Open GoogleCodeExporter opened 9 years ago
Please ignore above fix, it causes problems with larger attachments.
I ended up changing it to read all data at once.
...
$len = $end-$start;
$part = fread($this->stream, $len);
fwrite($temp_fp, $this->decode($part, $encoding));
} else if ($this->data) {
...
Original comment by jma...@gmail.com
on 19 Jan 2010 at 8:31
I think in Method getAttachmentStream from MimeMailParser.class.php it should
be:
line 399:
old:
if (($written+$write < $len )) {
new:
if (($written+$write > $len )) {
Original comment by palvoel...@gmail.com
on 27 Aug 2010 at 3:40
I believe palvoelgyi is right.
Jmared, could you test that with your attachments and see if it works for you
please?
I will commit the change if it is verified.
http://code.google.com/p/php-mime-mail-parser/source/browse/trunk/MimeMailParser
.class.php?r=16#399
PS: We need to read the data in chunks as some attachments may be larger then
PHP memory limit.
Original comment by buca...@gmail.com
on 27 Aug 2010 at 4:45
Just noticed this was from Jan, will commit once I test this.
Original comment by buca...@gmail.com
on 27 Aug 2010 at 4:46
Isn't there a double bug here? The modification above is in theory correct, it
does what code structure intended to do.
But the problem is that you cannot call the decode() function for the chunks of
file like this! I guess it almost always works if the encoding is
quoted-printable. Could break if there happens to be a character encoding
across the chunks. But for base64 encoding this fails probably in majority of
the cases.
The old version fails in a nice way, causing long (>2k) attachments to be
handled in one chunk. Thus working properly with the decode function. The old
version failed on attachments smaller than 2kB.
The following is not a nice way of coding this, but this fix works for me. For
both small and big attachments and for both quoted-printable and base64 encoded
attachments. Changed the same part of code to read as:
while($written < $len) {
$write = $len;
$part = fread($this->stream, $write);
fwrite($temp_fp, $this->decode($part, $encoding));
$written += $write;
}
Pertti
Original comment by info.fam...@gmail.com
on 5 Nov 2010 at 4:27
I believe the stream filters should be used here when possible:
http://us3.php.net/manual/en/function.stream-get-filters.php
That way we can keep from saving the whole encoded data into a variable for
decoding, which would not be efficient for large files.
Original comment by buca...@gmail.com
on 6 Nov 2010 at 5:16
I just spent a few hours figuring out why some attachments weren't being
decoded correctly only to find out it had already been discussed here.
Maybe this should be changed so it reads the whole encoded data for now? It's
exactly what it's doing already anyway, except small attachments aren't decoded
properly. The entire while loop could just be replaced by something like:
$part = fread($this->stream, $len);
fwrite($temp_fp, $this->decode($part, $encoding));
It's a temporary fix, but it could save others from wasting time debugging the
exact same problem.
-Adrian
Original comment by agpc...@gmail.com
on 23 Feb 2012 at 7:54
Thank you Adrian.
Original comment by m.michal...@gmail.com
on 25 Mar 2013 at 12:41
The problem is easily fixed by replacing line 400
$write = 2028;
with
$write = ($len<2028)?$len:2028;
Original comment by eckhard....@arcor.de
on 14 May 2013 at 6:13
[deleted comment]
This issue is reproduced, corrected and tested
(https://github.com/eXorus/php-mime-mail-parser/issues/2)
You could use my fork if you want :
https://github.com/eXorus/php-mime-mail-parser
Original comment by eXorus.m...@gmail.com
on 28 Jul 2013 at 4:22
Original issue reported on code.google.com by
jma...@gmail.com
on 19 Jan 2010 at 7:56