 |
BorlandTalk.com
Borland discussion newsgroups
|
| View previous topic :: View next topic |
| Author |
Message |
Bert van der Hoeven
Guest
|
 Posted: Mon Feb 06, 2006 10:02 pm Post subject: POP3 wrong decoding filename |
 |
|
This a part of a e-mail file genarated with lotus notes, whith indy10
snapshot 6 feb. 2006 the filename is not decoded.
My question is: is the mail wrong or is indy not correct translating the
name or have i decode the filename myself?
The attechment self is ok.
thanks
Bert.
--0__=4EBBFB95DFD462628f9e8a93df938690918c4EBBFB95DFD46262
Content-Type: application/msexcel;
name="=?UTF-8?B?cGxhbm5pbmcgZmFzZSAwIHZlcnNpZSAzYi54bHM=?="
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
filename="=?UTF-8?B?cGxhbm5pbmcgZmFzZSAwIHZlcnNpZSAzYi54bHM=?=" |
|
| Back to top |
|
 |
Remy Lebeau (TeamB)
Guest
|
| Posted: Tue Feb 07, 2006 1:03 am Post subject: Re: POP3 wrong decoding filename |
|
|
"Bert van der Hoeven" <GDvdHoeven (AT) planet (DOT) nl> wrote:
No.
| Quote: | or is indy not correct translating the name
|
Indy is not trying to translate the filename at all, because it is encoded using UTF-8, which Indy does not support at this time. The Indy developers are aware of the issue, but there is not much that can be done about it under Win32 since the string would have to be decoded into a WideString, which does not fit into Indy's current decoder architecture. UTF-8 support is currently being worked on for a future release under VCL.NET, where all strings are natively WideString.
| Quote: | or have i decode the filename myself?
|
At this point in time, yes. To decode it, look for strings that begin with '=?' and end with '?='. In between those two delimiters, there are three fields separated by '?'. The first field is the encoding used for the original data. The second field is the encoding used for transporting the data. The third field is the actual data.
So, given the string:
=?UTF-8?B?cGxhbm5pbmcgZmFzZSAwIHZlcnNpZSAzYi54bHM=?=
The data encoding is 'UTF-8'.
The transport encoding is 'B' (base64).
The data is 'cGxhbm5pbmcgZmFzZSAwIHZlcnNpZSAzYi54bHM='.
You would first decode the base64 data, and then second you would decode the resulting UTF-8 data to get the original value.
Gambit |
|
| Back to top |
|
|
theo
Guest
|
| Posted: Tue Feb 07, 2006 12:02 pm Post subject: Re: POP3 wrong decoding filename |
|
|
Bert van der Hoeven schrieb:
| Quote: | This a part of a e-mail file genarated with lotus notes, whith indy10
snapshot 6 feb. 2006 the filename is not decoded.
My question is: is the mail wrong or is indy not correct translating the
name or have i decode the filename myself?
|
With Synapse (http://www.ararat.cz/synapse/), this filenmame is
automatically decoded to "planning fase 0 versie 3b.xls" |
|
| Back to top |
|
|
Bert van der Hoeven
Guest
|
| Posted: Tue Feb 07, 2006 9:02 pm Post subject: Re: POP3 wrong decoding filename |
|
|
Thanks for this helpful information.
To pitty indy do not support this.
Can I use the Indy base64 decoder to decode the string?
and how can i decode UTF-8 to a delphi string?
"Remy Lebeau (TeamB)" <no.spam (AT) no (DOT) spam.com> schreef in bericht
news:43e7e12a$1 (AT) newsgroups (DOT) borland.com...
| Quote: |
"Bert van der Hoeven" <GDvdHoeven (AT) planet (DOT) nl> wrote:
is the mail wrong
No.
or is indy not correct translating the name
Indy is not trying to translate the filename at all, because it is encoded
using UTF-8, which Indy does not support at this time. The Indy
developers are aware of the issue, but there is not much that can be done
about it under Win32 since the string would have to be decoded into a
WideString, which does not fit into Indy's current decoder architecture.
UTF-8 support is currently being worked on for a future release under
VCL.NET, where all strings are natively WideString.
or have i decode the filename myself?
At this point in time, yes. To decode it, look for strings that begin
with '=?' and end with '?='. In between those two delimiters, there are
three fields separated by '?'. The first field is the encoding used for
the original data. The second field is the encoding used for transporting
the data. The third field is the actual data.
So, given the string:
=?UTF-8?B?cGxhbm5pbmcgZmFzZSAwIHZlcnNpZSAzYi54bHM=?=
The data encoding is 'UTF-8'.
The transport encoding is 'B' (base64).
The data is 'cGxhbm5pbmcgZmFzZSAwIHZlcnNpZSAzYi54bHM='.
You would first decode the base64 data, and then second you would decode
the resulting UTF-8 data to get the original value.
Gambit |
|
|
| Back to top |
|
|
Remy Lebeau (TeamB)
Guest
|
| Posted: Wed Feb 08, 2006 12:04 am Post subject: Re: POP3 wrong decoding filename |
|
|
"Bert van der Hoeven" <GDvdHoeven (AT) planet (DOT) nl> wrote in message
news:43e8fb20$1 (AT) newsgroups (DOT) borland.com...
| Quote: | To pitty indy do not support this.
|
It is not Indy's fault. Decoding UTF-8 strings almost always requires the
data to be placed into a WideString. The current architecture under Win32
is based on AnsiString instead of WideString.
| Quote: | Can I use the Indy base64 decoder to decode the string?
|
Yes.
| Quote: | how can i decode UTF-8 to a delphi string?
|
If you are using D6+, then the VCL has its own UTF-8 support functions, such
as UTF8Decode().
Alternatively, Indy 10 does have an IdUTF8 unit in the Protocols package.
You can use its UTF8ToUnicode() function.
Gambit |
|
| Back to top |
|
|
Bert van der Hoeven
Guest
|
| Posted: Wed Feb 08, 2006 9:05 pm Post subject: Re: POP3 wrong decoding filename |
|
|
Decoding works fine! Thanks to TeamB.
There is someting strange.
The filename starts with "=?UTF...... loking in then mail file (or
properties outlook express)
While debugging the code i found out that the file name starts with
"=_UTF......
The ? changed in _ underscore.
It is not a problem but i do no way this happens and if it can effect my
program.
Do anayone no?
Thanks |
|
| Back to top |
|
|
Remy Lebeau (TeamB)
Guest
|
| Posted: Wed Feb 08, 2006 10:05 pm Post subject: Re: POP3 wrong decoding filename |
|
|
"Bert van der Hoeven" <GDvdHoeven (AT) planet (DOT) nl> wrote in message
news:43ea4bce$1 (AT) newsgroups (DOT) borland.com...
| Quote: | While debugging the code i found out that the file name
starts with "=_UTF......
The ? changed in _ underscore.
|
At what point exactly? Please provide more details.
Gambit |
|
| Back to top |
|
|
Bert van der Hoeven
Guest
|
| Posted: Sat Feb 11, 2006 11:03 am Post subject: Re: POP3 wrong decoding filename |
|
|
| Quote: | At what point exactly? Please provide more details.
|
The message is loaded from a file in a TIdMessage.
MsgPart.loadfromfile......
loking directly in the file gives:
--0__=4EBBFB95DFD462628f9e8a93df938690918c4EBBFB95DFD46262
Content-Type: application/msexcel;
name="=?UTF-8?B?cGxhbm5pbmcgZmFzZSAwIHZlcnNpZSAzYi54bHM=?="
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
filename="=?UTF-8?B?cGxhbm5pbmcgZmFzZSAwIHZlcnNpZSAzYi54bHM=?="
Here is a snip from my code:
DS : string; //can this be the problem?
if (MsgPart.MessageParts.Items[intIndex] is TIdAttachmentFile) then
begin //general attachment
if TIdAttachmentFile(MsgPart.MessageParts.Items[intIndex]).ParentPart<0
then
begin //alleen zelfstandige bijlagen als parentpart<0
li := lvMessageParts.Items.Add;
if
uppercase(TIdAttachmentFile(MsgPart.MessageParts.Items[intIndex]).ContentType)='MESSAGE/RFC822'
then
begin
li.ImageIndex := 5;
li.Caption := 'E-mail bericht';
end
else
begin
li.ImageIndex := 8;
if
uppercase(TIdAttachmentFile(MsgPart.MessageParts.Items[intIndex]).ContentTransfer)='BASE64'
then
begin
//opdelen in juiste stukken.
DS:=TIdAttachmentFile(MsgPart.MessageParts.Items[intIndex]).Filename;
li.Caption:=DecodeBase64Data(DS);
//IdDecoderMIME1.Encode(TIdAttachmentFile(MsgPart.MessageParts.Items[intIndex]).Filename);
end
else li.Caption :=
TIdAttachmentFile(MsgPart.MessageParts.Items[intIndex]).Filename;
function SplitUTF8string(InDS : string;
var part1, part2: string;
var partData : string):Boolean;
var
DS,
UTF1,
UTF2 : string;
UTFData : string;
N : integer;
IPart : integer;
begin
DS:=InDS;
if (length(DS)>2) and (copy(DS,1,2)='=_') then
begin
IPart:=1;
UTF1:='';
UTF2:='';
UTFDATA:='';
for N:=3 to length(DS) do
begin
case IPart of
1 : begin
if DS[N]<>'_' then
UTF1:=UTF1+DS[N]
else
IPart:=2;
end;
2 : begin
if DS[N]<>'_' then
UTF2:=UTF2+DS[N]
else
IPart:=3;
end;
3 : begin
if DS[N]<>'_' then
UTFData:=UTFData+DS[N]
else
IPart:=4;
end;
end;
end;
Part1:=UTF1;
Part2:=UTF2;
PartData:=UTFData;
Result:=True;
end
else result:=False;
end;
function DecodeBase64Data(InData : string):string;
var
UTF1,
UTF2 : string;
UTFData : string;
DecoderMime : TIdDecoderMIME;
DS : string;
DUTF8 : UTF8String;
begin
Result:='';
if SplitUTF8string(InData,UTF1,UTF2,UTFData) then
begin
DecoderMime := TIdDecoderMIME.Create;
try
DUTF8:=DecoderMime.DecodeString(UTFData);
DS:=UTF8Decode(DUTF8);
Result:=DS;
finally
DecoderMime.Free;
end;
end
else result:=InData;
end;
I hope this gives the backgrount information.
Thanks,
Bert.
"Remy Lebeau (TeamB)" <no.spam (AT) no (DOT) spam.com> schreef in bericht
news:43ea5d6d$1 (AT) newsgroups (DOT) borland.com...
| Quote: |
"Bert van der Hoeven" <GDvdHoeven (AT) planet (DOT) nl> wrote in message
news:43ea4bce$1 (AT) newsgroups (DOT) borland.com...
While debugging the code i found out that the file name
starts with "=_UTF......
The ? changed in _ underscore.
At what point exactly? Please provide more details.
Gambit
|
|
|
| Back to top |
|
|
Ciaran Costelloe
Guest
|
| Posted: Sat Feb 11, 2006 11:03 pm Post subject: Re: POP3 wrong decoding filename |
|
|
Bert van der Hoeven wrote:
| Quote: | Decoding works fine! Thanks to TeamB.
There is someting strange.
The filename starts with "=?UTF...... loking in then mail file (or
properties outlook express)
While debugging the code i found out that the file name starts with
"=_UTF......
The ? changed in _ underscore.
|
With attachment filenames, Indy replaces any characters that are not
valid for a Windows filename with underscores. This is to avoid
problems with TIdAttachment's SaveToFile function. The ? is not valid
in a Windows filename since it is a DOS wild character.
Ciaran |
|
| Back to top |
|
|
Bert van der Hoeven
Guest
|
| Posted: Wed Feb 15, 2006 7:03 pm Post subject: Re: POP3 wrong decoding filename |
|
|
This make's sense. Thanks!
"Ciaran Costelloe" <ccostelloe (AT) flogas (DOT) ie> schreef in bericht
news:43ee6017$1 (AT) newsgroups (DOT) borland.com...
| Quote: | Bert van der Hoeven wrote:
Decoding works fine! Thanks to TeamB.
There is someting strange.
The filename starts with "=?UTF...... loking in then mail file (or
properties outlook express)
While debugging the code i found out that the file name starts with
"=_UTF......
The ? changed in _ underscore.
With attachment filenames, Indy replaces any characters that are not
valid for a Windows filename with underscores. This is to avoid
problems with TIdAttachment's SaveToFile function. The ? is not valid
in a Windows filename since it is a DOS wild character.
Ciaran |
|
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
