Revision: 7354
http://mahogany.svn.sourceforge.net/mahogany/?rev=7354&view=rev
Author: vadz
Date: 2007-08-31 14:58:51 -0700 (Fri, 31 Aug 2007)
Log Message:
-----------
fix encoding of consecutive encoded words in QP: we must encode the spaces as
they're ignored between encode words otherwise
Modified Paths:
--------------
trunk/M/src/mail/MimeDecode.cpp
trunk/M/tests/mime/decode.cpp
Modified: trunk/M/src/mail/MimeDecode.cpp
===================================================================
--- trunk/M/src/mail/MimeDecode.cpp 2007-08-30 20:24:02 UTC (rev 7353)
+++ trunk/M/src/mail/MimeDecode.cpp 2007-08-31 21:58:51 UTC (rev 7354)
@@ -422,16 +422,15 @@
return true;
}
+// encode the given text unconditionally, i.e. without checking if it must be
+// encoded (this is supposed to be done in the caller) and using the specified
+// encodings andcharset (which are supposed to be detected by the caller too)
static String
-EncodeWord(const String& in,
+EncodeText(const String& in,
wxFontEncoding enc,
MIME::Encoding enc2047,
const String& csName)
{
- if ( !NeedsEncoding(in) )
- return in;
-
-
// encode the word splitting it in the chunks such that they will be no
// longer than 75 characters each
wxCharBuffer buf(in.mb_str(wxCSConv(enc)));
@@ -608,24 +607,65 @@
String headerEnc;
headerEnc.reserve(2*in.length());
- // for QP we encode each header word separately so that the header remains
- // readable, but for Base64 it's useless to do this as it's unreadable
- // anyhow so we just encode everything at once
+ // for QP we encode each header word separately as some might not need being
+ // encoded at all and the header remains more readable if we don't encode
+ // them unnecessarily, but for Base64 it's useless to do this as it's
+ // unreadable anyhow so we just encode everything at once
if ( enc2047 == MIME::Encoding_QuotedPrintable )
{
- // encode each word of the header
+ // encode each word of the header if necessary, taking into account one
+ // added complication: white space between 2 consecutive encoded words is
+ // ignored during decoding, so we must encode 2 consecutive words both of
+ // which need encoding as one single encoded word or the space between
+ // them would be lost
+ bool lastWordEncoded = false;
const wxArrayString words(wxStringTokenize(in));
const size_t count = words.size();
for ( size_t n = 0; n < count; ++n )
{
- headerEnc += EncodeWord(words[n], enc, enc2047, csName);
- if ( n + 1 < count )
- headerEnc += ' ';
+ const wxString& word = words[n];
+ if ( NeedsEncoding(word) )
+ {
+ const String wordEnc = EncodeText(word, enc, enc2047, csName);
+
+ if ( lastWordEncoded )
+ {
+ // we need to merge the 2 consecutive encoded words together: we
+ // do it by removing "?=" suffix from the previous word, adding
+ // a space and remove the "=?charset?Q?" prefix from this word
+ ASSERT_MSG( headerEnc.length() > 7, "bad QP-encoded last word"
);
+ headerEnc.RemoveLast(2); // "?="
+
+ headerEnc += '_'; // space can be represented like this in QP
+
+ const size_t posText = wordEnc.find("?Q?");
+ ASSERT_MSG( posText != String::npos, "bad QP-encoded word" );
+ headerEnc += wordEnc.substr(posText + 3);
+ }
+ else // last word not encoded, just append this one
+ {
+ if ( !headerEnc.empty() )
+ headerEnc += ' ';
+
+ headerEnc += wordEnc;
+ }
+
+ lastWordEncoded = true;
+ }
+ else // this word doesn't need to be encoded, simply append it
+ {
+ if ( !headerEnc.empty() )
+ headerEnc += ' ';
+
+ headerEnc += word;
+
+ lastWordEncoded = false;
+ }
}
}
else // MIME::Encoding_Base64
{
- headerEnc = EncodeWord(in, enc, enc2047, csName);
+ headerEnc = EncodeText(in, enc, enc2047, csName);
}
return headerEnc.ToAscii();
Modified: trunk/M/tests/mime/decode.cpp
===================================================================
--- trunk/M/tests/mime/decode.cpp 2007-08-30 20:24:02 UTC (rev 7353)
+++ trunk/M/tests/mime/decode.cpp 2007-08-31 21:58:51 UTC (rev 7354)
@@ -363,6 +363,26 @@
"Ludovic P\303\251net <[EMAIL PROTECTED]>",
wxFONTENCODING_UTF8
},
+
+ {
+ "=?UTF-8?Q?=D0=92=D0=B0=D0=B4=D0=B8=D0=BC_"
+ "=D0=A6=D0=B5=D0=B9=D1=82=D0=BB=D0=B8=D0=BD?=",
+ "\xD0\x92\xD0\xB0\xD0\xB4\xD0\xB8\xD0\xBC "
+ "\xD0\xA6\xD0\xB5\xD0\xB9\xD1\x82\xD0\xBB\xD0\xB8\xD0\xBD",
+ wxFONTENCODING_UTF8
+ },
+
+ {
+ "2006 =?UTF-8?Q?=D0=92_=D0=A6?= 2007",
+ "2006 \xD0\x92 \xD0\xA6 2007",
+ wxFONTENCODING_UTF8
+ },
+
+ {
+ "=?UTF-8?Q?=D0=92_=D0=A6?= 2007",
+ "\xD0\x92 \xD0\xA6 2007",
+ wxFONTENCODING_UTF8
+ }
};
int rc = EXIT_SUCCESS;
This was sent by the SourceForge.net collaborative development platform, the
world's largest Open Source development site.
-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Mahogany-cvsupdates mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mahogany-cvsupdates
