Update of /cvsroot/htmlparser/htmlparser/src/org/htmlparser/lexer
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv32030
Modified Files:
Page.java
Log Message:
Incorporate feature request submitted by Bradford A. Folkens
#943197 Accept gzip / deflate content encodings
by setting request property "Accept-Encoding" to "gzip, deflate" in Page.setConnection(),
if possible, and handling those encodings.
No test case added because it needs a specially configured HTTP server.
Index: Page.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/lexer/Page.java,v
retrieving revision 1.35
retrieving revision 1.36
diff -C2 -d -r1.35 -r1.36
*** Page.java 7 May 2004 23:30:37 -0000 1.35
--- Page.java 23 May 2004 19:42:14 -0000 1.36
***************
*** 40,43 ****
--- 40,45 ----
import java.net.URLConnection;
import java.net.UnknownHostException;
+ import java.util.zip.GZIPInputStream;
+ import java.util.zip.InflaterInputStream;
import org.htmlparser.util.EncodingChangeException;
***************
*** 319,326 ****
--- 321,337 ----
String type;
String charset;
+ String contentEncoding;
mConnection = connection;
try
{
+ try
+ {
+ getConnection ().setRequestProperty ("Accept-Encoding", "gzip, deflate");
+ }
+ catch (IllegalStateException ise) // already connected
+ {
+ // assume all request properties have already been set
+ }
getConnection ().connect ();
}
***************
*** 343,347 ****
try
{
! stream = new Stream (getConnection ().getInputStream ());
try
{
--- 354,370 ----
try
{
! contentEncoding = connection.getContentEncoding();
! if ((null != contentEncoding) && (-1 != contentEncoding.indexOf ("gzip")))
! {
! stream = new Stream (new GZIPInputStream (getConnection ().getInputStream ()));
! }
! else if ((null != contentEncoding) && (-1 != contentEncoding.indexOf ("deflate")))
! {
! stream = new Stream (new InflaterInputStream (getConnection ().getInputStream ()));
! }
! else
! {
! stream = new Stream (getConnection ().getInputStream ());
! }
try
{
|
