public Page parse(SitemeshBuffer buffer) throws IOException {
CharArrayReader reader =
new CharArrayReader(buffer.getCharArray(), 0, buffer.getBufferLength());
CharArray _buffer = new CharArray(4096);
CharArray _body = new CharArray(4096);
CharArray _head = new CharArray(512);
CharArray _title = new CharArray(128);
Map _htmlProperties = null;
Map _metaProperties = new HashMap(6);
Map _sitemeshProperties = new HashMap(6);
Map _bodyProperties = null;
CharArray _currentTaggedContent = new CharArray(1024);
String _contentTagId = null;
boolean tagged = false;
boolean _frameSet = false;
int _state = STATE_TEXT;
int _tokenType = TOKEN_NONE;
int _pushBack = 0;
int _comment = 0;
int _quote = 0;
boolean hide = false;
int state = TAG_STATE_NONE;
int laststate = TAG_STATE_NONE;
boolean doneTitle = false;
// This tag object gets reused each iteration.
Tag tagObject = new Tag();
while (_tokenType != TOKEN_EOF) {
if (tagged) {
if (_tokenType == TOKEN_TAG || _tokenType == TOKEN_EMPTYTAG) {
if (_buffer == null || _buffer.length() == 0) {
_tokenType = TOKEN_NONE;
continue;
}
if (parseTag(tagObject, _buffer) == null) continue;
if (_buffer.compareLowerSubstr(
"/content")) // Note that the '/' survives the | 32 operation
{
tagged = false;
if (_contentTagId != null) {
state = TAG_STATE_NONE;
_sitemeshProperties.put(_contentTagId, _currentTaggedContent.toString());
_currentTaggedContent.setLength(0);
_contentTagId = null;
}
} else {
_currentTaggedContent.append('<').append(_buffer).append('>');
}
} else {
if (_buffer.length() > 0) _currentTaggedContent.append(_buffer);
}
} else {
if (_tokenType == TOKEN_TAG || _tokenType == TOKEN_EMPTYTAG) {
if (_buffer == null || _buffer.length() == 0) {
_tokenType = TOKEN_NONE;
continue;
}
if (parseTag(tagObject, _buffer) == null) {
_tokenType = TOKEN_TEXT;
continue;
}
int tagHash = _buffer.substrHashCode();
if (state == TAG_STATE_XML || state == TAG_STATE_XMP) {
writeTag(state, laststate, hide, _head, _buffer, _body);
if ((state == TAG_STATE_XML && tagHash == SLASH_XML_HASH)
|| (state == TAG_STATE_XMP && tagHash == SLASH_XMP_HASH)) {
state = laststate;
}
} else {
boolean doDefault = false;
switch (tagHash) {
case HTML_HASH:
if (!_buffer.compareLowerSubstr("html")) { // skip any accidental hash collisions
doDefault = true;
break;
}
state = TAG_STATE_HTML;
_htmlProperties = parseProperties(tagObject, _buffer).properties;
break;
case HEAD_HASH:
if (!_buffer.compareLowerSubstr("head")) { // skip any accidental hash collisions
doDefault = true;
break;
}
state = TAG_STATE_HEAD;
break;
case XML_HASH:
if (!_buffer.compareLowerSubstr("xml")) { // skip any accidental hash collisions
doDefault = true;
break;
}
laststate = state;
writeTag(state, laststate, hide, _head, _buffer, _body);
state = TAG_STATE_XML;
break;
case XMP_HASH:
if (!_buffer.compareLowerSubstr("xmp")) { // skip any accidental hash collisions
doDefault = true;
break;
}
laststate = state;
writeTag(state, laststate, hide, _head, _buffer, _body);
state = TAG_STATE_XMP;
break;
case TITLE_HASH:
if (!_buffer.compareLowerSubstr("title")) { // skip any accidental hash collisions
doDefault = true;
break;
}
if (doneTitle) {
hide = true;
} else {
laststate = state;
state = TAG_STATE_TITLE;
}
break;
case SLASH_TITLE_HASH:
if (!_buffer.compareLowerSubstr("/title")) { // skip any accidental hash collisions
doDefault = true;
break;
}
if (doneTitle) {
hide = false;
} else {
doneTitle = true;
state = laststate;
}
break;
case PARAMETER_HASH:
if (!_buffer.compareLowerSubstr(
"parameter")) { // skip any accidental hash collisions
doDefault = true;
break;
}
parseProperties(tagObject, _buffer);
String name = (String) tagObject.properties.get("name");
String value = (String) tagObject.properties.get("value");
if (name != null && value != null) {
_sitemeshProperties.put(name, value);
}
break;
case META_HASH:
if (!_buffer.compareLowerSubstr("meta")) { // skip any accidental hash collisions
doDefault = true;
break;
}
CharArray metaDestination = state == TAG_STATE_HEAD ? _head : _body;
metaDestination.append('<');
metaDestination.append(_buffer);
metaDestination.append('>');
parseProperties(tagObject, _buffer);
name = (String) tagObject.properties.get("name");
value = (String) tagObject.properties.get("content");
if (name == null) {
String httpEquiv = (String) tagObject.properties.get("http-equiv");
if (httpEquiv != null) {
name = "http-equiv." + httpEquiv;
}
}
if (name != null && value != null) {
_metaProperties.put(name, value);
}
break;
case SLASH_HEAD_HASH:
if (!_buffer.compareLowerSubstr("/head")) { // skip any accidental hash collisions
doDefault = true;
break;
}
state = TAG_STATE_HTML;
break;
case FRAME_HASH:
if (!_buffer.compareLowerSubstr("frame")) { // skip any accidental hash collisions
doDefault = true;
break;
}
_frameSet = true;
break;
case FRAMESET_HASH:
if (!_buffer.compareLowerSubstr(
"frameset")) { // skip any accidental hash collisions
doDefault = true;
break;
}
_frameSet = true;
break;
case BODY_HASH:
if (!_buffer.compareLowerSubstr("body")) { // skip any accidental hash collisions
doDefault = true;
break;
}
if (_tokenType == TOKEN_EMPTYTAG) {
state = TAG_STATE_BODY;
}
_bodyProperties = parseProperties(tagObject, _buffer).properties;
break;
case CONTENT_HASH:
if (!_buffer.compareLowerSubstr("content")) { // skip any accidental hash collisions
doDefault = true;
break;
}
state = TAG_STATE_NONE;
Map props = parseProperties(tagObject, _buffer).properties;
if (props != null) {
tagged = true;
_contentTagId = (String) props.get("tag");
}
break;
case SLASH_XMP_HASH:
if (!_buffer.compareLowerSubstr("/xmp")) { // skip any accidental hash collisions
doDefault = true;
break;
}
hide = false;
break;
case SLASH_BODY_HASH:
if (!_buffer.compareLowerSubstr("/body")) { // skip any accidental hash collisions
doDefault = true;
break;
}
state = TAG_STATE_NONE;
hide = true;
break;
case SLASH_HTML_HASH:
if (!_buffer.compareLowerSubstr("/html")) { // skip any accidental hash collisions
doDefault = true;
break;
}
state = TAG_STATE_NONE;
hide = true;
break;
default:
doDefault = true;
}
if (doDefault) writeTag(state, laststate, hide, _head, _buffer, _body);
}
} else if (!hide) {
if (_tokenType == TOKEN_TEXT) {
if (state == TAG_STATE_TITLE) {
_title.append(_buffer);
} else if (shouldWriteToHead(state, laststate)) {
_head.append(_buffer);
} else {
_body.append(_buffer);
}
} else if (_tokenType == TOKEN_COMMENT) {
final CharArray commentDestination =
shouldWriteToHead(state, laststate) ? _head : _body;
commentDestination.append("<!--");
commentDestination.append(_buffer);
commentDestination.append("-->");
} else if (_tokenType == TOKEN_CDATA) {
final CharArray commentDestination = state == TAG_STATE_HEAD ? _head : _body;
commentDestination.append("<![CDATA[");
commentDestination.append(_buffer);
commentDestination.append("]]>");
} else if (_tokenType == TOKEN_SCRIPT) {
final CharArray commentDestination = state == TAG_STATE_HEAD ? _head : _body;
commentDestination.append('<');
commentDestination.append(_buffer);
}
}
}
_buffer.setLength(0);
start:
while (true) {
int c;
if (_pushBack != 0) {
c = _pushBack;
_pushBack = 0;
} else {
try {
c = reader.read();
} catch (IOException e) {
_tokenType = TOKEN_EOF;
break start;
}
}
if (c < 0) {
int tmpstate = _state;
_state = STATE_EOF;
if (_buffer.length() > 0 && tmpstate == STATE_TEXT) {
_tokenType = TOKEN_TEXT;
break start;
} else {
_tokenType = TOKEN_EOF;
break start;
}
}
switch (_state) {
case STATE_TAG:
{
int buflen = _buffer.length();
if (c == '>') {
if (_buffer.length() > 1 && _buffer.charAt(_buffer.length() - 1) == '/') {
_tokenType = TOKEN_EMPTYTAG;
} else {
_tokenType = TOKEN_TAG;
}
_state = STATE_TEXT;
break start;
} else if (c == '/') {
_buffer.append('/');
} else if (c == '<' && buflen == 0) {
_buffer.append("<<");
_state = STATE_TEXT;
} else if (c == '-'
&& buflen == 2
&& _buffer.charAt(1) == '-'
&& _buffer.charAt(0) == '!') {
_buffer.setLength(0);
_state = STATE_COMMENT;
} else if (c == '['
&& buflen == 7
&& _buffer.charAt(0) == '!'
&& _buffer.charAt(1) == '['
&& _buffer.compareLower("cdata", 2)) {
_buffer.setLength(0);
_state = STATE_CDATA;
} else if ((c == 'e' || c == 'E')
&& buflen == 7
&& _buffer.charAt(0) == '!'
&& _buffer.compareLower("doctyp", 1)) {
_buffer.append((char) c);
_state = STATE_DOCTYPE;
} else if ((c == 'T' || c == 't')
&& buflen == 5
&& _buffer.compareLower("scrip", 0)) {
_buffer.append((char) c);
_state = STATE_SCRIPT;
} else if (c == '"' || c == '\'') {
_quote = c;
_buffer.append((char) c);
_state = STATE_TAG_QUOTE;
} else {
_buffer.append((char) c);
}
}
break;
case STATE_TEXT:
{
if (c == '<') {
_state = STATE_TAG;
if (_buffer.length() > 0) {
_tokenType = TOKEN_TEXT;
break start;
}
} else {
_buffer.append((char) c);
}
}
break;
case STATE_TAG_QUOTE:
{
if (c == '>') {
_pushBack = c;
_state = STATE_TAG;
} else {
_buffer.append((char) c);
if (c == _quote) {
_state = STATE_TAG;
}
}
}
break;
case STATE_COMMENT:
{
if (c == '>' && _comment >= 2) {
_buffer.setLength(_buffer.length() - 2);
_comment = 0;
_state = STATE_TEXT;
_tokenType = TOKEN_COMMENT;
break start;
} else if (c == '-') {
_comment++;
} else {
_comment = 0;
}
_buffer.append((char) c);
}
break;
case STATE_CDATA:
{
if (c == '>' && _comment >= 2) {
_buffer.setLength(_buffer.length() - 2);
_comment = 0;
_state = STATE_TEXT;
_tokenType = TOKEN_CDATA;
break start;
} else if (c == ']') {
_comment++;
} else {
_comment = 0;
}
_buffer.append((char) c);
}
break;
case STATE_SCRIPT:
{
_buffer.append((char) c);
if (c == '<') {
_comment = 0;
} else if ((c == '/' && _comment == 0)
|| ((c == 's' || c == 'S') && _comment == 1)
|| ((c == 'c' || c == 'C') && _comment == 2)
|| ((c == 'r' || c == 'R') && _comment == 3)
|| ((c == 'i' || c == 'I') && _comment == 4)
|| ((c == 'p' || c == 'P') && _comment == 5)
|| ((c == 't' || c == 'T') && _comment == 6)) {
_comment++;
} else if (c == '>' && _comment >= 7) {
_comment = 0;
_state = STATE_TEXT;
_tokenType = TOKEN_SCRIPT;
break start;
}
}
break;
case STATE_DOCTYPE:
{
_buffer.append((char) c);
if (c == '>') {
_state = STATE_TEXT;
_tokenType = TOKEN_DOCTYPE;
break start;
} else {
_comment = 0;
}
}
break;
}
}
}
// Help the GC
_currentTaggedContent = null;
_buffer = null;
return new FastPage(
buffer,
_sitemeshProperties,
_htmlProperties,
_metaProperties,
_bodyProperties,
_title.toString().trim(),
_head.toString().trim(),
_body.toString().trim(),
_frameSet);
}
