Source for gnu.java.net.protocol.http.ChunkedInputStream

   1: /* ChunkedInputStream.java --
   2:    Copyright (C) 2004, 2006 Free Software Foundation, Inc.
   3: 
   4: This file is part of GNU Classpath.
   5: 
   6: GNU Classpath is free software; you can redistribute it and/or modify
   7: it under the terms of the GNU General Public License as published by
   8: the Free Software Foundation; either version 2, or (at your option)
   9: any later version.
  10: 
  11: GNU Classpath is distributed in the hope that it will be useful, but
  12: WITHOUT ANY WARRANTY; without even the implied warranty of
  13: MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14: General Public License for more details.
  15: 
  16: You should have received a copy of the GNU General Public License
  17: along with GNU Classpath; see the file COPYING.  If not, write to the
  18: Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
  19: 02110-1301 USA.
  20: 
  21: Linking this library statically or dynamically with other modules is
  22: making a combined work based on this library.  Thus, the terms and
  23: conditions of the GNU General Public License cover the whole
  24: combination.
  25: 
  26: As a special exception, the copyright holders of this library give you
  27: permission to link this library with independent modules to produce an
  28: executable, regardless of the license terms of these independent
  29: modules, and to copy and distribute the resulting executable under
  30: terms of your choice, provided that you also meet, for each linked
  31: independent module, the terms and conditions of the license of that
  32: module.  An independent module is a module which is not derived from
  33: or based on this library.  If you modify this library, you may extend
  34: this exception to your version of the library, but you are not
  35: obligated to do so.  If you do not wish to do so, delete this
  36: exception statement from your version. */
  37: 
  38: 
  39: package gnu.java.net.protocol.http;
  40: 
  41: import gnu.java.lang.CPStringBuilder;
  42: 
  43: import java.io.IOException;
  44: import java.io.InputStream;
  45: import java.net.ProtocolException;
  46: 
  47: 
  48: //
  49: // Note that we rely on the implemtation of skip() in the super class
  50: // (InputStream) calling our read methods to account for chunk headers
  51: // while skipping.
  52: //
  53: 
  54: 
  55: /**
  56:  * Input stream wrapper for the "chunked" transfer-coding.
  57:  *
  58:  * @author Chris Burdess (dog@gnu.org)
  59:  */
  60: public class ChunkedInputStream
  61:   extends InputStream
  62: {
  63:   Headers headers;
  64: 
  65:   /** The underlying stream. */
  66:   private InputStream in;
  67: 
  68:   /** Size of the chunk we're reading.  */
  69:   int size;
  70:   /** Number of bytes we've read in this chunk.  */
  71:   int count;
  72:   /**
  73:    * True when we should read meta-information, false when we should
  74:    * read data.
  75:    */
  76:   boolean meta;
  77:   /** True when we've hit EOF.  */
  78:   boolean eof;
  79: 
  80:   /**
  81:    * Constructor.
  82:    * @param in the response socket input stream
  83:    * @param headers the headers to receive additional header lines
  84:    */
  85:   public ChunkedInputStream(InputStream in, Headers headers)
  86:   {
  87:     this.in = in;
  88:     this.headers = headers;
  89:     size = -1;
  90:     count = 0;
  91:     meta = true;
  92:   }
  93: 
  94:   public int read()
  95:     throws IOException
  96:   {
  97:     byte[] buf = new byte[1];
  98:     int len = read(buf, 0, 1);
  99:     if (len == -1)
 100:       {
 101:         return -1;
 102:       }
 103:     return 0xff & buf[0];
 104:   }
 105: 
 106:   public synchronized int read(byte[] buffer, int offset, int length)
 107:     throws IOException
 108:   {
 109:     if (eof)
 110:       {
 111:         return -1;
 112:       }
 113:     if (meta)
 114:       {
 115:         // Read chunk header
 116:         int c, last = 0;
 117:         boolean seenSemi = false;
 118:         CPStringBuilder buf = new CPStringBuilder();
 119:         do
 120:           {
 121:             c = in.read();
 122:             if (c == 0x3b) // ;
 123:               {
 124:                 seenSemi = true;
 125:               }
 126:             else if (c == 0x0a && last == 0x0d) // CRLF
 127:               {
 128:                 try
 129:                   {
 130:                     size = Integer.parseInt(buf.toString(), 16);
 131:                   }
 132:                 catch (NumberFormatException nfe)
 133:                   {
 134:                     IOException ioe = new IOException("Bad chunk header");
 135:                     ioe.initCause(nfe);
 136:                     // Unrecoverable.  Don't try to read more.
 137:                     in.close();
 138:                     throw ioe;
 139:                   }
 140:                 break;
 141:               }
 142:             else if (!seenSemi && c >= 0x30)
 143:               {
 144:                 buf.append ((char) c);
 145:               }
 146:             last = c;
 147:           }
 148:         while(c != -1);
 149:         count = 0;
 150:         meta = false;
 151:       }
 152:     if (size == 0)
 153:       {
 154:         // Read trailer
 155:         headers.parse(in);
 156:         eof = true;
 157:         return -1;
 158:       }
 159:     else
 160:       {
 161:         int canRead = Math.min(size - count, length);
 162:         int len = in.read(buffer, offset, canRead);
 163:         if (len == -1)
 164:           {
 165:             // This is an error condition but it isn't clear what we
 166:             // should do with it.
 167:             eof = true;
 168:             return -1;
 169:           }
 170:         count += len;
 171:         if (count == size)
 172:           {
 173:             // Read CRLF
 174:             int c1 = in.read();
 175:             int c2 = in.read();
 176:             if (c1 == -1 || c2 == -1)
 177:               {
 178:                 // EOF before CRLF: bad, but ignore
 179:                 eof = true;
 180:                 return -1;
 181:               }
 182:             if (c1 != 0x0d || c2 != 0x0a)
 183:               {
 184:                 throw new ProtocolException("expecting CRLF: " + c1 + "," + c2);
 185:               }
 186:             meta = true;
 187:           }
 188:         return len;
 189:       }
 190:   }
 191: 
 192:   /**
 193:    * This method returns the number of bytes that can be read from
 194:    * this stream before a read might block.  Even if the underlying
 195:    * InputStream has data available past the end of the current chunk,
 196:    * we have no way of knowing how large the next chunk header will
 197:    * be. So we cannot report available data past the current chunk.
 198:    *
 199:    * @return The number of bytes that can be read before a read might
 200:    * block
 201:    *
 202:    * @exception IOException If an error occurs
 203:    */
 204:   public int available() throws IOException
 205:   {
 206:     if (meta)
 207:       return 0;
 208: 
 209:     return Math.min(in.available(), size - count);
 210:   }
 211: 
 212:   /**
 213:    * This method closes the ChunkedInputStream by closing the underlying
 214:    * InputStream.
 215:    *
 216:    * @exception IOException If an error occurs
 217:    */
 218:   public void close() throws IOException
 219:   {
 220:     in.close();
 221:   }
 222: 
 223: }