1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/src/share/jaxws_classes/com/sun/xml/internal/ws/encoding/HeaderTokenizer.java Wed Apr 27 01:27:09 2016 +0800
1.3 @@ -0,0 +1,377 @@
1.4 +/*
1.5 + * Copyright (c) 1997, 2012, Oracle and/or its affiliates. All rights reserved.
1.6 + * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
1.7 + *
1.8 + * This code is free software; you can redistribute it and/or modify it
1.9 + * under the terms of the GNU General Public License version 2 only, as
1.10 + * published by the Free Software Foundation. Oracle designates this
1.11 + * particular file as subject to the "Classpath" exception as provided
1.12 + * by Oracle in the LICENSE file that accompanied this code.
1.13 + *
1.14 + * This code is distributed in the hope that it will be useful, but WITHOUT
1.15 + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
1.16 + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
1.17 + * version 2 for more details (a copy is included in the LICENSE file that
1.18 + * accompanied this code).
1.19 + *
1.20 + * You should have received a copy of the GNU General Public License version
1.21 + * 2 along with this work; if not, write to the Free Software Foundation,
1.22 + * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
1.23 + *
1.24 + * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
1.25 + * or visit www.oracle.com if you need additional information or have any
1.26 + * questions.
1.27 + */
1.28 +
1.29 +package com.sun.xml.internal.ws.encoding;
1.30 +
1.31 +import javax.xml.ws.WebServiceException;
1.32 +
1.33 +/**
1.34 + * This class tokenizes RFC822 and MIME headers into the basic
1.35 + * symbols specified by RFC822 and MIME. <p>
1.36 + *
1.37 + * This class handles folded headers (ie headers with embedded
1.38 + * CRLF SPACE sequences). The folds are removed in the returned
1.39 + * tokens.
1.40 + *
1.41 + * @version 1.9, 02/03/27
1.42 + * @author John Mani
1.43 + */
1.44 +
1.45 +class HeaderTokenizer {
1.46 +
1.47 + /**
1.48 + * The Token class represents tokens returned by the
1.49 + * HeaderTokenizer.
1.50 + */
1.51 + static class Token {
1.52 +
1.53 + private int type;
1.54 + private String value;
1.55 +
1.56 + /**
1.57 + * Token type indicating an ATOM.
1.58 + */
1.59 + public static final int ATOM = -1;
1.60 +
1.61 + /**
1.62 + * Token type indicating a quoted string. The value
1.63 + * field contains the string without the quotes.
1.64 + */
1.65 + public static final int QUOTEDSTRING = -2;
1.66 +
1.67 + /**
1.68 + * Token type indicating a comment. The value field
1.69 + * contains the comment string without the comment
1.70 + * start and end symbols.
1.71 + */
1.72 + public static final int COMMENT = -3;
1.73 +
1.74 + /**
1.75 + * Token type indicating end of input.
1.76 + */
1.77 + public static final int EOF = -4;
1.78 +
1.79 + /**
1.80 + * Constructor.
1.81 + * @param type Token type
1.82 + * @param value Token value
1.83 + */
1.84 + public Token(int type, String value) {
1.85 + this.type = type;
1.86 + this.value = value;
1.87 + }
1.88 +
1.89 + /**
1.90 + * Return the type of the token. If the token represents a
1.91 + * delimiter or a control character, the type is that character
1.92 + * itself, converted to an integer. Otherwise, it's value is
1.93 + * one of the following:
1.94 + * <ul>
1.95 + * <li><code>ATOM</code> A sequence of ASCII characters
1.96 + * delimited by either SPACE, CTL, "(", <"> or the
1.97 + * specified SPECIALS
1.98 + * <li><code>QUOTEDSTRING</code> A sequence of ASCII characters
1.99 + * within quotes
1.100 + * <li><code>COMMENT</code> A sequence of ASCII characters
1.101 + * within "(" and ")".
1.102 + * <li><code>EOF</code> End of header
1.103 + * </ul>
1.104 + */
1.105 + public int getType() {
1.106 + return type;
1.107 + }
1.108 +
1.109 + /**
1.110 + * Returns the value of the token just read. When the current
1.111 + * token is a quoted string, this field contains the body of the
1.112 + * string, without the quotes. When the current token is a comment,
1.113 + * this field contains the body of the comment.
1.114 + *
1.115 + * @return token value
1.116 + */
1.117 + public String getValue() {
1.118 + return value;
1.119 + }
1.120 + }
1.121 +
1.122 + private String string; // the string to be tokenized
1.123 + private boolean skipComments; // should comments be skipped ?
1.124 + private String delimiters; // delimiter string
1.125 + private int currentPos; // current parse position
1.126 + private int maxPos; // string length
1.127 + private int nextPos; // track start of next Token for next()
1.128 + private int peekPos; // track start of next Token for peek()
1.129 +
1.130 + /**
1.131 + * RFC822 specials
1.132 + */
1.133 + private final static String RFC822 = "()<>@,;:\\\"\t .[]";
1.134 +
1.135 + /**
1.136 + * MIME specials
1.137 + */
1.138 + final static String MIME = "()<>@,;:\\\"\t []/?=";
1.139 +
1.140 + // The EOF Token
1.141 + private final static Token EOFToken = new Token(Token.EOF, null);
1.142 +
1.143 + /**
1.144 + * Constructor that takes a rfc822 style header.
1.145 + *
1.146 + * @param header The rfc822 header to be tokenized
1.147 + * @param delimiters Set of delimiter characters
1.148 + * to be used to delimit ATOMS. These
1.149 + * are usually <code>RFC822</code> or
1.150 + * <code>MIME</code>
1.151 + * @param skipComments If true, comments are skipped and
1.152 + * not returned as tokens
1.153 + */
1.154 + HeaderTokenizer(String header, String delimiters,
1.155 + boolean skipComments) {
1.156 + string = (header == null) ? "" : header; // paranoia ?!
1.157 + this.skipComments = skipComments;
1.158 + this.delimiters = delimiters;
1.159 + currentPos = nextPos = peekPos = 0;
1.160 + maxPos = string.length();
1.161 + }
1.162 +
1.163 + /**
1.164 + * Constructor. Comments are ignored and not returned as tokens
1.165 + *
1.166 + * @param header The header that is tokenized
1.167 + * @param delimiters The delimiters to be used
1.168 + */
1.169 + HeaderTokenizer(String header, String delimiters) {
1.170 + this(header, delimiters, true);
1.171 + }
1.172 +
1.173 + /**
1.174 + * Constructor. The RFC822 defined delimiters - RFC822 - are
1.175 + * used to delimit ATOMS. Also comments are skipped and not
1.176 + * returned as tokens
1.177 + */
1.178 + HeaderTokenizer(String header) {
1.179 + this(header, RFC822);
1.180 + }
1.181 +
1.182 + /**
1.183 + * Parses the next token from this String. <p>
1.184 + *
1.185 + * Clients sit in a loop calling next() to parse successive
1.186 + * tokens until an EOF Token is returned.
1.187 + *
1.188 + * @return the next Token
1.189 + * @exception WebServiceException if the parse fails
1.190 + */
1.191 + Token next() throws WebServiceException {
1.192 + Token tk;
1.193 +
1.194 + currentPos = nextPos; // setup currentPos
1.195 + tk = getNext();
1.196 + nextPos = peekPos = currentPos; // update currentPos and peekPos
1.197 + return tk;
1.198 + }
1.199 +
1.200 + /**
1.201 + * Peek at the next token, without actually removing the token
1.202 + * from the parse stream. Invoking this method multiple times
1.203 + * will return successive tokens, until <code>next()</code> is
1.204 + * called. <p>
1.205 + *
1.206 + * @return the next Token
1.207 + * @exception WebServiceException if the parse fails
1.208 + */
1.209 + Token peek() throws WebServiceException {
1.210 + Token tk;
1.211 +
1.212 + currentPos = peekPos; // setup currentPos
1.213 + tk = getNext();
1.214 + peekPos = currentPos; // update peekPos
1.215 + return tk;
1.216 + }
1.217 +
1.218 + /**
1.219 + * Return the rest of the Header.
1.220 + *
1.221 + * @return String rest of header. null is returned if we are
1.222 + * already at end of header
1.223 + */
1.224 + String getRemainder() {
1.225 + return string.substring(nextPos);
1.226 + }
1.227 +
1.228 + /*
1.229 + * Return the next token starting from 'currentPos'. After the
1.230 + * parse, 'currentPos' is updated to point to the start of the
1.231 + * next token.
1.232 + */
1.233 + private Token getNext() throws WebServiceException {
1.234 + // If we're already at end of string, return EOF
1.235 + if (currentPos >= maxPos)
1.236 + return EOFToken;
1.237 +
1.238 + // Skip white-space, position currentPos beyond the space
1.239 + if (skipWhiteSpace() == Token.EOF)
1.240 + return EOFToken;
1.241 +
1.242 + char c;
1.243 + int start;
1.244 + boolean filter = false;
1.245 +
1.246 + c = string.charAt(currentPos);
1.247 +
1.248 + // Check or Skip comments and position currentPos
1.249 + // beyond the comment
1.250 + while (c == '(') {
1.251 + // Parsing comment ..
1.252 + int nesting;
1.253 + for (start = ++currentPos, nesting = 1;
1.254 + nesting > 0 && currentPos < maxPos;
1.255 + currentPos++) {
1.256 + c = string.charAt(currentPos);
1.257 + if (c == '\\') { // Escape sequence
1.258 + currentPos++; // skip the escaped character
1.259 + filter = true;
1.260 + } else if (c == '\r')
1.261 + filter = true;
1.262 + else if (c == '(')
1.263 + nesting++;
1.264 + else if (c == ')')
1.265 + nesting--;
1.266 + }
1.267 + if (nesting != 0)
1.268 + throw new WebServiceException("Unbalanced comments");
1.269 +
1.270 + if (!skipComments) {
1.271 + // Return the comment, if we are asked to.
1.272 + // Note that the comment start & end markers are ignored.
1.273 + String s;
1.274 + if (filter) // need to go thru the token again.
1.275 + s = filterToken(string, start, currentPos-1);
1.276 + else
1.277 + s = string.substring(start,currentPos-1);
1.278 +
1.279 + return new Token(Token.COMMENT, s);
1.280 + }
1.281 +
1.282 + // Skip any whitespace after the comment.
1.283 + if (skipWhiteSpace() == Token.EOF)
1.284 + return EOFToken;
1.285 + c = string.charAt(currentPos);
1.286 + }
1.287 +
1.288 + // Check for quoted-string and position currentPos
1.289 + // beyond the terminating quote
1.290 + if (c == '"') {
1.291 + for (start = ++currentPos; currentPos < maxPos; currentPos++) {
1.292 + c = string.charAt(currentPos);
1.293 + if (c == '\\') { // Escape sequence
1.294 + currentPos++;
1.295 + filter = true;
1.296 + } else if (c == '\r')
1.297 + filter = true;
1.298 + else if (c == '"') {
1.299 + currentPos++;
1.300 + String s;
1.301 +
1.302 + if (filter)
1.303 + s = filterToken(string, start, currentPos-1);
1.304 + else
1.305 + s = string.substring(start,currentPos-1);
1.306 +
1.307 + return new Token(Token.QUOTEDSTRING, s);
1.308 + }
1.309 + }
1.310 + throw new WebServiceException("Unbalanced quoted string");
1.311 + }
1.312 +
1.313 + // Check for SPECIAL or CTL
1.314 + if (c < 040 || c >= 0177 || delimiters.indexOf(c) >= 0) {
1.315 + currentPos++; // re-position currentPos
1.316 + char ch[] = new char[1];
1.317 + ch[0] = c;
1.318 + return new Token((int)c, new String(ch));
1.319 + }
1.320 +
1.321 + // Check for ATOM
1.322 + for (start = currentPos; currentPos < maxPos; currentPos++) {
1.323 + c = string.charAt(currentPos);
1.324 + // ATOM is delimited by either SPACE, CTL, "(", <">
1.325 + // or the specified SPECIALS
1.326 + if (c < 040 || c >= 0177 || c == '(' || c == ' ' ||
1.327 + c == '"' || delimiters.indexOf(c) >= 0)
1.328 + break;
1.329 + }
1.330 + return new Token(Token.ATOM, string.substring(start, currentPos));
1.331 + }
1.332 +
1.333 + // Skip SPACE, HT, CR and NL
1.334 + private int skipWhiteSpace() {
1.335 + char c;
1.336 + for (; currentPos < maxPos; currentPos++)
1.337 + if (((c = string.charAt(currentPos)) != ' ') &&
1.338 + (c != '\t') && (c != '\r') && (c != '\n'))
1.339 + return currentPos;
1.340 + return Token.EOF;
1.341 + }
1.342 +
1.343 + /* Process escape sequences and embedded LWSPs from a comment or
1.344 + * quoted string.
1.345 + */
1.346 + private static String filterToken(String s, int start, int end) {
1.347 + StringBuffer sb = new StringBuffer();
1.348 + char c;
1.349 + boolean gotEscape = false;
1.350 + boolean gotCR = false;
1.351 +
1.352 + for (int i = start; i < end; i++) {
1.353 + c = s.charAt(i);
1.354 + if (c == '\n' && gotCR) {
1.355 + // This LF is part of an unescaped
1.356 + // CRLF sequence (i.e, LWSP). Skip it.
1.357 + gotCR = false;
1.358 + continue;
1.359 + }
1.360 +
1.361 + gotCR = false;
1.362 + if (!gotEscape) {
1.363 + // Previous character was NOT '\'
1.364 + if (c == '\\') // skip this character
1.365 + gotEscape = true;
1.366 + else if (c == '\r') // skip this character
1.367 + gotCR = true;
1.368 + else // append this character
1.369 + sb.append(c);
1.370 + } else {
1.371 + // Previous character was '\'. So no need to
1.372 + // bother with any special processing, just
1.373 + // append this character
1.374 + sb.append(c);
1.375 + gotEscape = false;
1.376 + }
1.377 + }
1.378 + return sb.toString();
1.379 + }
1.380 +}
