/*
  Copyright (c) 2002, 2014, Oracle and/or its affiliates. All rights reserved.

  The MySQL Connector/J is licensed under the terms of the GPLv2
  <http://www.gnu.org/licenses/old-licenses/gpl-2.0.html>, like most MySQL Connectors.
  There are special exceptions to the terms and conditions of the GPLv2 as it is applied to
  this software, see the FLOSS License Exception
  <http://www.mysql.com/about/legal/licensing/foss-exception.html>.

  This program is free software; you can redistribute it and/or modify it under the terms
  of the GNU General Public License as published by the Free Software Foundation; version 2
  of the License.

  This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;
  without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
  See the GNU General Public License for more details.

  You should have received a copy of the GNU General Public License along with this
  program; if not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth
  Floor, Boston, MA 02110-1301  USA

 */

package com.mysql.jdbc;

/**
 * EscapeTokenizer breaks up an SQL statement into SQL and escape code parts.
 * 
 * @author Mark Matthews
 */
public class EscapeTokenizer {
	private static final char CHR_ESCAPE = '\\';
	private static final char CHR_SGL_QUOTE = '\'';
	private static final char CHR_DBL_QUOTE = '"';
	private static final char CHR_LF = '\n';
	private static final char CHR_CR = '\r';
	private static final char CHR_COMMENT = '-';
	private static final char CHR_BEGIN_TOKEN = '{';
	private static final char CHR_END_TOKEN = '}';
	private static final char CHR_VARIABLE = '@';

	private String source = null;
	private int sourceLength = 0;
	private int pos = 0;

	private boolean emittingEscapeCode = false;
	private boolean sawVariableUse = false;
	private int bracesLevel = 0;
	private boolean inQuotes = false;
	private char quoteChar = 0;

	/**
	 * Creates a new EscapeTokenizer object.
	 * 
	 * @param source
	 *            the string to tokenize
	 */
	public EscapeTokenizer(String source) {
		this.source = source;
		this.sourceLength = source.length();
		this.pos = 0;
	}

	/**
	 * Does this tokenizer have more tokens available?
	 * 
	 * @return if this tokenizer has more tokens available
	 */
	public synchronized boolean hasMoreTokens() {
		return (this.pos < this.sourceLength);
	}

	/**
	 * Returns the next token
	 * 
	 * @return the next token.
	 */
	public synchronized String nextToken() {
		StringBuffer tokenBuf = new StringBuffer();
		boolean backslashEscape = false;

		if (this.emittingEscapeCode) {
			// Previous token ended at the beginning of an escape code, so this token must start with '{'
			tokenBuf.append("{");
			this.emittingEscapeCode = false;
		}

		for (; this.pos < this.sourceLength; this.pos++) {
			char c = this.source.charAt(this.pos);

			// process escape char: (\)
			if (c == CHR_ESCAPE) {
				tokenBuf.append(c);
				backslashEscape = !backslashEscape;
				continue;
			}

			// process quotes: ('|")
			if ((c == CHR_SGL_QUOTE || c == CHR_DBL_QUOTE) && !backslashEscape) {
				tokenBuf.append(c);
				if (this.inQuotes) {
					if (c == this.quoteChar) {
						// look ahead for doubled quote
						if ((this.pos + 1 < this.sourceLength) && (this.source.charAt(this.pos + 1) == this.quoteChar)) {
							tokenBuf.append(c);
							this.pos++; // consume following char '\'' or '"'
						} else {
							this.inQuotes = false;
						}
					}
				} else {
					this.inQuotes = true;
					this.quoteChar = c;
				}
				continue;
			}

			// process new line: (\n|\r)
			if ((c == CHR_LF) || (c == CHR_CR)) {
				tokenBuf.append(c);
				backslashEscape = false;
				continue;
			}

			if (!this.inQuotes && !backslashEscape) {
				// process comments: (--)
				if (c == CHR_COMMENT) {
					tokenBuf.append(c);
					// look ahead for double hyphen
					if ((this.pos + 1 < this.sourceLength) && (this.source.charAt(this.pos + 1) == CHR_COMMENT)) {
						// consume following chars until new line or end of string
						while (++this.pos < this.sourceLength && c != CHR_LF && c != CHR_CR) {
							c = this.source.charAt(this.pos);
							tokenBuf.append(c);
						}
						this.pos--;
					}
					continue;
				}

				// process begin token: ({)
				if (c == CHR_BEGIN_TOKEN) {
					this.bracesLevel++;
					if (this.bracesLevel == 1) {
						this.emittingEscapeCode = true;
						this.pos++; // consume char '{' before returning
						return tokenBuf.toString();
					}
					tokenBuf.append(c);
					continue;
				}

				// process end token: (})
				if (c == CHR_END_TOKEN) {
					tokenBuf.append(c);
					this.bracesLevel--;
					if (this.bracesLevel == 0) {
						this.pos++; // consume char '}' before returning
						return tokenBuf.toString();
					}
					continue;
				}

				// detect variable usage: (@)
				if (c == CHR_VARIABLE) {
					this.sawVariableUse = true;
				}
			}

			tokenBuf.append(c);
			backslashEscape = false;
		}

		return tokenBuf.toString();
	}

	/**
	 * Returns true if a variable reference was found. Note that this information isn't accurate until finishing to
	 * process all tokens from source String. It also can't be used as per token basis.
	 * 
	 * @return true if a variable reference was found.
	 */
	boolean sawVariableUse() {
		return this.sawVariableUse;
	}
}