1 /* 2 * Copyright (c) 1998, 2015, Oracle and/or its affiliates. All rights reserved. 3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. 4 * 5 * This code is free software; you can redistribute it and/or modify it 6 * under the terms of the GNU General Public License version 2 only, as 7 * published by the Free Software Foundation. Oracle designates this 8 * particular file as subject to the "Classpath" exception as provided 9 * by Oracle in the LICENSE file that accompanied this code. 10 * 11 * This code is distributed in the hope that it will be useful, but WITHOUT 12 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or 13 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License 14 * version 2 for more details (a copy is included in the LICENSE file that 15 * accompanied this code). 16 * 17 * You should have received a copy of the GNU General Public License version 18 * 2 along with this work; if not, write to the Free Software Foundation, 19 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. 20 * 21 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA 22 * or visit www.oracle.com if you need additional information or have any 23 * questions. 24 */ 25 26 package com.sun.xml.internal.dtdparser; 27 28 import org.xml.sax.EntityResolver; 29 import org.xml.sax.InputSource; 30 31 import java.io.File; 32 import java.io.FileInputStream; 33 import java.io.IOException; 34 import java.io.InputStream; 35 import java.net.URL; 36 import java.net.URLConnection; 37 import java.util.Hashtable; 38 import java.util.Locale; 39 40 /** 41 * This entity resolver class provides a number of utilities which can help 42 * managment of external parsed entities in XML. These are commonly used 43 * to hold markup declarations that are to be used as part of a Document 44 * Type Declaration (DTD), or to hold text marked up with XML. 45 * <p> 46 * <P> Features include: <UL> 47 * 48 * <LI> Static factory methods are provided for constructing SAX InputSource 49 * objects from Files, URLs, or MIME objects. This eliminates a class of 50 * error-prone coding in applications.</LI> 51 * 52 * <LI> Character encodings for XML documents are correctly supported:<UL> 53 * 54 * <LI> The encodings defined in the RFCs for MIME content types 55 * (2046 for general MIME, and 2376 for XML in particular), are 56 * supported, handling <em>charset=...</em> attributes and accepting 57 * content types which are known to be safe for use with XML;</LI> 58 * 59 * <LI> The character encoding autodetection algorithm identified 60 * in the XML specification is used, and leverages all of 61 * the JDK 1.1 (and later) character encoding support.</LI> 62 * 63 * <LI> The use of MIME typing may optionally be disabled, forcing the 64 * use of autodetection, to support web servers which don't correctly 65 * report MIME types for XML. For example, they may report text that 66 * is encoded in EUC-JP as being US-ASCII text, leading to fatal 67 * errors during parsing.</LI> 68 * 69 * <LI> The InputSource objects returned by this class always 70 * have a <code>java.io.Reader</code> available as the "character 71 * stream" property.</LI> 72 * 73 * </UL></LI> 74 * 75 * <LI> Catalog entries can map public identifiers to Java resources or 76 * to local URLs. These are used to reduce network dependencies and loads, 77 * and will often be used for external DTD components. For example, packages 78 * shipping DTD files as resources in JAR files can eliminate network traffic 79 * when accessing them, and sites may provide local caches of common DTDs. 80 * Note that no particular catalog syntax is supported by this class, only 81 * the notion of a set of entries.</LI> 82 * 83 * </UL> 84 * <p> 85 * <P> Subclasses can perform tasks such as supporting new URI schemes for 86 * URIs which are not URLs, such as URNs (see RFC 2396) or for accessing 87 * MIME entities which are part of a <em>multipart/related</em> group 88 * (see RFC 2387). They may also be used to support particular catalog 89 * syntaxes, such as the <a href="http://www.oasis-open.org/html/a401.htm"> 90 * SGML/Open Catalog (SOCAT)</a> which supports the SGML notion of "Formal 91 * Public Identifiers (FPIs). 92 * 93 * @author David Brownell 94 * @author Janet Koenig 95 * @version 1.3 00/02/24 96 */ 97 public class Resolver implements EntityResolver { 98 private boolean ignoringMIME; 99 100 // table mapping public IDs to (local) URIs 101 private Hashtable id2uri; 102 103 // tables mapping public IDs to resources and classloaders 104 private Hashtable id2resource; 105 private Hashtable id2loader; 106 107 // 108 // table of MIME content types (less attributes!) known 109 // to be mostly "OK" to use with XML MIME entities. the 110 // idea is to rule out obvious braindamage ("image/jpg") 111 // not the subtle stuff ("text/html") that might actually 112 // be (or become) safe. 113 // 114 private static final String types [] = { 115 "application/xml", 116 "text/xml", 117 "text/plain", 118 "text/html", // commonly mis-inferred 119 "application/x-netcdf", // this is often illegal XML 120 "content/unknown" 121 }; 122 123 /** 124 * Constructs a resolver. 125 */ 126 public Resolver() { 127 } 128 129 /** 130 * <p>Returns an input source, using the MIME type information and URL 131 * scheme to statically determine the correct character encoding if 132 * possible and otherwise autodetecting it. MIME carefully specifies 133 * the character encoding defaults, and how attributes of the content 134 * type can change it. XML further specifies two mandatory encodings 135 * (UTF-8 and UTF-16), and includes an XML declaration which can be 136 * used to internally label most documents encoded using US-ASCII 137 * supersets (such as Shift_JIS, EUC-JP, ISO-2022-*, ISO-8859-*, and 138 * more).</p> 139 * 140 * <p> This method can be used to access XML documents which do not 141 * have URIs (such as servlet input streams, or most JavaMail message 142 * entities) and to support access methods such as HTTP POST or PUT. 143 * (URLs normally return content using the GET method.)</p> 144 * 145 * <p> <em> The caller should set the system ID in order for relative URIs 146 * found in this document to be interpreted correctly.</em> In some cases, 147 * a custom resolver will need to be used; for example, documents 148 * may be grouped in a single MIME "multipart/related" bundle, and 149 * relative URLs would refer to other documents in that bundle.</p> 150 * 151 * @param contentType The MIME content type for the source for which 152 * an InputSource is desired, such as <em>text/xml;charset=utf-8</em>. 153 * @param stream The input byte stream for the input source. 154 * @param checkType If true, this verifies that the content type is known 155 * to support XML documents, such as <em>application/xml</em>. 156 * @param scheme Unless this is "file", unspecified MIME types 157 * default to US-ASCII. Files are always autodetected since most 158 * file systems discard character encoding information. 159 */ 160 public static InputSource createInputSource(String contentType, 161 InputStream stream, 162 boolean checkType, 163 String scheme) throws IOException { 164 InputSource retval; 165 String charset = null; 166 167 if (contentType != null) { 168 int index; 169 170 contentType = contentType.toLowerCase(Locale.ENGLISH); 171 index = contentType.indexOf(';'); 172 if (index != -1) { 173 String attributes; 174 175 attributes = contentType.substring(index + 1); 176 contentType = contentType.substring(0, index); 177 178 // use "charset=..." if it's available 179 index = attributes.indexOf("charset"); 180 if (index != -1) { 181 attributes = attributes.substring(index + 7); 182 // strip out subsequent attributes 183 if ((index = attributes.indexOf(';')) != -1) 184 attributes = attributes.substring(0, index); 185 // find start of value 186 if ((index = attributes.indexOf('=')) != -1) { 187 attributes = attributes.substring(index + 1); 188 // strip out rfc822 comments 189 if ((index = attributes.indexOf('(')) != -1) 190 attributes = attributes.substring(0, index); 191 // double quotes are optional 192 if ((index = attributes.indexOf('"')) != -1) { 193 attributes = attributes.substring(index + 1); 194 attributes = attributes.substring(0, 195 attributes.indexOf('"')); 196 } 197 charset = attributes.trim(); 198 // XXX "\;", "\)" etc were mishandled above 199 } 200 } 201 } 202 203 // 204 // Check MIME type. 205 // 206 if (checkType) { 207 boolean isOK = false; 208 for (int i = 0; i < types.length; i++) 209 if (types[i].equals(contentType)) { 210 isOK = true; 211 break; 212 } 213 if (!isOK) 214 throw new IOException("Not XML: " + contentType); 215 } 216 217 // 218 // "text/*" MIME types have hard-wired character set 219 // defaults, as specified in the RFCs. For XML, we 220 // ignore the system "file.encoding" property since 221 // autodetection is more correct. 222 // 223 if (charset == null) { 224 contentType = contentType.trim(); 225 if (contentType.startsWith("text/")) { 226 if (!"file".equalsIgnoreCase(scheme)) 227 charset = "US-ASCII"; 228 } 229 // "application/*" has no default 230 } 231 } 232 233 retval = new InputSource(XmlReader.createReader(stream, charset)); 234 retval.setByteStream(stream); 235 retval.setEncoding(charset); 236 return retval; 237 } 238 239 240 /** 241 * Creates an input source from a given URI. 242 * 243 * @param uri the URI (system ID) for the entity 244 * @param checkType if true, the MIME content type for the entity 245 * is checked for document type and character set encoding. 246 */ 247 static public InputSource createInputSource(URL uri, boolean checkType) 248 throws IOException { 249 250 URLConnection conn = uri.openConnection(); 251 InputSource retval; 252 253 if (checkType) { 254 String contentType = conn.getContentType(); 255 retval = createInputSource(contentType, conn.getInputStream(), 256 false, uri.getProtocol()); 257 } else { 258 retval = new InputSource(XmlReader.createReader(conn.getInputStream())); 259 } 260 retval.setSystemId(conn.getURL().toString()); 261 return retval; 262 } 263 264 265 /** 266 * Creates an input source from a given file, autodetecting 267 * the character encoding. 268 */ 269 static public InputSource createInputSource(File file) 270 throws IOException { 271 InputSource retval; 272 String path; 273 274 retval = new InputSource(XmlReader.createReader(new FileInputStream(file))); 275 276 // On JDK 1.2 and later, simplify this: 277 // "path = file.toURL ().toString ()". 278 path = file.getAbsolutePath(); 279 if (File.separatorChar != '/') 280 path = path.replace(File.separatorChar, '/'); 281 if (!path.startsWith("/")) 282 path = "/" + path; 283 if (!path.endsWith("/") && file.isDirectory()) 284 path = path + "/"; 285 286 retval.setSystemId("file:" + path); 287 return retval; 288 } 289 290 291 /** 292 * <b>SAX:</b> 293 * Resolve the given entity into an input source. If the name can't 294 * be mapped to a preferred form of the entity, the URI is used. To 295 * resolve the entity, first a local catalog mapping names to URIs is 296 * consulted. If no mapping is found there, a catalog mapping names 297 * to java resources is consulted. Finally, if neither mapping found 298 * a copy of the entity, the specified URI is used. 299 * <p> 300 * <P> When a URI is used, <a href="#createInputSource"> 301 * createInputSource</a> is used to correctly deduce the character 302 * encoding used by this entity. No MIME type checking is done. 303 * 304 * @param name Used to find alternate copies of the entity, when 305 * this value is non-null; this is the XML "public ID". 306 * @param uri Used when no alternate copy of the entity is found; 307 * this is the XML "system ID", normally a URI. 308 */ 309 @Override 310 public InputSource resolveEntity(String name, String uri) 311 throws IOException { 312 InputSource retval; 313 String mappedURI = name2uri(name); 314 InputStream stream; 315 316 // prefer explicit URI mappings, then bundled resources... 317 if (mappedURI == null && (stream = mapResource(name)) != null && id2resource != null) { 318 uri = "java:resource:" + (String) id2resource.get(name); 319 retval = new InputSource(XmlReader.createReader(stream)); 320 321 // ...and treat all URIs the same (as URLs for now). 322 } else { 323 URL url; 324 URLConnection conn; 325 326 if (mappedURI != null) 327 uri = mappedURI; 328 else if (uri == null) 329 return null; 330 331 url = new URL(uri); 332 conn = url.openConnection(); 333 uri = conn.getURL().toString(); 334 // System.out.println ("++ URI: " + url); 335 if (ignoringMIME) 336 retval = new InputSource(XmlReader.createReader(conn.getInputStream())); 337 else { 338 String contentType = conn.getContentType(); 339 retval = createInputSource(contentType, 340 conn.getInputStream(), 341 false, url.getProtocol()); 342 } 343 } 344 retval.setSystemId(uri); 345 retval.setPublicId(name); 346 return retval; 347 } 348 349 350 /** 351 * Returns true if this resolver is ignoring MIME types in the documents 352 * it returns, to work around bugs in how servers have reported the 353 * documents' MIME types. 354 */ 355 public boolean isIgnoringMIME() { 356 return ignoringMIME; 357 } 358 359 /** 360 * Tells the resolver whether to ignore MIME types in the documents it 361 * retrieves. Many web servers incorrectly assign text documents a 362 * default character encoding, even when that is incorrect. For example, 363 * all HTTP text documents default to use ISO-8859-1 (used for Western 364 * European languages), and other MIME sources default text documents 365 * to use US-ASCII (a seven bit encoding). For XML documents which 366 * include text encoding declarations (as most should do), these server 367 * bugs can be worked around by ignoring the MIME type entirely. 368 */ 369 public void setIgnoringMIME(boolean value) { 370 ignoringMIME = value; 371 } 372 373 374 // maps the public ID to an alternate URI, if one is registered 375 private String name2uri(String publicId) { 376 if (publicId == null || id2uri == null) 377 return null; 378 return (String) id2uri.get(publicId); 379 } 380 381 382 /** 383 * Registers the given public ID as corresponding to a particular 384 * URI, typically a local copy. This URI will be used in preference 385 * to ones provided as system IDs in XML entity declarations. This 386 * mechanism would most typically be used for Document Type Definitions 387 * (DTDs), where the public IDs are formally managed and versioned. 388 * 389 * @param publicId The managed public ID being mapped 390 * @param uri The URI of the preferred copy of that entity 391 */ 392 public void registerCatalogEntry(String publicId, 393 String uri) { 394 if (id2uri == null) 395 id2uri = new Hashtable(17); 396 id2uri.put(publicId, uri); 397 } 398 399 400 // return the resource as a stream 401 private InputStream mapResource(String publicId) { 402 // System.out.println ("++ PUBLIC: " + publicId); 403 if (publicId == null || id2resource == null) 404 return null; 405 406 String resourceName = (String) id2resource.get(publicId); 407 ClassLoader loader = null; 408 409 if (resourceName == null) 410 return null; 411 // System.out.println ("++ Resource: " + resourceName); 412 413 if (id2loader != null) 414 loader = (ClassLoader) id2loader.get(publicId); 415 // System.out.println ("++ Loader: " + loader); 416 if (loader == null) 417 return ClassLoader.getSystemResourceAsStream(resourceName); 418 return loader.getResourceAsStream(resourceName); 419 } 420 421 /** 422 * Registers a given public ID as corresponding to a particular Java 423 * resource in a given class loader, typically distributed with a 424 * software package. This resource will be preferred over system IDs 425 * included in XML documents. This mechanism should most typically be 426 * used for Document Type Definitions (DTDs), where the public IDs are 427 * formally managed and versioned. 428 * <p> 429 * <P> If a mapping to a URI has been provided, that mapping takes 430 * precedence over this one. 431 * 432 * @param publicId The managed public ID being mapped 433 * @param resourceName The name of the Java resource 434 * @param loader The class loader holding the resource, or null if 435 * it is a system resource. 436 */ 437 public void registerCatalogEntry(String publicId, 438 String resourceName, 439 ClassLoader loader) { 440 if (id2resource == null) 441 id2resource = new Hashtable(17); 442 id2resource.put(publicId, resourceName); 443 444 if (loader != null) { 445 if (id2loader == null) 446 id2loader = new Hashtable(17); 447 id2loader.put(publicId, loader); 448 } 449 } 450 }