* The TibiXmlParser is a small XML pull parser, * specially designed for constrained environments such as *
** The TibiXmlParser is based on two classes. *
** A main change is a replacing org.xmlpull.v1.XmlPullParserException with java.lang.Exception. *
* ** A minimal example for using this API may look as follows: *
*
* public class SimpleXmlPullApp
* {
*
* public static void main (String args[])
* throws Exception, IOException
* {
* TibiXmlParser xpp = new TibiXmlParser();
*
* xpp.setInput( new StringReader ( "<foo>Hello World!</foo>" ) );
* int eventType = xpp.getEventType();
* while (eventType != xpp.END_DOCUMENT) {
* if(eventType == xpp.START_DOCUMENT) {
* System.out.println("Start document");
* } else if(eventType == xpp.END_DOCUMENT) {
* System.out.println("End document");
* } else if(eventType == xpp.START_TAG) {
* System.out.println("Start tag "+xpp.getName());
* } else if(eventType == xpp.END_TAG) {
* System.out.println("End tag "+xpp.getName());
* } else if(eventType == xpp.TEXT) {
* System.out.println("Text "+xpp.getText());
* }
* eventType = xpp.next();
* }
* }
* }
*
*
* * The above example will generate the following output: *
** Start document * Start tag foo * Text Hello World! * End tag foo ** * @see #defineEntityReplacementText * @see #getName * @see #getNamespace * @see #getText * @see #next * @see #nextToken * @see #setInput * @see #FEATURE_PROCESS_DOCDECL * @see #FEATURE_VALIDATION * @see #START_DOCUMENT * @see #START_TAG * @see #TEXT * @see #END_TAG * @see #END_DOCUMENT * * @author NI-Lab. * @version 1.1.0 */ public class TibiXmlParser { /** This constant represents the default namespace (empty string "") */ public static final String NO_NAMESPACE = ""; // ---------------------------------------------------------------------------- // EVENT TYPES as reported by next() /** * Signalize that parser is at the very beginning of the document * and nothing was read yet. * This event type can only be observed by calling getEvent() * before the first call to next(), nextToken, or nextTag()). * * @see #next * @see #nextToken */ public final static int START_DOCUMENT = 0; /** * Logical end of the xml document. Returned from getEventType, next() * and nextToken() * when the end of the input document has been reached. *
NOTE: calling again * next() or nextToken() * will result in exception being thrown. * * @see #next * @see #nextToken */ public final static int END_DOCUMENT = 1; /** * Returned from getEventType(), * next(), nextToken() when * a start tag was read. * The name of start tag is available from getName(), its namespace and prefix are * available from getNamespace() and getPrefix() * if namespaces are enabled. * See getAttribute* methods to retrieve element attributes. * See getNamespace* methods to retrieve newly declared namespaces. * * @see #next * @see #nextToken * @see #getName * @see #getPrefix * @see #getNamespace * @see #getAttributeCount * @see #getDepth * @see #getNamespaceCount * @see #getNamespace * @see #FEATURE_PROCESS_NAMESPACES */ public final static int START_TAG = 2; /** * Returned from getEventType(), next(), or * nextToken() when an end tag was read. * The name of start tag is available from getName(), its * namespace and prefix are * available from getNamespace() and getPrefix(). * * @see #next * @see #nextToken * @see #getName * @see #getPrefix * @see #getNamespace * @see #FEATURE_PROCESS_NAMESPACES */ public final static int END_TAG = 3; /** * Character data was read and will is available by calling getText(). *
Please note: next() will * accumulate multiple * events into one TEXT event, skipping IGNORABLE_WHITESPACE, * PROCESSING_INSTRUCTION and COMMENT events, * In contrast, nextToken() will stop reading * text when any other event is observed. * Also, when the state was reached by calling next(), the text value will * be normalized, whereas getText() will * return unnormalized content in the case of nextToken(). This allows * an exact roundtrip without chnanging line ends when examining low * level events, whereas for high level applications the text is * normalized apropriately. * * @see #next * @see #nextToken * @see #getText */ public final static int TEXT = 4; // ---------------------------------------------------------------------------- // additional events exposed by lower level nextToken() /** * A CDATA sections was just read; * this token is available only from calls to nextToken(). * A call to next() will accumulate various text events into a single event * of type TEXT. The text contained in the CDATA section is available * by callling getText(). * * @see #nextToken * @see #getText */ public final static int CDSECT = 5; /** * An entity reference was just read; * this token is available from nextToken() * only. The entity name is available by calling getName(). If available, * the replacement text can be obtained by calling getTextt(); otherwise, * the user is responsibile for resolving the entity reference. * This event type is never returned from next(); next() will * accumulate the replacement text and other text * events to a single TEXT event. * * @see #nextToken * @see #getText */ public final static int ENTITY_REF = 6; /** * Ignorable whitespace was just read. * This token is available only from nextToken()). * For non-validating * parsers, this event is only reported by nextToken() when outside * the root element. * Validating parsers may be able to detect ignorable whitespace at * other locations. * The ignorable whitespace string is available by calling getText() * *
NOTE: this is different from calling the * isWhitespace() method, since text content * may be whitespace but not ignorable. * * Ignorable whitespace is skipped by next() automatically; this event * type is never returned from next(). * * @see #nextToken * @see #getText */ public static final int IGNORABLE_WHITESPACE = 7; /** * An XML processing instruction declaration was just read. This * event type is available only via nextToken(). * getText() will return text that is inside the processing instruction. * Calls to next() will skip processing instructions automatically. * @see #nextToken * @see #getText */ public static final int PROCESSING_INSTRUCTION = 8; /** * An XML comment was just read. This event type is this token is * available via nextToken() only; * calls to next() will skip comments automatically. * The content of the comment can be accessed using the getText() * method. * * @see #nextToken * @see #getText */ public static final int COMMENT = 9; /** * An XML document type declaration was just read. This token is * available from nextToken() only. * The unparsed text inside the doctype is available via * the getText() method. * * @see #nextToken * @see #getText */ public static final int DOCDECL = 10; /** * This array can be used to convert the event type integer constants * such as START_TAG or TEXT to * to a string. For example, the value of TYPES[START_TAG] is * the string "START_TAG". * * This array is intended for diagnostic output only. Relying * on the contents of the array may be dangerous since malicous * applications may alter the array, although it is final, due * to limitations of the Java language. */ public static final String [] TYPES = { "START_DOCUMENT", "END_DOCUMENT", "START_TAG", "END_TAG", "TEXT", "CDSECT", "ENTITY_REF", "IGNORABLE_WHITESPACE", "PROCESSING_INSTRUCTION", "COMMENT", "DOCDECL" }; // ---------------------------------------------------------------------------- // namespace related features /** * This feature determines whether the parser processes * namespaces. As for all features, the default value is false. *
NOTE: The value can not be changed during * parsing an must be set before parsing. * * @see #getFeature * @see #setFeature */ public static final String FEATURE_PROCESS_NAMESPACES = "http://xmlpull.org/v1/doc/features.html#process-namespaces"; /** * This feature determines whether namespace attributes are * exposed via the attribute access methods. Like all features, * the default value is false. This feature cannot be changed * during parsing. * * @see #getFeature * @see #setFeature */ public static final String FEATURE_REPORT_NAMESPACE_ATTRIBUTES = "http://xmlpull.org/v1/doc/features.html#report-namespace-prefixes"; /** * This feature determines whether the document declaration * is processed. If set to false, * the DOCDECL event type is reported by nextToken() * and ignored by next(). * * If this featue is activated, then the document declaration * must be processed by the parser. * *
Please note: If the document type declaration * was ignored, entity references may cause exceptions * later in the parsing process. * The default value of this feature is false. It cannot be changed * during parsing. * * @see #getFeature * @see #setFeature */ public static final String FEATURE_PROCESS_DOCDECL = "http://xmlpull.org/v1/doc/features.html#process-docdecl"; /** * If this feature is activated, all validation errors as * defined in the XML 1.0 sepcification are reported. * This implies that FEATURE_PROCESS_DOCDECL is true and both, the * internal and external document type declaration will be processed. *
Please Note: This feature can not be changed * during parsing. The default value is false. * * @see #getFeature * @see #setFeature */ public static final String FEATURE_VALIDATION = "http://xmlpull.org/v1/doc/features.html#validation"; private Object location; static final private String UNEXPECTED_EOF = "Unexpected EOF"; static final private String ILLEGAL_TYPE = "Wrong event type"; static final private int LEGACY = 999; static final private int XML_DECL = 998; // general private String version; private Boolean standalone; private boolean processNsp; private boolean relaxed; private Hashtable entityMap; private int depth; private String[] elementStack = new String[16]; private String[] nspStack = new String[8]; private int[] nspCounts = new int[4]; // source private Reader reader; private String encoding; private char[] srcBuf; private int srcPos; private int srcCount; private int line; private int column; // txtbuffer private char[] txtBuf = new char[128]; private int txtPos; // Event-related private int type; //private String text; private boolean isWhitespace; private String namespace; private String prefix; private String name; private boolean degenerated; private int attributeCount; private String[] attributes = new String[16]; private int stackMismatch = 0; private String error; /** * A separate peek buffer seems simpler than managing * wrap around in the first level read buffer */ private int[] peek = new int[2]; private int peekCount; private boolean wasCR; private boolean unresolved; private boolean token; public TibiXmlParser() { srcBuf = new char[Runtime.getRuntime().freeMemory() >= 1048576 ? 8192 : 128]; } private final boolean isProp(String n1, boolean prop, String n2) { if (!n1.startsWith("http://xmlpull.org/v1/doc/")) return false; if (prop) return n1.substring(42).equals(n2); else return n1.substring(40).equals(n2); } private final boolean adjustNsp() throws Exception { boolean any = false; for (int i = 0; i < attributeCount << 2; i += 4) { // * 4 - 4; i >= 0; i -= 4) { String attrName = attributes[i + 2]; int cut = attrName.indexOf(':'); String prefix; if (cut != -1) { prefix = attrName.substring(0, cut); attrName = attrName.substring(cut + 1); } else if (attrName.equals("xmlns")) { prefix = attrName; attrName = null; } else continue; if (!prefix.equals("xmlns")) { any = true; } else { int j = (nspCounts[depth]++) << 1; nspStack = ensureCapacity(nspStack, j + 2); nspStack[j] = attrName; nspStack[j + 1] = attributes[i + 3]; if (attrName != null && attributes[i + 3].equals("")) error("illegal empty namespace"); // prefixMap = new PrefixMap (prefixMap, attrName, attr.getValue ()); //System.out.println (prefixMap); System.arraycopy( attributes, i + 4, attributes, i, ((--attributeCount) << 2) - i); i -= 4; } } if (any) { for (int i = (attributeCount << 2) - 4; i >= 0; i -= 4) { String attrName = attributes[i + 2]; int cut = attrName.indexOf(':'); if (cut == 0 && !relaxed) throw new RuntimeException( "illegal attribute name: " + attrName + " at " + this); else if (cut != -1) { String attrPrefix = attrName.substring(0, cut); attrName = attrName.substring(cut + 1); String attrNs = getNamespace(attrPrefix); if (attrNs == null && !relaxed) throw new RuntimeException( "Undefined Prefix: " + attrPrefix + " in " + this); attributes[i] = attrNs; attributes[i + 1] = attrPrefix; attributes[i + 2] = attrName; /* if (!relaxed) { for (int j = (attributeCount << 2) - 4; j > i; j -= 4) if (attrName.equals(attributes[j + 2]) && attrNs.equals(attributes[j])) exception( "Duplicate Attribute: {" + attrNs + "}" + attrName); } */ } } } int cut = name.indexOf(':'); if (cut == 0) error("illegal tag name: " + name); if (cut != -1) { prefix = name.substring(0, cut); name = name.substring(cut + 1); } this.namespace = getNamespace(prefix); if (this.namespace == null) { if (prefix != null) error("undefined prefix: " + prefix); this.namespace = NO_NAMESPACE; } return any; } private final String[] ensureCapacity(String[] arr, int required) { if (arr.length >= required) return arr; String[] bigger = new String[required + 16]; System.arraycopy(arr, 0, bigger, 0, arr.length); return bigger; } private final void error(String desc) throws Exception { if (relaxed) { if (error == null) error = "ERR: " + desc; } else exception(desc); } private final void exception(String desc) throws Exception { throw new Exception( desc.length() < 100 ? desc : desc.substring(0, 100) + "\n" + this); } /** * common base for next and nextToken. Clears the state, except from * txtPos and whitespace. Does not set the type variable */ private final void nextImpl() throws IOException, Exception { if (reader == null) exception("No Input specified"); if (type == END_TAG) depth--; while (true) { attributeCount = -1; // degenerated needs to be handled before error because of possible // processor expectations(!) if (degenerated) { degenerated = false; type = END_TAG; return; } if (error != null) { for (int i = 0; i < error.length(); i++) push(error.charAt(i)); // text = error; error = null; type = COMMENT; return; } if (relaxed && (stackMismatch > 0 || (peek(0) == -1 && depth > 0))) { int sp = (depth - 1) << 2; type = END_TAG; namespace = elementStack[sp]; prefix = elementStack[sp + 1]; name = elementStack[sp + 2]; if (stackMismatch != 1) error = "missing end tag /" + name + " inserted"; if (stackMismatch > 0) stackMismatch--; return; } prefix = null; name = null; namespace = null; // text = null; type = peekType(); switch (type) { case ENTITY_REF : pushEntity(); return; case START_TAG : parseStartTag(false); return; case END_TAG : parseEndTag(); return; case END_DOCUMENT : return; case TEXT : pushText('<', !token); if (depth == 0) { if (isWhitespace) type = IGNORABLE_WHITESPACE; // make exception switchable for instances.chg... !!!! // else // exception ("text '"+getText ()+"' not allowed outside root element"); } return; default : type = parseLegacy(token); if (type != XML_DECL) return; } } } private final int parseLegacy(boolean push) throws IOException, Exception { String req = ""; int term; int result; int prev = 0; read(); // < int c = read(); if (c == '?') { if ((peek(0) == 'x' || peek(0) == 'X') && (peek(1) == 'm' || peek(1) == 'M')) { if (push) { push(peek(0)); push(peek(1)); } read(); read(); if ((peek(0) == 'l' || peek(0) == 'L') && peek(1) <= ' ') { if (line != 1 || column > 4) error("PI must not start with xml"); parseStartTag(true); if (attributeCount < 1 || !"version".equals(attributes[2])) error("version expected"); version = attributes[3]; int pos = 1; if (pos < attributeCount && "encoding".equals(attributes[2 + 4])) { encoding = attributes[3 + 4]; pos++; } if (pos < attributeCount && "standalone".equals(attributes[4 * pos + 2])) { String st = attributes[3 + 4 * pos]; if ("yes".equals(st)) standalone = new Boolean(true); else if ("no".equals(st)) standalone = new Boolean(false); else error("illegal standalone value: " + st); pos++; } if (pos != attributeCount) error("illegal xmldecl"); isWhitespace = true; txtPos = 0; return XML_DECL; } } /* int c0 = read (); int c1 = read (); int */ term = '?'; result = PROCESSING_INSTRUCTION; } else if (c == '!') { if (peek(0) == '-') { result = COMMENT; req = "--"; term = '-'; } else if (peek(0) == '[') { result = CDSECT; req = "[CDATA["; term = ']'; push = true; } else { result = DOCDECL; req = "DOCTYPE"; term = -1; } } else { error("illegal: <" + c); return COMMENT; } for (int i = 0; i < req.length(); i++) read(req.charAt(i)); if (result == DOCDECL) parseDoctype(push); else { while (true) { c = read(); if (c == -1){ error(UNEXPECTED_EOF); return COMMENT; } if (push) push(c); if ((term == '?' || c == term) && peek(0) == term && peek(1) == '>') break; prev = c; } if (term == '-' && prev == '-') error("illegal comment delimiter: --->"); read(); read(); if (push && term != '?') txtPos--; } return result; } /** precondition: <! consumed */ private final void parseDoctype(boolean push) throws IOException, Exception { int nesting = 1; boolean quoted = false; // read(); while (true) { int i = read(); switch (i) { case -1 : error(UNEXPECTED_EOF); return; case '\'' : quoted = !quoted; break; case '<' : if (!quoted) nesting++; break; case '>' : if (!quoted) { if ((--nesting) == 0) return; } break; } if (push) push(i); } } /* precondition: </ consumed */ private final void parseEndTag() throws IOException, Exception { read(); // '<' read(); // '/' name = readName(); skip(); read('>'); int sp = (depth - 1) << 2; if (depth == 0) { error("element stack empty"); type = COMMENT; return; } if (!name.equals(elementStack[sp + 3])) { error("expected: /" + elementStack[sp + 3] + " read: " + name); // become case insensitive in relaxed mode int probe = sp; while (probe >= 0 && !name.toLowerCase().equals(elementStack[probe + 3].toLowerCase())) { stackMismatch++; probe -= 4; } if (probe < 0) { stackMismatch = 0; // text = "unexpected end tag ignored"; type = COMMENT; return; } } namespace = elementStack[sp]; prefix = elementStack[sp + 1]; name = elementStack[sp + 2]; } private final int peekType() throws IOException { switch (peek(0)) { case -1 : return END_DOCUMENT; case '&' : return ENTITY_REF; case '<' : switch (peek(1)) { case '/' : return END_TAG; case '?' : case '!' : return LEGACY; default : return START_TAG; } default : return TEXT; } } private final String get(int pos) { return new String(txtBuf, pos, txtPos - pos); } /* private final String pop (int pos) { String result = new String (txtBuf, pos, txtPos - pos); txtPos = pos; return result; } */ private final void push(int c) { isWhitespace &= c <= ' '; if (txtPos == txtBuf.length) { char[] bigger = new char[txtPos * 4 / 3 + 4]; System.arraycopy(txtBuf, 0, bigger, 0, txtPos); txtBuf = bigger; } txtBuf[txtPos++] = (char) c; } /** Sets name and attributes */ private final void parseStartTag(boolean xmldecl) throws IOException, Exception { if (!xmldecl) read(); name = readName(); attributeCount = 0; while (true) { skip(); int c = peek(0); if (xmldecl) { if (c == '?') { read(); read('>'); return; } } else { if (c == '/') { degenerated = true; read(); skip(); read('>'); break; } if (c == '>' && !xmldecl) { read(); break; } } if (c == -1) { error(UNEXPECTED_EOF); //type = COMMENT; return; } String attrName = readName(); if (attrName.length() == 0) { error("attr name expected"); //type = COMMENT; break; } int i = (attributeCount++) << 2; attributes = ensureCapacity(attributes, i + 4); attributes[i++] = ""; attributes[i++] = null; attributes[i++] = attrName; skip(); if (peek(0) != '=') { error("Attr.value missing f. "+attrName); attributes[i] = "1"; } else { read('='); skip(); int delimiter = peek(0); if (delimiter != '\'' && delimiter != '"') { error("attr value delimiter missing!"); delimiter = ' '; } else read(); int p = txtPos; pushText(delimiter, true); attributes[i] = get(p); txtPos = p; if (delimiter != ' ') read(); // skip endquote } } int sp = depth++ << 2; elementStack = ensureCapacity(elementStack, sp + 4); elementStack[sp + 3] = name; if (depth >= nspCounts.length) { int[] bigger = new int[depth + 4]; System.arraycopy(nspCounts, 0, bigger, 0, nspCounts.length); nspCounts = bigger; } nspCounts[depth] = nspCounts[depth - 1]; /* if(!relaxed){ for (int i = attributeCount - 1; i > 0; i--) { for (int j = 0; j < i; j++) { if (getAttributeName(i).equals(getAttributeName(j))) exception("Duplicate Attribute: " + getAttributeName(i)); } } } */ if (processNsp) adjustNsp(); else namespace = ""; elementStack[sp] = namespace; elementStack[sp + 1] = prefix; elementStack[sp + 2] = name; } /** result: isWhitespace; if the setName parameter is set, the name of the entity is stored in "name" */ private final void pushEntity() throws IOException, Exception { read(); // & int pos = txtPos; while (true) { int c = read(); if (c == ';') break; if (c < 128 && (c < '0' || c > '9') && (c < 'a' || c > 'z') && (c < 'A' || c > 'Z') && c != '_' && c != '-' && c != '#') { error("unterminated entity ref"); //; ends with:"+(char)c); if (c != -1) push(c); return; } push(c); } String code = get(pos); txtPos = pos; if (token && type == ENTITY_REF) name = code; if (code.charAt(0) == '#') { int c = (code.charAt(1) == 'x' ? Integer.parseInt(code.substring(2), 16) : Integer.parseInt(code.substring(1))); push(c); return; } String result = (String) entityMap.get(code); unresolved = result == null; if (unresolved) { if (!token) error("unresolved: &" + code + ";"); } else { for (int i = 0; i < result.length(); i++) push(result.charAt(i)); } } /** types: '<': parse to any token (for nextToken ()) '"': parse to quote ' ': parse to whitespace or '>' */ private final void pushText(int delimiter, boolean resolveEntities) throws IOException, Exception { int next = peek(0); int cbrCount = 0; while (next != -1 && next != delimiter) { // covers eof, '<', '"' if (delimiter == ' ') if (next <= ' ' || next == '>') break; if (next == '&') { if (!resolveEntities) break; pushEntity(); } else if (next == '\n' && type == START_TAG) { read(); push(' '); } else push(read()); if (next == '>' && cbrCount >= 2 && delimiter != ']') error("Illegal: ]]>"); if (next == ']') cbrCount++; else cbrCount = 0; next = peek(0); } } private final void read(char c) throws IOException, Exception { int a = read(); if (a != c) error("expected: '" + c + "' actual: '" + ((char) a) + "'"); } private final int read() throws IOException { int result; if (peekCount == 0) result = peek(0); else { result = peek[0]; peek[0] = peek[1]; } // else { // result = peek[0]; // System.arraycopy (peek, 1, peek, 0, peekCount-1); // } peekCount--; column++; if (result == '\n') { line++; column = 1; } return result; } /** Does never read more than needed */ private final int peek(int pos) throws IOException { while (pos >= peekCount) { int nw; if (srcBuf.length <= 1) nw = reader.read(); else if (srcPos < srcCount) nw = srcBuf[srcPos++]; else { srcCount = reader.read(srcBuf, 0, srcBuf.length); if (srcCount <= 0) nw = -1; else nw = srcBuf[0]; srcPos = 1; } if (nw == '\r') { wasCR = true; peek[peekCount++] = '\n'; } else { if (nw == '\n') { if (!wasCR) peek[peekCount++] = '\n'; } else peek[peekCount++] = nw; wasCR = false; } } return peek[pos]; } private final String readName() throws IOException, Exception { int pos = txtPos; int c = peek(0); if ((c < 'a' || c > 'z') && (c < 'A' || c > 'Z') && c != '_' && c != ':' && c < 0x0c0 && !relaxed) error("name expected"); do { push(read()); c = peek(0); } while ((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') || (c >= '0' && c <= '9') || c == '_' || c == '-' || c == ':' || c == '.' || c >= 0x0b7); String result = get(pos); txtPos = pos; return result; } private final void skip() throws IOException { while (true) { int c = peek(0); if (c > ' ' || c == -1) break; read(); } } //--------------- public part starts here... --------------- /** * Set the input source for parser to the given reader and * resets the parser. The event type is set to the initial value * START_DOCUMENT. * Setting the reader to null will just stop parsing and * reset parser state, * allowing the parser to free internal resources * such as parsing buffers. */ public void setInput(Reader reader) throws Exception { this.reader = reader; line = 1; column = 0; type = START_DOCUMENT; name = null; namespace = null; degenerated = false; attributeCount = -1; encoding = null; version = null; standalone = null; if (reader == null) return; srcPos = 0; srcCount = 0; peekCount = 0; depth = 0; entityMap = new Hashtable(); entityMap.put("amp", "&"); entityMap.put("apos", "'"); entityMap.put("gt", ">"); entityMap.put("lt", "<"); entityMap.put("quot", "\""); } /** * Sets the input stream the parser is going to process. * This call resets the parser state and sets the event type * to the initial value START_DOCUMENT. * *
NOTE: If an input encoding string is passed, * it MUST be used. Otherwise, * if inputEncoding is null, the parser SHOULD try to determine * input encoding following XML 1.0 specification (see below). * If encoding detection is supported then following feature * http://xmlpull.org/v1/doc/features.html#detect-encoding * MUST be true amd otherwise it must be false * * @param is contains a raw byte input stream of possibly * unknown encoding (when inputEncoding is null). * * @param _enc if not null it MUST be used as encoding for inputStream */ public void setInput(InputStream is, String _enc) throws Exception { srcPos = 0; srcCount = 0; String enc = _enc; if (is == null) throw new IllegalArgumentException(); try { if (enc == null) { // read four bytes int chk = 0; while (srcCount < 4) { int i = is.read(); if (i == -1) break; chk = (chk << 8) | i; srcBuf[srcCount++] = (char) i; } if (srcCount == 4) { switch (chk) { case 0x00000FEFF : enc = "UTF-32BE"; srcCount = 0; break; case 0x0FFFE0000 : enc = "UTF-32LE"; srcCount = 0; break; case 0x03c : enc = "UTF-32BE"; srcBuf[0] = '<'; srcCount = 1; break; case 0x03c000000 : enc = "UTF-32LE"; srcBuf[0] = '<'; srcCount = 1; break; case 0x0003c003f : enc = "UTF-16BE"; srcBuf[0] = '<'; srcBuf[1] = '?'; srcCount = 2; break; case 0x03c003f00 : enc = "UTF-16LE"; srcBuf[0] = '<'; srcBuf[1] = '?'; srcCount = 2; break; case 0x03c3f786d : while (true) { int i = is.read(); if (i == -1) break; srcBuf[srcCount++] = (char) i; if (i == '>') { String s = new String(srcBuf, 0, srcCount); int i0 = s.indexOf("encoding"); if (i0 != -1) { while (s.charAt(i0) != '"' && s.charAt(i0) != '\'') i0++; char deli = s.charAt(i0++); int i1 = s.indexOf(deli, i0); enc = s.substring(i0, i1); } break; } } default : if ((chk & 0x0ffff0000) == 0x0FEFF0000) { enc = "UTF-16BE"; srcBuf[0] = (char) ((srcBuf[2] << 8) | srcBuf[3]); srcCount = 1; } else if ((chk & 0x0ffff0000) == 0x0fffe0000) { enc = "UTF-16LE"; srcBuf[0] = (char) ((srcBuf[3] << 8) | srcBuf[2]); srcCount = 1; } else if ((chk & 0x0ffffff00) == 0x0EFBBBF) { enc = "UTF-8"; srcBuf[0] = srcBuf[3]; srcCount = 1; } } } } if (enc == null) enc = "UTF-8"; int sc = srcCount; setInput(new InputStreamReader(is, enc)); encoding = _enc; srcCount = sc; } catch (Exception e) { throw new Exception( "Invalid stream or encoding: " + e.toString() + ": " + this + ": " + e.toString()); } } /** * Returns the current value of the given feature. *
Please note: unknown features are * always returned as false. * * @param feature The name of feature to be retrieved. * @return The value of the feature. * @exception IllegalArgumentException if string the feature name is null */ public boolean getFeature(String feature) { if (FEATURE_PROCESS_NAMESPACES.equals(feature)) return processNsp; else if (isProp(feature, false, "relaxed")) return relaxed; else return false; } /** * Returns the input encoding if known, null otherwise. * If setInput(InputStream, inputEncoding) was called with an inputEncoding * value other than null, this value must be returned * from this method. Otherwise, if inputEncoding is null and * the parser suppports the encoding detection feature * (http://xmlpull.org/v1/doc/features.html#detect-encoding), * it must return the detected encoding. * If setInput(Reader) was called, null is returned. * After first call to next if XML declaration was present this method * will return encoding declared. */ public String getInputEncoding() { return encoding; } /** * Set new value for entity replacement text as defined in * XML 1.0 Section 4.5 * Construction of Internal Entity Replacement Text. * If FEATURE_PROCESS_DOCDECL or FEATURE_VALIDATION are set, calling this * function will result in an exception -- when processing of DOCDECL is * enabled, there is no need to the entity replacement text manually. * *
The motivation for this function is to allow very small * implementations of XMLPULL that will work in J2ME environments. * Though these implementations may not be able to process the document type * declaration, they still can work with known DTDs by using this function. * *
Please notes: The given value is used literally as replacement text * and it corresponds to declaring entity in DTD that has all special characters * escaped: left angle bracket is replaced with <, ampersnad with & * and so on. * *
Note: The given value is the literal replacement text and must not * contain any other entity reference (if it contains any entity reference * there will be no further replacement). * *
Note: The list of pre-defined entity names will * always contain standard XML entities such as * amp (&), lt (<), gt (>), quot ("), and apos ('). * Those cannot be redefined by this method! * * @see #setInput * @see #FEATURE_PROCESS_DOCDECL * @see #FEATURE_VALIDATION */ public void defineEntityReplacementText(String entity, String value) throws Exception { if (entityMap == null) throw new RuntimeException("entity replacement text must be defined after setInput!"); entityMap.put(entity, value); } /** * Look up the value of a property. * * The property name is any fully-qualified URI. *
NOTE: unknown properties are always * returned as null. * * @param property The name of property to be retrieved. * @return The value of named property. */ public Object getProperty(String property) { if (isProp(property, true, "xmldecl-version")) return version; if (isProp(property, true, "xmldecl-standalone")) return standalone; if (isProp(property, true, "location")) return location != null ? location : reader.toString(); return null; } /** * Returns the numbers of elements in the namespace stack for the given * depth. * If namespaces are not enabled, 0 is returned. * *
NOTE: when parser is on END_TAG then it is allowed to call * this function with getDepth()+1 argument to retrieve position of namespace * prefixes and URIs that were declared on corresponding START_TAG. *
NOTE: to retrieve lsit of namespaces declared in current element:
* TibiXmlParser pp = ...
* int nsStart = pp.getNamespaceCount(pp.getDepth()-1);
* int nsEnd = pp.getNamespaceCount(pp.getDepth());
* for (int i = nsStart; i < nsEnd; i++) {
* String prefix = pp.getNamespacePrefix(i);
* String ns = pp.getNamespaceUri(i);
* // ...
* }
*
*
* @see #getNamespacePrefix
* @see #getNamespaceUri
* @see #getNamespace()
* @see #getNamespace(String)
*/
public int getNamespaceCount(int depth) {
if (depth > this.depth)
throw new IndexOutOfBoundsException();
return nspCounts[depth];
}
/**
* Returns the namespace prefixe for the given position
* in the namespace stack.
* Default namespace declaration (xmlns='...') will have null as prefix.
* If the given index is out of range, an exception is thrown.
* Please note: when the parser is on an END_TAG, * namespace prefixes that were declared * in the corresponding START_TAG are still accessible * although they are no longer in scope. */ public String getNamespacePrefix(int pos) { return nspStack[pos << 1]; } /** * Returns the namespace URI for the given position in the * namespace stack * If the position is out of range, an exception is thrown. *
NOTE: when parser is on END_TAG then namespace prefixes that were declared * in corresponding START_TAG are still accessible even though they are not in scope */ public String getNamespaceUri(int pos) { return nspStack[(pos << 1) + 1]; } /** * Returns the URI corresponding to the given prefix, * depending on current state of the parser. * *
If the prefix was not declared in the current scope, * null is returned. The default namespace is included * in the namespace table and is available via * getNamespace (null). * *
This method is a convenience method for * *
* for (int i = getNamespaceCount(getDepth ())-1; i >= 0; i--) {
* if (getNamespacePrefix(i).equals( prefix )) {
* return getNamespaceUri(i);
* }
* }
* return null;
*
*
* Please note: parser implementations * may provide more efifcient lookup, e.g. using a Hashtable. * The 'xml' prefix is bound to "http://www.w3.org/XML/1998/namespace", as * defined in the * Namespaces in XML * specification. Analogous, the 'xmlns' prefix is resolved to * http://www.w3.org/2000/xmlns/ * * @see #getNamespaceCount * @see #getNamespacePrefix * @see #getNamespaceUri */ public String getNamespace(String prefix) { if ("xml".equals(prefix)) return "http://www.w3.org/XML/1998/namespace"; if ("xmlns".equals(prefix)) return "http://www.w3.org/2000/xmlns/"; for (int i = (getNamespaceCount(depth) << 1) - 2; i >= 0; i -= 2) { if (prefix == null) { if (nspStack[i] == null) return nspStack[i + 1]; } else if (prefix.equals(nspStack[i])) return nspStack[i + 1]; } return null; } /** * Returns the current depth of the element. * Outside the root element, the depth is 0. The * depth is incremented by 1 when a start tag is reached. * The depth is decremented AFTER the end tag * event was observed. * *
* <!-- outside --> 0
* <root> 1
* sometext 1
* <foobar> 2
* </foobar> 2
* </root> 1
* <!-- outside --> 0
*
*/
public int getDepth() {
return depth;
}
/**
* Returns a short text describing the current parser state, including
* the position, a
* description of the current event and the data source if known.
* This method is especially useful to provide meaningful
* error messages and for debugging purposes.
*/
public String getPositionDescription() {
StringBuffer buf =
new StringBuffer(type < TYPES.length ? TYPES[type] : "unknown");
buf.append(' ');
if (type == START_TAG || type == END_TAG) {
if (degenerated)
buf.append("(empty) ");
buf.append('<');
if (type == END_TAG)
buf.append('/');
if (prefix != null)
buf.append("{" + namespace + "}" + prefix + ":");
buf.append(name);
int cnt = attributeCount << 2;
for (int i = 0; i < cnt; i += 4) {
buf.append(' ');
if (attributes[i + 1] != null)
buf.append(
"{" + attributes[i] + "}" + attributes[i + 1] + ":");
buf.append(attributes[i + 2] + "='" + attributes[i + 3] + "'");
}
buf.append('>');
}
else if (type == IGNORABLE_WHITESPACE);
else if (type != TEXT)
buf.append(getText());
else if (isWhitespace)
buf.append("(whitespace)");
else {
String text = getText();
if (text.length() > 16)
text = text.substring(0, 16) + "...";
buf.append(text);
}
buf.append("@"+line + ":" + column);
buf.append(" in ");
buf.append(location == null ? reader.toString() : location);
return buf.toString();
}
/**
* Returns the current line number, starting from 1.
* When the parser does not know the current line number
* or can not determine it, -1 is returned (e.g. for WBXML).
*
* @return current line number or -1 if unknown.
*/
public int getLineNumber() {
return line;
}
/**
* Returns the current column number, starting from 0.
* When the parser does not know the current column number
* or can not determine it, -1 is returned (e.g. for WBXML).
*
* @return current column number or -1 if unknown.
*/
public int getColumnNumber() {
return column;
}
/**
* Checks whether the current TEXT event contains only whitespace
* characters.
* For IGNORABLE_WHITESPACE, this is always true.
* For TEXT and CDSECT, false is returned when the current event text
* contains at least one non-white space character. For any other
* event type an exception is thrown.
*
* Please note: non-validating parsers are not * able to distinguish whitespace and ignorable whitespace, * except from whitespace outside the root element. Ignorable * whitespace is reported as separate event, which is exposed * via nextToken only. * */ public boolean isWhitespace() throws Exception { if (type != TEXT && type != IGNORABLE_WHITESPACE && type != CDSECT) exception(ILLEGAL_TYPE); return isWhitespace; } /** * Returns the text content of the current event as String. * The value returned depends on current event type, * for example for TEXT event it is element content * (this is typical case when next() is used). * * See description of nextToken() for detailed description of * possible returned values for different types of events. * *
NOTE: in case of ENTITY_REF, this method returns * the entity replacement text (or null if not available). This is * the only case where * getText() and getTextCharacters() return different values. * * @see #getEventType * @see #next * @see #nextToken */ public String getText() { return type < TEXT || (type == ENTITY_REF && unresolved) ? null : get(0); } /** * Returns the buffer that contains the text of the current event, * as well as the start offset and length relevant for the current * event. See getText(), next() and nextToken() for description of possible returned values. * *
Please note: this buffer must not * be modified and its content MAY change after a call to * next() or nextToken(). This method will always return the * same value as getText(), except for ENTITY_REF. In the case * of ENTITY ref, getText() returns the replacement text and * this method returns the actual input buffer containing the * entity name. * If getText() returns null, this method returns null as well and * the values returned in the holder array MUST be -1 (both start * and length). * * @see #getText * @see #next * @see #nextToken * * @param poslen Must hold an 2-element int array * into which the start offset and length values will be written. * @return char buffer that contains the text of the current event * (null if the current event has no text associated). */ public char[] getTextCharacters(int[] poslen) { if (type >= TEXT) { if (type == ENTITY_REF) { poslen[0] = 0; poslen[1] = name.length(); return name.toCharArray(); } poslen[0] = 0; poslen[1] = txtPos; return txtBuf; } poslen[0] = -1; poslen[1] = -1; return null; } /** * Returns the namespace URI of the current element. * The default namespace is represented * as empty string. * If namespaces are not enabled, an empty String ("") is always returned. * The current event must be START_TAG or END_TAG; otherwise, * null is returned. */ public String getNamespace() { return namespace; } /** * For START_TAG or END_TAG events, the (local) name of the current * element is returned when namespaces are enabled. When namespace * processing is disabled, the raw name is returned. * For ENTITY_REF events, the entity name is returned. * If the current event is not START_TAG, END_TAG, or ENTITY_REF, * null is returned. *
Please note: To reconstruct the raw element name * when namespaces are enabled and the prefix is not null, * you will need to add the prefix and a colon to localName.. * */ public String getName() { return name; } /** * Returns the prefix of the current element. * If the element is in the default namespace (has no prefix), * null is returned. * If namespaces are not enabled, or the current event * is not START_TAG or END_TAG, null is returned. */ public String getPrefix() { return prefix; } /** * Returns true if the current event is START_TAG and the tag * is degenerated * (e.g. <foobar/>). *
NOTE: if the parser is not on START_TAG, an exception * will be thrown. */ public boolean isEmptyElementTag() throws Exception { if (type != START_TAG) exception(ILLEGAL_TYPE); return degenerated; } /** * Returns the number of attributes of the current start tag, or * -1 if the current event type is not START_TAG * * @see #getAttributeNamespace * @see #getAttributeName * @see #getAttributePrefix * @see #getAttributeValue */ public int getAttributeCount() { return attributeCount; } /** * Returns the type of the specified attribute * If parser is non-validating it MUST return CDATA. * * @param index zero based index of attribute * @return attribute type (null is never returned) */ public String getAttributeType(int index) { return "CDATA"; } /** * Returns if the specified attribute was not in input was declared in XML. * If parser is non-validating it MUST always return false. * This information is part of XML infoset: * * @param index zero based index of attribute * @return false if attribute was in input */ public boolean isAttributeDefault(int index) { return false; } /** * Returns the namespace URI of the attribute * with the given index (starts from 0). * Returns an empty string ("") if namespaces are not enabled * or the attribute has no namespace. * Throws an IndexOutOfBoundsException if the index is out of range * or the current event type is not START_TAG. * *
NOTE: if FEATURE_REPORT_NAMESPACE_ATTRIBUTES is set * then namespace attributes (xmlns:ns='...') must be reported * with namespace * http://www.w3.org/2000/xmlns/ * (visit this URL for description!). * The default namespace attribute (xmlns="...") will be reported with empty namespace. *
NOTE:The xml prefix is bound as defined in * Namespaces in XML * specification to "http://www.w3.org/XML/1998/namespace". * * @param index zero based index of attribute * @return attribute namespace, * empty string ("") is returned if namesapces processing is not enabled or * namespaces processing is enabled but attribute has no namespace (it has no prefix). */ public String getAttributeNamespace(int index) { if (index >= attributeCount) throw new IndexOutOfBoundsException(); return attributes[index << 2]; } /** * Returns the local name of the specified attribute * if namespaces are enabled or just attribute name if namespaces are disabled. * Throws an IndexOutOfBoundsException if the index is out of range * or current event type is not START_TAG. * * @param index zero based index of attribute * @return attribute name (null is never returned) */ public String getAttributeName(int index) { if (index >= attributeCount) throw new IndexOutOfBoundsException(); return attributes[(index << 2) + 2]; } /** * Returns the prefix of the specified attribute * Returns null if the element has no prefix. * If namespaces are disabled it will always return null. * Throws an IndexOutOfBoundsException if the index is out of range * or current event type is not START_TAG. * * @param index zero based index of attribute * @return attribute prefix or null if namespaces processing is not enabled. */ public String getAttributePrefix(int index) { if (index >= attributeCount) throw new IndexOutOfBoundsException(); return attributes[(index << 2) + 1]; } /** * Returns the given attributes value. * Throws an IndexOutOfBoundsException if the index is out of range * or current event type is not START_TAG. * *
NOTE: attribute value must be normalized * (including entity replacement text if PROCESS_DOCDECL is false) as described in * XML 1.0 section * 3.3.3 Attribute-Value Normalization * * @see #defineEntityReplacementText * * @param index zero based index of attribute * @return value of attribute (null is never returned) */ public String getAttributeValue(int index) { if (index >= attributeCount) throw new IndexOutOfBoundsException(); return attributes[(index << 2) + 3]; } /** * Returns the attributes value identified by namespace URI and namespace localName. * If namespaces are disabled namespace must be null. * If current event type is not START_TAG then IndexOutOfBoundsException will be thrown. * *
NOTE: attribute value must be normalized * (including entity replacement text if PROCESS_DOCDECL is false) as described in * XML 1.0 section * 3.3.3 Attribute-Value Normalization * * @see #defineEntityReplacementText * * @param namespace Namespace of the attribute if namespaces are enabled otherwise must be null * @param name If namespaces enabled local name of attribute otherwise just attribute name * @return value of attribute or null if attribute with given name does not exist */ public String getAttributeValue(String namespace, String name) { for (int i = (attributeCount << 2) - 4; i >= 0; i -= 4) { if (attributes[i + 2].equals(name) && (namespace == null || attributes[i].equals(namespace))) return attributes[i + 3]; } return null; } /** * Returns the type of the current event (START_TAG, END_TAG, TEXT, etc.) * * @see #next() * @see #nextToken() */ public int getEventType() throws Exception { return type; } /** * Get next parsing event - element content wil be coalesced and only one * TEXT event must be returned for whole element content * (comments and processing instructions will be ignored and emtity references * must be expanded or exception mus be thrown if entity reerence can not be exapnded). * If element content is empty (content is "") then no TEXT event will be reported. * *
NOTE: empty element (such as <tag/>) will be reported * with two separate events: START_TAG, END_TAG - it must be so to preserve * parsing equivalency of empty element to <tag></tag>. * (see isEmptyElementTag ()) * * @see #isEmptyElementTag * @see #START_TAG * @see #TEXT * @see #END_TAG * @see #END_DOCUMENT */ public int next() throws Exception, IOException { txtPos = 0; isWhitespace = true; int minType = 9999; token = false; do { nextImpl(); if (type < minType) minType = type; // if (curr <= TEXT) type = curr; } while (minType > ENTITY_REF // ignorable || (minType >= TEXT && peekType() >= TEXT)); type = minType; if (type > TEXT) type = TEXT; return type; } /** * This method works similarly to next() but will expose * additional event types (COMMENT, CDSECT, DOCDECL, ENTITY_REF, PROCESSING_INSTRUCTION, or * IGNORABLE_WHITESPACE) if they are available in input. * *
If special feature * FEATURE_XML_ROUNDTRIP * (identified by URI: http://xmlpull.org/v1/doc/features.html#xml-roundtrip) * is enabled it is possible to do XML document round trip ie. reproduce * exectly on output the XML input using getText(): * returned content is always unnormalized (exactly as in input). * Otherwise returned content is end-of-line normalized as described * XML 1.0 End-of-Line Handling * and. Also when this feature is enabled exact content of START_TAG, END_TAG, * DOCDECL and PROCESSING_INSTRUCTION is available. * *
Here is the list of tokens that can be returned from nextToken() * and what getText() and getTextCharacters() returns:
* " titlepage SYSTEM "http://www.foo.bar/dtds/typo.dtd"
* [<!ENTITY % active.links "INCLUDE">]"
* for input document that contained:
* <!DOCTYPE titlepage SYSTEM "http://www.foo.bar/dtds/typo.dtd"
* [<!ENTITY % active.links "INCLUDE">]>
* otherwise if FEATURE_XML_ROUNDTRIP is false and PROCESS_DOCDECL is true
* then what is returned is undefined (it may be even null)
* NOTE: there is no gurantee that there will only one TEXT or * IGNORABLE_WHITESPACE event from nextToken() as parser may chose to deliver element content in * multiple tokens (dividing element content into chunks) * *
NOTE: whether returned text of token is end-of-line normalized * is depending on FEATURE_XML_ROUNDTRIP. * *
NOTE: XMLDecl (<?xml ...?>) is not reported but its content * is available through optional properties (see class description above). * * @see #next * @see #START_TAG * @see #TEXT * @see #END_TAG * @see #END_DOCUMENT * @see #COMMENT * @see #DOCDECL * @see #PROCESSING_INSTRUCTION * @see #ENTITY_REF * @see #IGNORABLE_WHITESPACE */ public int nextToken() throws Exception, IOException { isWhitespace = true; txtPos = 0; token = true; nextImpl(); return type; } //---------------------------------------------------------------------- // utility methods to make XML parsing easier ... /** * Call next() and return event if it is START_TAG or END_TAG * otherwise throw an exception. * It will skip whitespace TEXT before actual tag if any. * *
essentially it does this *
* int eventType = next();
* if(eventType == TEXT && isWhitespace()) { // skip whitespace
* eventType = next();
* }
* if (eventType != START_TAG && eventType != END_TAG) {
* throw new Exception("expected start or end tag", this, null);
* }
* return eventType;
*
*/
public int nextTag() throws Exception, IOException {
next();
if (type == TEXT && isWhitespace)
next();
if (type != END_TAG && type != START_TAG)
exception("unexpected type");
return type;
}
/**
* Test if the current event is of the given type and if the
* namespace and name do match. null will match any namespace
* and any name. If the test is not passed, an exception is
* thrown. The exception text indicates the parser position,
* the expected event and the current event that is not meeting the
* requirement.
*
* Essentially it does this *
* if (type != getEventType()
* || (namespace != null && !namespace.equals( getNamespace () ) )
* || (name != null && !name.equals( getName() ) ) )
* throw new Exception( "expected "+ TYPES[ type ]+getPositionDescription());
*
*/
public void require(int type, String namespace, String name)
throws Exception, IOException {
if (type != this.type
|| (namespace != null && !namespace.equals(getNamespace()))
|| (name != null && !name.equals(getName())))
exception(
"expected: " + TYPES[type] + " {" + namespace + "}" + name);
}
/**
* If current event is START_TAG then if next element is TEXT then element content is returned
* or if next event is END_TAG then empty string is returned, otherwise exception is thrown.
* After calling this function successfully parser will be positioned on END_TAG.
*
* The motivation for this function is to allow to parse consistently both * empty elements and elements that has non empty content, for example for input:
* p.nextTag()
* p.requireEvent(p.START_TAG, "", "tag");
* String content = p.nextText();
* p.requireEvent(p.END_TAG, "", "tag");
*
* This function together with nextTag make it very easy to parse XML that has
* no mixed content.
*
*
* Essentially it does this *
* if(getEventType() != START_TAG) {
* throw new Exception(
* "parser must be on START_TAG to read next text", this, null);
* }
* int eventType = next();
* if(eventType == TEXT) {
* String result = getText();
* eventType = next();
* if(eventType != END_TAG) {
* throw new Exception(
* "event TEXT it must be immediately followed by END_TAG", this, null);
* }
* return result;
* } else if(eventType == END_TAG) {
* return "";
* } else {
* throw new Exception(
* "parser must be on START_TAG or TEXT to read text", this, null);
* }
*
*/
public String nextText() throws Exception, IOException {
if (type != START_TAG)
exception("precondition: START_TAG");
next();
String result;
if (type == TEXT) {
result = getText();
next();
}
else
result = "";
if (type != END_TAG)
exception("END_TAG expected");
return result;
}
/**
* Use this call to change the general behaviour of the parser,
* such as namespace processing or doctype declaration handling.
* This method must be called before the first call to next or
* nextToken. Otherwise, an exception is thrown.
* Example: call setFeature(FEATURE_PROCESS_NAMESPACES, true) in order
* to switch on namespace processing. The initial settings correspond
* to the properties requested from the XML Pull Parser factory.
* If none were requested, all feautures are deactivated by default.
*
* @exception Exception If the feature is not supported or can not be set
* @exception IllegalArgumentException If string with the feature name is null
*/
public void setFeature(String feature, boolean value)
throws Exception {
if (FEATURE_PROCESS_NAMESPACES.equals(feature))
processNsp = value;
else if (isProp(feature, false, "relaxed"))
relaxed = value;
else
exception("unsupported feature: " + feature);
}
/**
* Set the value of a property.
*
* The property name is any fully-qualified URI.
*
* @exception Exception If the property is not supported or can not be set
* @exception IllegalArgumentException If string with the property name is null
*/
public void setProperty(String property, Object value)
throws Exception {
if(isProp(property, true, "location"))
location = value;
else
throw new Exception("unsupported property: " + property);
}
/**
* Skip sub tree that is currently porser positioned on.
*
NOTE: parser must be on START_TAG and when funtion returns
* parser will be positioned on corresponding END_TAG.
*/
// Implementation copied from Alek's mail...
public void skipSubTree() throws Exception, IOException {
require(START_TAG, null, null);
int level = 1;
while (level > 0) {
int eventType = next();
if (eventType == END_TAG) {
--level;
}
else if (eventType == START_TAG) {
++level;
}
}
}
}