Email's attachments showed in Outlook 2013 as .DAT

Pietro_Beccegato · February 24, 2016, 11:43am

My app sends emails with attachments (pdf/xls/ecc).
Everything works fine; emails are sended and the recipient receive the emails.

This is the problem:
If I open the email through a webmail, the attachment has the correct filename, correct extension and I can open it.
If I use Outlook 2013 to open the email, the attachment has a different name and “DAT” as extension. So if I save the xxx.DAT and rename it as .PDF or .XLS, I can open it.

If I send the same email with thunderbird, the recipients receive the attachment correctly also using Outlook 2013.

I googled to search something to solve the problem, but after some test, I still got the problem.

Any idea?

Beatrix_Willius · February 24, 2016, 12:05pm

Have a look at the source of the email to see the difference between the mail that Xojo sends and what Thunderbird does. Which Xojo version do you use? I’ve had a problem with getting duplicate attachment names in 2015r4. Instead of “Session Log data time.txt.gz” I’m getting “Session Log data time.txt.gzSession Log data time.txt.gz”.

Christian_Schmitz · February 24, 2016, 12:48pm

How do you send email?

via SMTPSocket? or via CURLSMBS class?

Pietro_Beccegato · February 24, 2016, 12:49pm

Via SMTPSocket.

Pietro_Beccegato · February 24, 2016, 12:53pm

This is a suspected part of email’s source compared:
From Xojo code:

Content-Type: application/base64; name="CLIORD-00647-2015.pdf"; name*=utf-8''CLIORD-00647-2015.pdf x-mac-type="3F3F3F3F" x-mac-creator="3F3F3F3F" Content-transfer-encoding: Base64 Content-disposition: attachment; filename="CLIORD-00647-2015.pdf"; filename*=utf-8''CLIORD-00647-2015.pdf

From Thunderbird:

Content-Type: application/pdf; name="CLIORD-00647-2015.pdf" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="CLIORD-00647-2015.pdf"

Sam_Rowlands · February 25, 2016, 3:09am

It seems that the code generated by your Xojo app is confusing Outlook, it’s been a long time since I worked with e-mail apps, but the Content-Type: application/base64; seems like a potential suspect.

I would also send yourself a PDF via Outlook just to double check what does work in Outlook. Don’t be surprised if you get something totally different.

I had to write a program to extract the raw files from e-mails that our accountant sends, all we get is .dat files from him (he uses Outlook something or rather).

Beatrix_Willius · February 25, 2016, 5:27am

This looks like a variation of my bug with the duplicate file name. See <https://xojo.com/issue/42655>.

Try to remove the parts with name* and filename*. As far as I remember the * means concatenation.

Pietro_Beccegato · February 25, 2016, 5:35am

[quote]
This looks like a variation of my bug with the duplicate file name. See Feedback Case #42655.
Try to remove the parts with name* and filename*. As far as I remember the * means concatenation.[/quote]

Yes, it seems very similar at Feedback Case #42655

This a the part of my xojo code

file = New EmailAttachment file.loadFromFile Element mail.Attachments.Append file

so the code

Content-Type: application/base64; name="CLIORD-00647-2015.pdf"; name*=utf-8''CLIORD-00647-2015.pdf x-mac-type="3F3F3F3F" x-mac-creator="3F3F3F3F" Content-transfer-encoding: Base64 Content-disposition: attachment; filename="CLIORD-00647-2015.pdf"; filename*=utf-8''CLIORD-00647-2015.pdf

is automatically inserted by Xojo and I don’t know how I can obtain a code as one generated in thunderbird

Content-Type: application/pdf; name="CLIORD-00647-2015.pdf" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="CLIORD-00647-2015.pdf"

Beatrix_Willius · February 25, 2016, 6:00am

I’ve use a simple regex to remove the parts that I don’t need.

Can you add your information to the case? Xojo can’t reproduce the problem.

Beatrix_Willius · February 25, 2016, 10:09am

The issue has been verified.

Pietro_Beccegato · February 25, 2016, 10:26am

How do you use regex to remove the parts you don’t need?

Beatrix_Willius · February 25, 2016, 1:04pm

I only have gz attachments. It’s working half so far. The source of the email looks good but Mail is duplicating the attachments (not the names) now.

dim temp(-1) as string = StringUtils.SplitByRegEx(MailToCompany.Source, "filename\\*.*gz", false) dim tempString as string = Join(temp, "") temp = StringUtils.SplitByRegEx(MailToCompany.Source, "name\\*.*gz", false) tempString = Join(temp, "") MailToCompany.Source = tempString

SplitByRegex:

[code] Dim out(-1) As String
Dim startPos As Integer

Dim re As New RegEx
re.SearchPattern = delimPattern
Dim rm As RegExMatch = re.Search(source)
while rm <> nil
out.Append MidB(source, startPos + 1, rm.SubExpressionStartB(0) - startPos)
startPos = re.SearchStartPosition
rm = re.Search
wend

if startPos < source.LenB then out.Append MidB(source, startPos + 1)
if UBound(out) > 0 then return out[/code]

Beatrix_Willius · February 25, 2016, 1:07pm

Ugh… Mail is such a diva. Now the mails show up blank.

Pietro_Beccegato · February 25, 2016, 2:57pm

By first, thanks to Beatrix Willius for the EmailMessage.source tip !

I edited the EmailMessage.source with this method:

[code]Function CleanEmailSource(EmailSource as String) As String
Dim Rows() As String = Split(EmailSource, EndOfLine)

For i as integer=0 to Rows.Ubound
Dim row as String = Rows(i)
if InStr(0,Row,“x-mac-type”)>0 Then Row =""
if InStr(0,Row,“x-mac-creator”)>0 Then Row =""
if InStr(0,Row,“filename*=”)>0 or InStr(0,Row,“name*=”)>0 Then
Dim s() as String = Split(Row, “;”)
call s.Pop
Row = Join(s, “;”) + “;”
End If

rows(i)=Row

Return result
End Function
[/code]

Now the received email source has no more “filename*=” code, but using Thunderbird, the email appears as up blank also for me… but… if I open the email through webmail, it appears correctly.

Brandon_Warlick · February 25, 2016, 3:35pm

After some trial and error and testing with various iphones, this code ended up working for me. The recipient received that PDF as an attachment that was recognized as a PDF.

'This code creates the correct email header
oAttachment.LoadFromFile(oFileList.Item(i))
oAttachment.MIMEType = "application/pdf"

Email Header:
Content-Type: application/pdf; name=“testfile.pdf”;
x-mac-type=“3F3F3F3F”
x-mac-creator=“3F3F3F3F”
Content-transfer-encoding: Base64
Content-disposition: attachment

Pietro_Beccegato · February 25, 2016, 3:38pm

[quote]Brandon Warlick

'This code creates the correct email header
oAttachment.LoadFromFile(oFileList.Item(i))
oAttachment.MIMEType = “application/pdf”[/quote]
The attachment’s filename received is the same of the one sended?

Christian_Schmitz · February 25, 2016, 3:46pm

3F3F3F3F for the Mac type looks strange.

Beatrix_Willius · February 25, 2016, 3:50pm

The funny thing is I’ve got 2 computers running the Captain. One shows a blank mail and the other shows the mail fine.

An old mail has the following for the attachment:

----==_mimepart_569ddc29187f9_531d3fba133aefdc366933
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
filename=Session_Log_2016-01-19_07-37-00.txt.gz
Content-Type: application/x-gzip;
charset=UTF-8;
filename=Session_Log_2016-01-19_07-37-00.txt.gz;
name=Session_Log_2016-01-19_07-37-00.txt.gz
Content-Id: 569ddc28b03e4_531d3fba133aefdc3662c0@worker6.mail

I’ll try to get to this tomorrow.

Greg_O_Lone · March 1, 2016, 3:05pm

That’s Mail deciding that you’ve already seen this content and don’t need to see it again. If you double-click the message to open a new message window, you should be able to see the entire contents.

FWIW - This mechanism is really handy when you’re viewing a long threaded email conversation

Pietro_Beccegato · March 1, 2016, 3:13pm

I created a <https://xojo.com/issue/42788> for a similar trouble