107 public final void parse(Reader reader) throws IOException, Exn {
112 clear(); // clean up possible mid-way linked-list element
115 // process the stream
118 if (current.qName == null) break;
119 throw new Exn("reached eof without closing <"+current.qName+"> element", Exn.WFC, getLine(), getCol());
122 if (buf[off] == '<') readTag();
123 readChars(current.qName != null);
125 } finally { clear(); } // clean up elements
128 // Stuff below here is Adam's hack //////////////////////////////////////////////////////////////////////////////
130 boolean done = false;
131 public static class Pull extends XML {
132 public Pull(Reader in) { this.in = in; off = len = 0; line = col = 1; clear(); }
133 StringBuffer sb = new StringBuffer();
134 Element pending = null;
135 boolean emptytag = true;
136 public int level = 0;
137 public final void startElement(Element e) throws Exn { emptytag = false; level++; pending = e; }
138 public final void endElement(Element e) throws Exn, IOException { emptytag=pending!=null; level--; }
139 public final void whitespace(char[] ch, int start, int length) throws Exn, IOException { }
140 public final void characters(char[] ch, int start, int length) throws Exn, IOException {
141 emptytag=false; sb.append(ch,start,length);}
142 public Object read() throws Exn, IOException {
144 if (pending != null) { Element ret = pending; pending = null; ret.level = level-(emptytag?0:1); return ret; }
145 if (sb.length() > 0) { String ret = sb.toString(); sb.setLength(0); return ret; }
147 if (done) return null;
148 throw new Exn("reached eof without closing <"+current.qName+"> element", Exn.WFC, getLine(), getCol());
150 if (buf[off] == '<') readTag(); else readChars(!done);
157 // Stuff above here is Adam's hack //////////////////////////////////////////////////////////////////////////////
159 /** remove any leftover elements from the linked list and queue them */
161 for (Element last = current; current.parent != null; ) {
162 current = current.parent;
164 elements.append(last);
169 /** reads in a tag. expects <tt>buf[off] == '<'</tt> */
170 final void readTag() throws IOException, Exn {
171 // Start Tag '<' Name (S Attribute)* S? '>'
172 boolean starttag = true;
174 // End Tag '</' Name S? '>'
175 boolean endtag = false;
177 // if (starttag & endtag) then: EmptyElemTag '<' Name (S Attribute)* S? '/>'
179 // Position in the name of the ':' namespace prefix
185 if (!buffer(1)) throw new EOFException("Unexpected EOF processing element tag");
187 // work out what we can from the beginning of the tag
190 // definitions here don't necessarily conform to xml spec (as DTDs not yet implemented)
192 if (!buffer(4)) throw new EOFException("Unexpected EOF processing <! element");
197 if (buf[off+1] != '-') { bad = true; break; }
198 col += 2; off += 2; len -= 2;
200 // Comment '<!--' ((Char - '-') | ('-' (Char - '-')))* '-->'
201 readChars(false, "-->", false);
202 col += 3; off += 3; len -= 3;
205 // we don't care about the following definitions
209 || buf[off+1] != 'T' || buf[off+2] != 'T' || buf[off+3] != 'L'
210 || buf[off+4] != 'I' || buf[off+5] != 'S' || buf[off+6] != 'T') {
213 col += 7; off += 7; len -= 7;
215 // ATTLIST '<!ATTLIST' (Char* - '>') '>'
216 readChars(false, ">", true);
221 || buf[off+1] != 'O' || buf[off+2] != 'C' || buf[off+3] != 'T'
222 || buf[off+4] != 'Y' || buf[off+5] != 'P' || buf[off+6] != 'E') {
225 col += 7; off += 7; len -= 7;
227 // DTD '<!DOCTYPE' (Char* - '>') '>'
228 readChars(false, ">", true);
234 } else if (buf[off+1] == 'L' && buf[off+2] == 'E' && buf[off+3] == 'M'
235 && buf[off+4] == 'E' && buf[off+5] == 'N' && buf[off+6] == 'T') {
236 // ELEMENT '<!ELEMENT' (Char* - '>') '>'
237 readChars(false, ">", true);
240 } else if (buf[off+1] == 'N' && buf[off+2] == 'T' && buf[off+3] == 'I'
241 && buf[off+4] == 'T' && buf[off+5] == 'Y') {
242 // ENTITY '<!ENTITY' (Char* - '>') '>'
243 readChars(false, ">", true);
253 || buf[off+1] != 'O' || buf[off+2] != 'T' || buf[off+3] != 'A' || buf[off+4] != 'T'
254 || buf[off+5] != 'I' || buf[off+6] != 'O' || buf[off+7] != 'N') {
257 col += 8; off += 8; len -= 8;
258 // NOTATION '<!NOTATION' (Char* - '>') '>'
259 readChars(false, ">", true);
266 if (bad) throw new Exn("element tag start character is invalid", Exn.MARKUP, getLine(), getCol());
268 } else if (s == '?') {
269 // PI (Ignored) '<?' (Char* - (Char* '?>' Char*)) '?>'
271 readChars(false, "?>", true);
272 if (!buffer(2)) throw new EOFException("Unexpected EOF at end of Processing Instruction");
273 col += 2; off += 2; len -= 2;
275 } else if (s == '[') {
277 || buf[off+1] != 'C' || buf[off+2] != 'D' || buf[off+3] != 'A'
278 || buf[off+4] != 'T' || buf[off+5] != 'A' || buf[off+6] != '[') {
280 // Conditional '<![' (Char* - (Char* ']]>' Char*)) ']]>'
281 readChars(false, "]]>", false);
283 col += 7; off += 7; len -=7;
284 // CDATA '<![CDATA[' (Char* - (Char* ']]>' Char*)) ']]>'
285 readChars(true, "]]>", false);
287 col += 3; off += 3; len -= 3;
290 // End Tag '</' Name S? '>'
295 if (!buffer(1)) throw new EOFException("Unexpected EOF processing end tag");
299 if (!Name(s)) throw new Exn("invalid starting character in element name", Exn.MARKUP, getLine(), getCol());
301 // find the element name (defined in XML Spec: section 2.3)
302 for (namelen = 0; ; namelen++) {
303 if (!buffer(namelen+1)) throw new EOFException("Unexpected EOF in element tag name");
305 s = buf[off+namelen];
307 if (S(s) || s == '>') {
309 } else if (s == '/') {
312 } else if (s == ':' && namelen > 0 && prefix < 1) {
313 // we have a definition of the prefix range available
315 } else if (!NameChar(s)) {
316 throw new Exn("element name contains invalid character", Exn.MARKUP, getLine(), getCol());
320 // process name (based on calculated region)
321 if (namelen < 1) throw new Exn("element name is null", Exn.MARKUP, getLine(), getCol());
323 // we have marked out the name region, so turn it into a string and move on
324 String qName = new String(buf, off, namelen);
326 col += namelen; off += namelen; len -= namelen;
329 // create the in-memory element representation of this beast
330 // if current.qName == null then this is the root element we're dealing with
331 if (current.qName != null) {
332 Element next = newElement();
333 //next.clear(); // TODO: remove as elements now checked as they're added to the queue
334 next.parent = current;
338 current.qName = qName;
341 current.prefix = current.qName.substring(0, prefix);
342 current.localName = current.qName.substring(prefix+1);
344 current.prefix = null;
345 current.localName = current.qName;
348 // process attributes
350 if (!buffer(1)) throw new EOFException("Unexpected EOF - processing attributes part 1");
351 while (buf[off] != '/' && buf[off] != '>') {
353 if (!buffer(1)) throw new EOFException("Unexpected EOF - processing attributes part 2");
357 // work out the uri of this element
358 current.uri = current.getUri(current.getPrefix());
359 if (current.getUri().equals("") && current.getPrefix() != null)
360 current.addError(new Exn("undefined prefix '"+current.getPrefix()+"'", Exn.NC, getLine(), getCol()));
363 // this is an end-of-element tag
364 if (!qName.equals(current.getQName())) throw new Exn(
365 "end tag </"+qName+"> does not line up with start tag <"+current.getQName()+">", Exn.WFC, getLine(), getCol()
369 // deal with whitespace
373 if (!buffer(1)) throw new EOFException("Unexpected EOF before end of tag");
374 if (buf[off] == '/') {
378 if (!buffer(1)) throw new EOFException("Unexpected EOF before end of endtag");
379 if (buf[off] == '>') {
382 throw new Exn("missing '>' character from element '"+qName+"'", Exn.MARKUP, getLine(), getCol());
385 // send element signals
386 if (starttag) startElement(current);
390 // we just closed an element, so remove it from the element 'stack'
391 if (current.getParent() == null) {
392 // we just finished the root element
395 Element last = current;
396 current = current.parent;
397 //last.clear(); FIXME
398 elements.append(last);
405 /** reads in an attribute of an element. expects Name(buf[off]) */
406 private final void readAttribute() throws IOException, Exn {
409 String n, v, p, u; // attribute name, value, prefix and uri respectively
410 n = v = p = u = null;
413 // find the element name (defined in XML Spec: section 2.3)
414 for (ref= 0; ; ref++) {
415 if (!buffer(ref+1)) throw new EOFException("Unexpected EOF in read attribute loop part 1");
419 if (s == '=' || S(s)) {
421 } else if (s == ':' && ref > 0 && prefix < 1) {
422 // we have a definition of the prefix range available
424 } else if (!NameChar(s)) {
425 throw new Exn("attribute name contains invalid characters", Exn.MARKUP, getLine(), getCol());
429 // determine prefix and key name
431 p = new String(buf, off, prefix-1);
432 col += prefix; off += prefix; len -= prefix; ref -= prefix;
434 n = new String(buf, off, ref);
435 col += ref; off += ref; len -= ref;
437 // find name/value divider ('=')
439 if (!buffer(1)) throw new EOFException("Unexpected EOF before attribute '=' divider");
440 if (buf[off] != '=') throw new Exn("attribute name not followed by '=' sign", Exn.MARKUP, getLine(), getCol());
445 if (!buffer(1)) throw new EOFException("Unexpected EOF after attribute '=' divider");
448 if (buf[off] == '\'' || buf[off] == '"') {
451 throw new Exn("attribute '"+n+"' must have attribute wrapped in ' or \"", Exn.MARKUP, getLine(), getCol());
455 // find the attribute value
456 attval: for (ref = 0; ; ref++) {
457 if (!buffer(ref+1)) throw new EOFException("Unexpected EOF in attribute value");
459 if (buf[off+ref] == wrap) {
461 } else if (buf[off+ref] == '<') {
462 throw new Exn("attribute value for '"+n+"' must not contain '<'", Exn.WFC, getLine(), getCol());
466 v = new String(buf, off, ref);
467 col += ref; off += ref; len -= ref;
469 // remove end wrapper character
473 if (p != null && p.equals("xmlns")) {
474 current.addUri(n, v);
475 } else if (n.equals("xmlns")) {
476 if (current.getUri().equals("")) {
477 current.addUri("", v);
479 current.addError(new Exn("default namespace definition repeated", Exn.NC, getLine(), getCol()));
482 // find attribute uri
483 u = current.getUri(p);
484 if (p != null && u.equals("")) current.addError(new Exn("undefined attribute prefix '"+p+"'", Exn.NC, getLine(), getCol()));
486 // check to see if attribute is a repeat
487 for (int i=0; current.len > i; i++) if (n.equals(current.getAttrKey(i)) && u.equals(current.getAttrUri(i))) throw new Exn(
488 "attribute name '"+n+"' may not appear more than once in the same element tag", Exn.WFC, getLine(), getCol()
491 current.addAttr(n, v, u);
495 /** reads an entity and processes out its value. expects buf[off] == '&' */
496 private final void readEntity() throws IOException, Exn {
498 if (!buffer(2)) throw new EOFException("Unexpected EOF reading entity");
500 boolean unknown = false;
506 if (buf[off] == 'x') { off++; len--; radix = 16; } else { radix = 10; }
509 // read in each char, then shift total value to the left and add the extra
510 // style of loop is slightly different from all the others, as this should run a limited number of times
511 findchar: while (true) {
512 if (!buffer(1)) throw new EOFException("Unexpected EOF reading entity");
513 int d = Character.digit(buf[off], radix);
515 if (buf[off] != ';') throw new Exn("illegal characters in entity reference", Exn.WFC, getLine(), getCol());
524 singlechar[0] = Character.forDigit(c, radix);
525 characters(singlechar, 0, 1);
529 if (buffer(4) && buf[off+1] == 'm' && buf[off+2] == 'p' && buf[off+3] == ';') {
530 characters(single_amp, 0, 1); // &
531 off += 4; len -= 4; col++;
532 } else if (buffer(5) && buf[off+1] == 'p' && buf[off+2] == 'o' && buf[off+3] == 's' && buf[off+4] == ';') {
533 characters(single_apos, 0, 1); // '
534 off += 5; len -= 5; col++;
541 if (buffer(3) && buf[off+1] == 't' && buf[off+2] == ';') {
542 characters(single_gt, 0, 1); // >
543 off += 3; len -= 3; col++;
550 if (buffer(3) && buf[off+1] == 't' && buf[off+2] == ';') {
551 characters(single_lt, 0, 1); // <
552 off += 3; len -= 3; col++;
559 if (buffer(5) && buf[off+1] == 'u' && buf[off+2] == 'o' && buf[off+3] == 't' && buf[off+4] == ';') {
560 characters(single_quot, 0, 1); // "
561 off += 5; len -= 5; col++;
567 // TODO: check a parser-level Hash of defined entities
570 if (unknown) throw new Exn("unknown entity (<!ENTITY> not supported)", Exn.WFC, getLine(), getCol());
573 /** reads until the passed string is encountered. */
574 final void readChars(boolean p, String match, boolean entities) throws IOException, Exn {
576 char[] end = match.toCharArray();
578 for (boolean more = true; more;) {
579 if (!buffer(1)) return;
581 buf: for (ref = 0; ref < len; ref++) {
582 switch (buf[off+ref]) {
583 case '\r': // windows or macos9 newline
584 // normalise and process
585 buf[off+ref] = '\n'; ref++;
586 if (p) characters(buf, off, ref);
587 off += ref; len -= ref; ref = -1;
590 // windows double-char newline; skip the next char
591 if (!buffer(1)) return;
592 if (buf[off] == '\n') { off++; len--; }
595 case '\n': // unix newline
597 if (p) characters(buf, off, ref);
598 off += ref; len -= ref; ref = -1;
605 if (ref > 0) characters(buf, off, ref);
606 off += ref; len -= ref; ref = -1;
613 if (!buffer(ref+end.length)) continue buf;
614 for (int i=0; end.length > i; i++) if (end[i] != buf[off+ref+i]) continue buf;
620 if (p && ref > 0) characters(buf, off, ref);
621 off += ref; len -= ref; col += ref;
626 * reads until a <tt><</tt> symbol is encountered
627 * @param p If true call the characters(char[],int,int) funciton for the processed characters
629 final void readChars(boolean p) throws IOException, Exn {
632 for (boolean more = true; more;) {
633 if (!buffer(1)) return;
635 buf: for (ref = 0; ref < len; ref++) {
636 switch (buf[off+ref]) {
637 case '\r': // windows or macos9 newline
638 // normalise and process
639 buf[off+ref] = '\n'; ref++;
640 if (p) characters(buf, off, ref);
641 off += ref; len -= ref; ref = -1;
644 // windows double-char newline; skip the next char
645 if (!buffer(1)) return;
646 if (buf[off] == '\n') { off++; len--; }
649 case '\n': // unix newline
651 if (p) characters(buf, off, ref);
652 off += ref; len -= ref; ref = -1;
658 if (ref > 0) characters(buf, off, ref);
659 off += ref; len -= ref; ref = -1;
664 case '<': // end of chars section
670 if (p && ref > 0) characters(buf, off, ref);
671 off += ref; len -= ref; col += ref;
675 /** reads until a non-whitespace symbol is encountered */
676 private final void readWhitespace() throws IOException, Exn {
679 for (boolean more = true; more;) {
680 if (!buffer(1)) return;
682 buf: for (ref = 0; ref < len; ref++) {
683 switch (buf[off+ref]) {
684 case '\r': // windows or macos9 newline
685 // normalise and process
687 whitespace(buf, off, ++ref);
688 off += ref; len -= ref; ref = -1;
691 // windows double-char newline; skip the next char
692 if (!buffer(1)) return;
693 if (buf[off] == '\n') { off++; len--; }
696 case '\n': // unix newline
697 whitespace(buf, off, ++ref);
698 off += ref; len -= ref; ref = -1;
706 default: // end of whitespace
712 off += ref; len -= ref; col += ref;
717 * attempt to fill the buffer.
719 * @param min Minimum number of characters to read (even if we have to block to do it).
720 * @return return false if min can't be reached.
722 final boolean buffer(int min) throws IOException {
723 if (len > min) return true;
725 if (buf.length - (off+len) >= min) {
726 // plenty of space left on the end of the buffer
727 } else if (off >= min) {
728 // moving offset data to start will leave enough free space on the end
729 System.arraycopy(buf, off, buf, 0, len);
733 // buffer size will have to be increased
734 char[] newbuf = new char[buf.length * 2];
735 System.arraycopy(buf, off, newbuf, 0, len);
742 int newlen = in.read(buf, off+len, buf.length-(off+len));
743 if (newlen < 0) return false;
751 /////////////////////////////////////////////////////////////////////////////////////////////
752 // Abstract SAX-Like Interface
753 /////////////////////////////////////////////////////////////////////////////////////////////
756 * Called when the start of an element is processed.
758 * <p><b>DO NOT</b> store a reference to the Element object, as
759 * they are reused by XML Parser.</p>
761 public abstract void startElement(Element e) throws Exn;
764 * Represents up to a line of character data.
766 * <p>Newlines are all normalised to the Unix \n as per the XML Spec,
767 * and a newline will only appear as the last character in the passed
770 * <p>XML.getLine() and XML.getCol() report the position at the
771 * beginning of this character segment, which can be processed in a
772 * line-by-line fashion due to the above newline restriction.</p>
774 public abstract void characters(char[] ch, int start, int length) throws Exn, IOException;
776 /** Represents up to a line of ignorable whitespace. */
777 public abstract void whitespace(char[] ch, int start, int length) throws Exn, IOException;
779 /** Represents the end of an Element. */
780 public abstract void endElement(Element e) throws Exn, IOException;
782 /////////////////////////////////////////////////////////////////////////////////////////////
783 // Inner Classes for Parser Support
784 /////////////////////////////////////////////////////////////////////////////////////////////
787 * Represents an element in an XML document. Stores a reference to its
788 * parent, forming a one-way linked list.
790 * Element objects are reused, so client code making use of them must
791 * drop their references after the specific element process function
794 public static final class Element {
796 private static final int DEFAULT_ATTR_SIZE = 10;
798 protected Element parent = null;
800 protected String uri = null;
801 protected String localName = null;
802 protected String qName = null;
803 protected String prefix = null;
804 public int level = 0;
806 protected Hash urimap = new Hash(3,3);
808 protected String[] keys = new String[DEFAULT_ATTR_SIZE];
809 protected String[] vals = new String[DEFAULT_ATTR_SIZE];
810 protected String[] uris = new String[DEFAULT_ATTR_SIZE];
811 protected int len = 0;
813 protected Exn[] errors = new Exn[] {};
815 /** Parent of current element. */
816 public Element getParent() { return parent; }
818 /** Qualified Name of current element. XML Namespace Spec 14-Jan-1999 [6] */
819 public String getQName() { return qName; }
821 /** LocalPart of current element. XML Namespace Spec 14-Jan-1999 [8] */
822 public String getLocalName() { return localName; }
824 /** Prefix of current element. Substring of qName. XML Namespace Spec 14-Jan-1999 [7] */
825 public String getPrefix() { return prefix; }
828 public Hash getUriMap() {
829 Hash map = new Hash();
830 for (Element e = this; e != null; e = e.getParent()) {
831 java.util.Enumeration en = e.urimap.keys();
832 while(en.hasMoreElements()) {
833 String key = (String)en.nextElement();
834 String val = getUri(key);
841 /** URI of current tag. XML Namespace Spec 14-Jan-1999 section 1 */
842 public String getUri() { return getUri(prefix); }
844 /** URI of a given prefix. Never returns null, instead gives "". */
845 public String getUri(String p) {
847 for (Element e = this; e != null && ret == null; e = e.getParent()) {
848 ret = (String)e.urimap.get(p == null ? "" : p);
850 return ret == null ? "" : ret;
853 /** An array of attribute names. */
854 public String getAttrKey(int pos) { return len > pos ? keys[pos] : null; }
856 /** An array of attribute values. */
857 public String getAttrVal(int pos) { return len > pos ? vals[pos] : null; }
859 /** An array of attribute uris. */
860 public String getAttrUri(int pos) { return len > pos ? uris[pos] : null; }
862 /** Current number of attributes in the element. */
863 public int getAttrLen() { return len; }
865 /** Poor performance, but easier to use when speed is not a concern */
866 public Hash getAttrHash() {
867 Hash ret = new Hash(getAttrLen() * 2, 3);
868 for(int i=0; i<len; i++)
869 ret.put(getAttrKey(i), getAttrVal(i));
873 /** Poor performance, but easier to use */
874 public String getAttrVal(String key) {
875 for(int i=0; i<len; i++) if (keys[i].equals(key)) return vals[i];
879 /** An array of non-fatal errors related to this element. */
880 public Exn[] getErrors() { return errors; }
882 protected Element() { }
884 /** Add (replace if exists in current element) a Namespace prefix/uri map. */
885 public void addUri(String name, String value) {
886 urimap.put(name, value);
889 /** Add an attribute. */
890 protected void addAttr(String key, String val, String uri) {
891 if (len == keys.length) {
892 // increase the size of the attributes arrays
893 String[] newkeys = new String[keys.length*2];
894 String[] newvals = new String[vals.length*2];
895 String[] newuris = new String[uris.length*2];
896 System.arraycopy(keys, 0, newkeys, 0, keys.length);
897 System.arraycopy(vals, 0, newvals, 0, vals.length);
898 System.arraycopy(uris, 0, newuris, 0, uris.length);
899 keys = newkeys; vals = newvals; uris = newuris;
909 protected void addError(Exn e) {
910 // it doesn't really matter about continually expanding the array, as this case is quite rare
911 Exn[] newe = new Exn[errors.length+1];
912 System.arraycopy(errors, 0, newe, 0, errors.length);
913 newe[errors.length] = e;
917 /** Empty out all the data from the Element. */
918 protected void clear() {
920 uri = localName = qName = prefix = null;
923 if (keys.length != vals.length || vals.length != uris.length) {
924 keys = new String[DEFAULT_ATTR_SIZE];
925 vals = new String[DEFAULT_ATTR_SIZE];
926 uris = new String[DEFAULT_ATTR_SIZE];
928 for (int i=0; keys.length > i; i++) { keys[i] = null; vals[i] = null; uris[i] = null; };
932 errors = new Exn[] {};
936 /** Parse or Structural Error */
937 public static class Exn extends Exception {
938 /** Violation of Markup restrictions in XML Specification - Fatal Error */
939 public static final int MARKUP = 1;
941 /** Well-Formedness Constraint Violation - Fatal Error */
942 public static final int WFC = 2;
944 /** Namespace Constraint Violation - Recoverable Error */
945 public static final int NC = 3;
947 /** Schema Violation - Fatal Error */
948 public static final int SCHEMA = 4;
950 private String error;
955 public Exn(String e) { this(e, MARKUP, -1, -1); }
957 public Exn(String e, int type, int line, int col) {
964 public int getType() { return this.type; }
965 public int getLine() { return this.line; }
966 public int getCol() { return this.col; }
967 public String getMessage() { return this.error + (line >= 0 && col >= 0 ? " at " + line + ":" + col: ""); }
971 /////////////////////////////////////////////////////////////////////////////////////////////
972 // Static Support Functions for the XML Specification
973 /////////////////////////////////////////////////////////////////////////////////////////////
975 // attempt to avoid these functions unless you *expect* the input to fall in the given range.
977 /** First Character of Name - XML Specification 1.0 [5] */
978 private static final boolean Name(char c) {
979 return BaseCharAscii(c) || c == '_' || c == ':' || Letter(c);
982 /** NameChar - XML Specification 1.0 [4] */
983 private static final boolean NameChar(char c) {
984 return BaseCharAscii(c) || c == '.' || c == '-' || c == '_' || c == ':'
985 || Digit(c) || Letter(c) || Extender(c); // TODO: || CombiningChar(c);
988 /** BaseChar - XMl Specification 1.0 [84] */
989 private static final boolean Letter(char c) {
990 return BaseChar(c) || Ideographic(c);
993 /** Elements of BaseChar that exist in ASCII. */
994 private static final boolean BaseCharAscii(char c) {
995 return (c >= '\u0041' && c <= '\u005A') || (c >= '\u0061' && c <= '\u007A');
998 /** Char - XML Specification 1.0 [2] */
999 private static final boolean Char(char c) {
1000 // u000A == r and u000D == n, but the javac compiler can't handle the \ u form
1001 return c == '\u0009' || c == '\r' || c == '\n'
1002 || (c >= '\u0020' && c <= '\uD7FF')
1003 || (c >= '\uE000' && c <= '\uFFFD');
1006 /** BaseChar - XML Specification 1.0 [85] */
1007 private static final boolean BaseChar(char c) {
1008 return BaseCharAscii(c) || (c >= '\u00C0' && c <= '\u00D6')
1009 || (c >= '\u00D8' && c <= '\u00F6') || (c >= '\u00F8' && c <= '\u00FF') || (c >= '\u0100' && c <= '\u0131')
1010 || (c >= '\u0134' && c <= '\u013E') || (c >= '\u0141' && c <= '\u0148') || (c >= '\u014A' && c <= '\u017E')
1011 || (c >= '\u0180' && c <= '\u01C3') || (c >= '\u01CD' && c <= '\u01F0') || (c >= '\u01F4' && c <= '\u01F5')
1012 || (c >= '\u01FA' && c <= '\u0217') || (c >= '\u0250' && c <= '\u02A8') || (c >= '\u02BB' && c <= '\u02C1')
1013 || (c == '\u0386') || (c >= '\u0388' && c <= '\u038A') || (c == '\u038C')
1014 || (c >= '\u038E' && c <= '\u03A1') || (c >= '\u03A3' && c <= '\u03CE') || (c >= '\u03D0' && c <= '\u03D6')
1015 || (c == '\u03DA') || (c == '\u03DC') || (c == '\u03DE')
1017 || (c >= '\u03E2' && c <= '\u03F3') || (c >= '\u0401' && c <= '\u040C') || (c >= '\u040E' && c <= '\u044F')
1018 || (c >= '\u0451' && c <= '\u045C') || (c >= '\u045E' && c <= '\u0481') || (c >= '\u0490' && c <= '\u04C4')
1019 || (c >= '\u04C7' && c <= '\u04C8') || (c >= '\u04CB' && c <= '\u04CC') || (c >= '\u04D0' && c <= '\u04EB')
1020 || (c >= '\u04EE' && c <= '\u04F5') || (c >= '\u04F8' && c <= '\u04F9') || (c >= '\u0531' && c <= '\u0556')
1022 || (c >= '\u0561' && c <= '\u0586') || (c >= '\u05D0' && c <= '\u05EA') || (c >= '\u05F0' && c <= '\u05F2')
1023 || (c >= '\u0621' && c <= '\u063A') || (c >= '\u0641' && c <= '\u064A') || (c >= '\u0671' && c <= '\u06B7')
1024 || (c >= '\u06BA' && c <= '\u06BE') || (c >= '\u06C0' && c <= '\u06CE') || (c >= '\u06D0' && c <= '\u06D3')
1026 || (c >= '\u06E5' && c <= '\u06E6') || (c >= '\u0905' && c <= '\u0939')
1028 || (c >= '\u0958' && c <= '\u0961') || (c >= '\u0985' && c <= '\u098C') || (c >= '\u098F' && c <= '\u0990')
1029 || (c >= '\u0993' && c <= '\u09A8') || (c >= '\u09AA' && c <= '\u09B0')
1031 || (c >= '\u09B6' && c <= '\u09B9') || (c >= '\u09DF' && c <= '\u09E1') || (c >= '\u09F0' && c <= '\u09F1')
1032 || (c >= '\u0A05' && c <= '\u0A0A') || (c >= '\u0A0F' && c <= '\u0A10') || (c >= '\u0A13' && c <= '\u0A28')
1033 || (c >= '\u0A2A' && c <= '\u0A30') || (c >= '\u0A32' && c <= '\u0A33') || (c >= '\u0A35' && c <= '\u0A36')
1034 || (c >= '\u0A38' && c <= '\u0A39') || (c >= '\u0A59' && c <= '\u0A5C')
1036 || (c >= '\u0A72' && c <= '\u0A74') || (c >= '\u0A85' && c <= '\u0A8B')
1038 || (c >= '\u0A8F' && c <= '\u0A91') || (c >= '\u0A93' && c <= '\u0AA8') || (c >= '\u0AAA' && c <= '\u0AB0')
1039 || (c >= '\u0AB2' && c <= '\u0AB3') || (c >= '\u0AB5' && c <= '\u0AB9')
1042 || (c >= '\u0B05' && c <= '\u0B0C') || (c >= '\u0B0F' && c <= '\u0B10') || (c >= '\u0B13' && c <= '\u0B28')
1043 || (c >= '\u0B2A' && c <= '\u0B30') || (c >= '\u0B32' && c <= '\u0B33') || (c >= '\u0B36' && c <= '\u0B39')
1045 || (c >= '\u0B5C' && c <= '\u0B5D') || (c >= '\u0B5F' && c <= '\u0B61') || (c >= '\u0B85' && c <= '\u0B8A')
1046 || (c >= '\u0B8E' && c <= '\u0B90') || (c >= '\u0B92' && c <= '\u0B95') || (c >= '\u0B99' && c <= '\u0B9A')
1048 || (c >= '\u0B9E' && c <= '\u0B9F') || (c >= '\u0BA3' && c <= '\u0BA4') || (c >= '\u0BA8' && c <= '\u0BAA')
1049 || (c >= '\u0BAE' && c <= '\u0BB5') || (c >= '\u0BB7' && c <= '\u0BB9') || (c >= '\u0C05' && c <= '\u0C0C')
1050 || (c >= '\u0C0E' && c <= '\u0C10') || (c >= '\u0C12' && c <= '\u0C28') || (c >= '\u0C2A' && c <= '\u0C33')
1051 || (c >= '\u0C35' && c <= '\u0C39') || (c >= '\u0C60' && c <= '\u0C61') || (c >= '\u0C85' && c <= '\u0C8C')
1052 || (c >= '\u0C8E' && c <= '\u0C90') || (c >= '\u0C92' && c <= '\u0CA8') || (c >= '\u0CAA' && c <= '\u0CB3')
1053 || (c >= '\u0CB5' && c <= '\u0CB9')
1055 || (c >= '\u0CE0' && c <= '\u0CE1') || (c >= '\u0D05' && c <= '\u0D0C') || (c >= '\u0D0E' && c <= '\u0D10')
1056 || (c >= '\u0D12' && c <= '\u0D28') || (c >= '\u0D2A' && c <= '\u0D39') || (c >= '\u0D60' && c <= '\u0D61')
1057 || (c >= '\u0E01' && c <= '\u0E2E')
1059 || (c >= '\u0E32' && c <= '\u0E33') || (c >= '\u0E40' && c <= '\u0E45') || (c >= '\u0E81' && c <= '\u0E82')
1061 || (c >= '\u0E87' && c <= '\u0E88')
1064 || (c >= '\u0E94' && c <= '\u0E97') || (c >= '\u0E99' && c <= '\u0E9F') || (c >= '\u0EA1' && c <= '\u0EA3')
1067 || (c >= '\u0EAA' && c <= '\u0EAB') || (c >= '\u0EAD' && c <= '\u0EAE')
1069 || (c >= '\u0EB2' && c <= '\u0EB3')
1071 || (c >= '\u0EC0' && c <= '\u0EC4') || (c >= '\u0F40' && c <= '\u0F47') || (c >= '\u0F49' && c <= '\u0F69')
1072 || (c >= '\u10A0' && c <= '\u10C5') || (c >= '\u10D0' && c <= '\u10F6')
1074 || (c >= '\u1102' && c <= '\u1103') || (c >= '\u1105' && c <= '\u1107')
1076 || (c >= '\u110B' && c <= '\u110C') || (c >= '\u110E' && c <= '\u1112')
1083 || (c >= '\u1154' && c <= '\u1155')
1085 || (c >= '\u115F' && c <= '\u1161')
1090 || (c >= '\u116D' && c <= '\u116E') || (c >= '\u1172' && c <= '\u1173')
1095 || (c >= '\u11AE' && c <= '\u11AF') || (c >= '\u11B7' && c <= '\u11B8')
1097 || (c >= '\u11BC' && c <= '\u11C2')
1101 || (c >= '\u1E00' && c <= '\u1E9B') || (c >= '\u1EA0' && c <= '\u1EF9') || (c >= '\u1F00' && c <= '\u1F15')
1102 || (c >= '\u1F18' && c <= '\u1F1D') || (c >= '\u1F20' && c <= '\u1F45') || (c >= '\u1F48' && c <= '\u1F4D')
1103 || (c >= '\u1F50' && c <= '\u1F57')
1107 || (c >= '\u1F5F' && c <= '\u1F7D') || (c >= '\u1F80' && c <= '\u1FB4') || (c >= '\u1FB6' && c <= '\u1FBC')
1109 || (c >= '\u1FC2' && c <= '\u1FC4') || (c >= '\u1FC6' && c <= '\u1FCC') || (c >= '\u1FD0' && c <= '\u1FD3')
1110 || (c >= '\u1FD6' && c <= '\u1FDB') || (c >= '\u1FE0' && c <= '\u1FEC') || (c >= '\u1FF2' && c <= '\u1FF4')
1111 || (c >= '\u1FF6' && c <= '\u1FFC')
1113 || (c >= '\u212A' && c <= '\u212B')
1115 || (c >= '\u2180' && c <= '\u2182') || (c >= '\u3041' && c <= '\u3094') || (c >= '\u30A1' && c <= '\u30FA')
1116 || (c >= '\u3105' && c <= '\u312C') || (c >= '\uAC00' && c <= '\uD7A3');
1119 /** BaseChar - XMl Specification 1.0 [86] */
1120 private static final boolean Ideographic(char c) {
1121 return (c >= '\u4E00' && c <= '\u9FA5') || c == '\u3007' || (c >= '\u3021' && c <= '\u3029');
1124 /** CombiningChar - XMl Specification 1.0 [87] */
1125 /*private static final boolean CombiningChar(char c) {
1126 return (c >= '\u0300' && c <= '\u0345')
1127 || (c >= '\u0360' && c <= '\u0361') || (c >= '\u0483' && c <= '\u0486') || (c >= '\u0591' && c <= '\u05A1')
1128 || (c >= '\u05A3' && c <= '\u05B9') || (c >= '\u05BB' && c <= '\u05BD')
1130 || (c >= '\u05C1' && c <= '\u05C2')
1132 || (c >= '\u064B' && c <= '\u0652')
1134 || (c >= '\u06D6' && c <= '\u06DC') || (c >= '\u06DD' && c <= '\u06DF') || (c >= '\u06E0' && c <= '\u06E4')
1135 || (c >= '\u06E7' && c <= '\u06E8') || (c >= '\u06EA' && c <= '\u06ED') || (c >= '\u0901' && c <= '\u0903')
1137 || (c >= '\u093E' && c <= '\u094C')
1139 || (c >= '\u0951' && c <= '\u0954') || (c >= '\u0962' && c <= '\u0963') || (c >= '\u0981' && c <= '\u0983')
1143 || (c >= '\u09C0' && c <= '\u09C4') || (c >= '\u09C7' && c <= '\u09C8') || (c >= '\u09CB' && c <= '\u09CD')
1145 || (c >= '\u09E2' && c <= '\u09E3')
1150 || (c >= '\u0A40' && c <= '\u0A42') || (c >= '\u0A47' && c <= '\u0A48') || (c >= '\u0A4B' && c <= '\u0A4D')
1151 || (c >= '\u0A70' && c <= '\u0A71') || (c >= '\u0A81' && c <= '\u0A83')
1153 || (c >= '\u0ABE' && c <= '\u0AC5') || (c >= '\u0AC7' && c <= '\u0AC9') || (c >= '\u0ACB' && c <= '\u0ACD')
1154 || (c >= '\u0B01' && c <= '\u0B03')
1156 || (c >= '\u0B3E' && c <= '\u0B43') || (c >= '\u0B47' && c <= '\u0B48') || (c >= '\u0B4B' && c <= '\u0B4D')
1157 || (c >= '\u0B56' && c <= '\u0B57') || (c >= '\u0B82' && c <= '\u0B83') || (c >= '\u0BBE' && c <= '\u0BC2')
1158 || (c >= '\u0BC6' && c <= '\u0BC8') || (c >= '\u0BCA' && c <= '\u0BCD')
1160 || (c >= '\u0C01' && c <= '\u0C03') || (c >= '\u0C3E' && c <= '\u0C44') || (c >= '\u0C46' && c <= '\u0C48')
1161 || (c >= '\u0C4A' && c <= '\u0C4D') || (c >= '\u0C55' && c <= '\u0C56') || (c >= '\u0C82' && c <= '\u0C83')
1162 || (c >= '\u0CBE' && c <= '\u0CC4') || (c >= '\u0CC6' && c <= '\u0CC8') || (c >= '\u0CCA' && c <= '\u0CCD')
1163 || (c >= '\u0CD5' && c <= '\u0CD6') || (c >= '\u0D02' && c <= '\u0D03') || (c >= '\u0D3E' && c <= '\u0D43')
1164 || (c >= '\u0D46' && c <= '\u0D48') || (c >= '\u0D4A' && c <= '\u0D4D')
1167 || (c >= '\u0E34' && c <= '\u0E3A') || (c >= '\u0E47' && c <= '\u0E4E')
1169 || (c >= '\u0EB4' && c <= '\u0EB9') || (c >= '\u0EBB' && c <= '\u0EBC') || (c >= '\u0EC8' && c <= '\u0ECD')
1170 || (c >= '\u0F18' && c <= '\u0F19')
1176 || (c >= '\u0F71' && c <= '\u0F84') || (c >= '\u0F86' && c <= '\u0F8B') || (c >= '\u0F90' && c <= '\u0F95')
1178 || (c >= '\u0F99' && c <= '\u0FAD') || (c >= '\u0FB1' && c <= '\u0FB7')
1180 || (c >= '\u20D0' && c <= '\u20DC')
1182 || (c >= '\u302A' && c <= '\u302F')
1187 /** Digit - XMl Specification 1.0 [88] */
1188 private static final boolean Digit(char c) {
1189 return (c >= '\u0030' && c <= '\u0039') || (c >= '\u0660' && c <= '\u0669') || (c >= '\u06F0' && c <= '\u06F9')
1190 || (c >= '\u0966' && c <= '\u096F') || (c >= '\u09E6' && c <= '\u09EF') || (c >= '\u0A66' && c <= '\u0A6F')
1191 || (c >= '\u0AE6' && c <= '\u0AEF') || (c >= '\u0B66' && c <= '\u0B6F') || (c >= '\u0BE7' && c <= '\u0BEF')
1192 || (c >= '\u0C66' && c <= '\u0C6F') || (c >= '\u0CE6' && c <= '\u0CEF') || (c >= '\u0D66' && c <= '\u0D6F')
1193 || (c >= '\u0E50' && c <= '\u0E59') || (c >= '\u0ED0' && c <= '\u0ED9') || (c >= '\u0F20' && c <= '\u0F29');
1196 /** Extender - XMl Specification 1.0 [89] */
1197 private static final boolean Extender(char c) {
1198 return c == '\u00B7' || c == '\u02D0' || c == '\u02D1' || c == '\u0387'
1199 || c == '\u0640' || c == '\u0E46' || c == '\u0EC6' || c == '\u3005'
1200 || (c >= '\u3031' && c <= '\u3035') || (c >= '\u309D' && c <= '\u309E') || (c >= '\u30FC' && c <= '\u30FE');
1203 /** Whitespace - XML Specification 1.0 [3] */
1204 private static final boolean S(char c) {
1205 return c == '\u0020' || c == '\u0009' || c == '\r' || c == '\n';