685112 - Attached .doc file is replaced by the email's text upon Open, if broken mime-type like =?iso-8859-1?q?application/msword; is specified in Content-Type: header

Status: RESOLVED DUPLICATE of bug 659355

(MailNews Core :: Attachments, defect)

Description

[Frederic.Meynadier]

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:6.0.2) Gecko/20100101 Firefox/6.0.2
Build ID: 20110905175240

Steps to reproduce:

Config : TB 6.0.1, Ubuntu 11.04 64 bits, LibreOffice for reading word documents.

I am connected to an IMAP mail server.
I have received an email with an attached Word document (.doc format).
In the message panel, I click on the attachment's name.
A dialog box appears, proposing to open the file with LibreOffice writer : I click OK


Actual results:

LibreOffice launches, but what appears is a copy of the email's body text, instead of the intended attachment.

I remember it already occured some months ago, with an earlier version of TB. Back then I thought it was a sender's error.


Expected results:

I should have had the .doc file imported and displayed in LibreOffice writer.

Additionnal comments :
- I have opened the very same email from a webmail : the attachment is the right one.
- From thunderbird, if I save the attachment (right-click, "save" in the context menu), I get the right file too. 
- If I copy this mail in another folder, the issue is still there : I get the wrong content if I left-click on the attachment name, but I get the right content if I right-click and save the attachment.

Unfortunately I can't forward the mail itself as it contains professional informations.

Possibly relevant headers :

User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; fr; rv:1.9.2.21) Gecko/20110830 Thunderbird/3.1.13
MIME-Version: 1.0
Content-Type: multipart/mixed;
 boundary="------------050902040609080307040602"
This is a multi-part message in MIME format.
--------------050902040609080307040602

Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

[The email's text, which erroneously appears in LibreOffice]

--------------050902040609080307040602
Content-Type: =?iso-8859-1?q?application/msword;
 name="[filename].doc"
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
 filename="[filename].doc"

[The attachement's code in base64]

--------------050902040609080307040602--

Comment 1

[WADA:World Anti-bad-Duping Agency]

Attachment: Mail folder file which contains 6 test mails

Six mails has same structure and identical data, except (a) content-type of mail body part(text/plain or text/html), (b) content-type of part which is opened by application.

[Mail structure]
  Content-Type: multipart/mixed; boundary="Bdy"
  Part-1, mail body          : Content-Type: text/{Body_Type}
  Part-2, text attachment #1 [details] [diff] [review] : Content-Type: text/plain
  Part-3, .doc attachment    ; Content-Type: {Broken_or_Undefined}
  Part-4, text attachment #2 [details] [diff] [review] : Content-Type: text/plain

[Test mails]
              {Body_Type}   {Broken_or_Undefined}
  mail-A-01 :  plain         =?iso-8859-1?q?application/msword;
  mail-A-02 :  plain         unknown-major/unknown-sub;
  mail-A-03 :  plain         ?corrupted?major?/?broken?sub?;
  mail-B-01 :  plain         =?iso-8859-1?q?application/msword;
  mail-B-02 :  plain         unknown-major/unknown-sub;
  mail-B-03 :  plain         ?corrupted?major?/?broken?sub?;

When Part-3 is opened by Notepad via Open of context menu, shown data was:
  mail-x-01/mail-x-03 : HTML which has content of Part-1, Part-2, Part-4.
    Text in Part-2, Part-4, is embed in <PRE> tag.
    When x=A, text in Part-1(Mail Body) is also embed in <PRE> tag.
    When x=B, text in Part-1(Mail Body) is placed as-is becasue of text/html.
  mail-x-02 : Content of Part-2, as expectd.
 
Problem looks;
  If mime-type in Content-Type: is broken, all text parts is gathered,
  and converted to HTML, then passed to application.

Confirmed with Tb 6.0 and next trunk nightly build.
> Mozilla/5.0 (Windows NT 5.1; rv:9.0a1) Gecko/20110828 Thunderbird/9.0a1

Comment 2

[WADA:World Anti-bad-Duping Agency]

{Body_Type} for mail-B-nn was wrong. Read as next, please. Sorry for spam.
              {Body_Type}   {Broken_or_Undefined}
  mail-B-01 :  html          =?iso-8859-1?q?application/msword;
  mail-B-02 :  html          unknown-major/unknown-sub;
  mail-B-03 :  html          ?corrupted?major?/?broken?sub?;

Comment 3

[WADA:World Anti-bad-Duping Agency]

Bug 659355 was found for same external symptom.
> Bug 659355 Cannot open PDF attachment in some mails (Content-Type: =?windows-1252?q?application/pdf)
But that bug was for mail generated by Tb.
  1. mail of broken mime-type is sent to Tb user.
  2. when user tried to open the broken mime-type attachment, he requested "always
     open with the application", then bad mimeTypes.rdf entry was generated.
  3. when user send mail with attachment, Tb generates broken mime-type.
  4. thus, any Tb user can't open the attachment of broken mime-type(this bug)
That bug now focuses on 2./3. instead of 4.
To analyze/resolve problem of 4., this bug should be kept open separately.

IIRC, phenomenon of this bug was already reported in some bugs. I'll try to search such bugs.
I think this bug is better kept, because crisp bug report and pointing to "email's text" is done by you.

Comment 4

[WADA:World Anti-bad-Duping Agency]

A bug I saw in the past was Bug 580971, and was for broken mime-type of =?windows-1252?q?application/pdf;. This bug and that bug are report of identical problem.
> Bug 580971 Thunderbird gets an incorrect mime part when opening an attachment of a received message with malformed content type

Comment 5

[WADA:World Anti-bad-Duping Agency]

Text shown by Notepad when Open of the .doc part(mail-A-01/mail-A-03) 
> <div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> [The email's text, which erroneously appears in LibreOffice]
> </pre></div><BR><FIELDSET CLASS="mimeAttachmentHeader"></FIELDSET><BR/><div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> Text of correct text/plain part #1
> </pre></div><BR><FIELDSET CLASS="mimeAttachmentHeader"></FIELDSET><BR/><div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> Text of correct text/plain part #2
> </pre></div>

Text shown by Notepad when Open of the .doc part(mail-B-01/mail-B-03) 
> <p>[The email's text, which erroneously appears in LibreOffice]</p>
> <BR><FIELDSET CLASS="mimeAttachmentHeader"></FIELDSET><BR/><div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> Text of correct text/plain part #1
> </pre></div><BR><FIELDSET CLASS="mimeAttachmentHeader"></FIELDSET><BR/><div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> Text of correct text/plain part #2
> </pre></div>

If mail body is text/html, and if text part is mail body only, phenomenon looks "text/html part(or mail body) is selected instead of part to which Open is requested".

Comment 6

[WADA:World Anti-bad-Duping Agency]

Attachment: Test mail to see "funy cid: url replacement" and "what is shown by Open of attachment with broken mime-type"

As funny cid: url was seen with test mail of Bug 580971, I checked "html mail which uses cid: url" case.

[Mail structure]
  multipart/mixed mail
    part-1 : multipart/related
      part-1-1 : text/html with <iframe src="cid:...">
      part-1-2 : text/plain part referred by the cid: url
    part-2 : Content-Type: ?broken?/?corrupted?
    part-3 : multipart/related
      part-3-1 : text/html with <iframe src="cid:...">
      part-3-2 : text/plain part referred by the cid: url
    part-4 : Content-Type: ?broken?/?corrupted?
    part-5 : multipart/related
      part-5-1 : text/html with <iframe src="cid:...">
      part-5-2 : text/plain part referred by the cid: url

[Data shown by Notepad]

(1) View/Message Body As/... HTML, Display Attachment Inline=Off
(part-2 is opened)
> <html><head></head><body><p>[This is mail body  #1]</p><div><iframe src="mailbox:///C|/wada/@@@/Mail0/bug-685112-02?number=0&part=1.1.2&filename=brokendocfile-1.doc&type=?broken?/?corrupted?&filename=brokendocfile-1.doc&filename=sigfile_1.txt" width="320" height="40~>This is iframe 1</iframe></div></body></html>
(part-4 is opened)
> <html><head></head><body><p>[This is mail body  #1]</p><div><iframe src="mailbox:///C|/wada/@@@/Mail0/bug-685112-02?number=0&part=1.1.2&filename=brokendocfile-2.doc&type=?broken?/?corrupted?&filename=brokendocfile-2.doc&filename=sigfile_1.txt" width="320" height="40~>This is iframe 1</iframe></div></body></html>

(2) View/Message Body As/... HTML, Display Attachment Inline=On
(part-2 is opened)
> <html><head></head><body><p>[This is mail body  #1]</p><div><iframe src="mailbox:///C|/wada/@@@/Mail0/bug-685112-02?number=0&part=1.1.2&filename=brokendocfile-1.doc&type=?broken?/?corrupted?&filename=brokendocfile-1.doc&filename=sigfile_1.txt" width="320" height="40~>This is iframe 1</iframe></div></body></html>
> <BR><FIELDSET CLASS="mimeAttachmentHeader"><LEGEND CLASS="mimeAttachmentHeaderName">html_attach_02.html</LEGEND></FIELDSET><BR/><html><head></head><body><p>[Attached html #2]</p><div><iframe src="mailbox:///C|/wada/@@@/Mail0/bug-685112-02?number=0&part=1.3.2&filename=brokendocfile-1.doc&type=?broken?/?corrupted?&filename=brokendocfile-1.doc&filename=sigfile_2.txt" width="320" height="40~>This is iframe 2</iframe></div></body></html>
> <BR><FIELDSET CLASS="mimeAttachmentHeader"><LEGEND CLASS="mimeAttachmentHeaderName">html_attach_03.html</LEGEND></FIELDSET><BR/><html><head></head><body><p>[Attached html #3]</p><div><iframe src="mailbox:///C|/wada/@@@/Mail0/bug-685112-02?number=0&part=1.5.2&filename=brokendocfile-1.doc&type=?broken?/?corrupted?&filename=brokendocfile-1.doc&filename=sigfile_3.txt" width="320" height="40~>This is iframe 3</iframe></div></body></html>
(part-4 is opened)
> <html><head></head><body><p>[This is mail body  #1]</p><div><iframe src="mailbox:///C|/wada/@@@/Mail0/bug-685112-02?number=0&part=1.1.2&filename=brokendocfile-2.doc&type=?broken?/?corrupted?&filename=brokendocfile-2.doc&filename=sigfile_1.txt" width="320" height="40~>This is iframe 1</iframe></div></body></html>
> <BR><FIELDSET CLASS="mimeAttachmentHeader"><LEGEND CLASS="mimeAttachmentHeaderName">html_attach_02.html</LEGEND></FIELDSET><BR/><html><head></head><body><p>[Attached html #2]</p><div><iframe src="mailbox:///C|/wada/@@@/Mail0/bug-685112-02?number=0&part=1.3.2&filename=brokendocfile-2.doc&type=?broken?/?corrupted?&filename=brokendocfile-2.doc&filename=sigfile_2.txt" width="320" height="40~>This is iframe 2</iframe></div></body></html>
> <BR><FIELDSET CLASS="mimeAttachmentHeader"><LEGEND CLASS="mimeAttachmentHeaderName">html_attach_03.html</LEGEND></FIELDSET><BR/><html><head></head><body><p>[Attached html #3]</p><div><iframe src="mailbox:///C|/wada/@@@/Mail0/bug-685112-02?number=0&part=1.5.2&filename=brokendocfile-2.doc&type=?broken?/?corrupted?&filename=brokendocfile-2.doc&filename=sigfile_3.txt" width="320" height="40~>This is iframe 3</iframe></div></body></html>

(3) View/Message Body As/Plan Text, Display Attachment Inline=On
(part-2 is opened)
> <div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> [This is mail body #1]
> </pre></div>
> <BR><FIELDSET CLASS="mimeAttachmentHeader"><LEGEND CLASS="mimeAttachmentHeaderName">html_attach_02.html</LEGEND></FIELDSET><BR/><div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> [Attached html #2]
> </pre></div>
> <BR><FIELDSET CLASS="mimeAttachmentHeader"><LEGEND CLASS="mimeAttachmentHeaderName">html_attach_03.html</LEGEND></FIELDSET><BR/><div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> [Attached html #3]
> </pre></div>
(part-4 is opened)
> <div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> [This is mail body #1]
> </pre></div>
> <BR><FIELDSET CLASS="mimeAttachmentHeader"><LEGEND CLASS="mimeAttachmentHeaderName">html_attach_02.html</LEGEND></FIELDSET><BR/><div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> [Attached html #2]
> </pre></div><BR><FIELDSET CLASS="mimeAttachmentHeader"><LEGEND CLASS="mimeAttachmentHeaderName">html_attach_03.html</LEGEND></FIELDSET><BR/><div class="moz-text-plain" wrap=true graphical-quote=true style="font-family: -moz-fixed"><pre wrap>
> [Attached html #3]
> </pre></div>

Phenemenon was;
(i) Data was "email's body text" as you say, instead of "gathered text parts".
    "Internal HTML source currently used for mail display by UI"
     was passed to application.
(ii) cid: url in <iframe> refers correct part number(&part= portion).
     However, &type= and &filename=  portion of the cid: url looks somehow
     replaced by mime-type and filename of opened attachment.

Symptom was not "wrong part(mail body part) is selected" which Bug 580971 says.

Comment 7

[WADA:World Anti-bad-Duping Agency]

Creation of such corrupted Content-Type: by Tb is due to bug 503309.
Setting dependency.

Comment 8

[WADA:World Anti-bad-Duping Agency]

Fix was proposed to Bug 659355, so closing as DUP.