| Ticket UUID: | 3483716 | |||
| Title: | transfer encoding ignored for message/* | |||
| Type: | Bug | Version: | None | |
| Submitter: | cnassau | Created on: | 2012-02-03 10:21:10 | |
| Subsystem: | mime | Assigned To: | andreas_kupries | |
| Priority: | 5 Medium | Severity: | ||
| Status: | Closed | Last Modified: | 2012-02-24 00:37:05 | |
| Resolution: | Fixed | Closed By: | andreas_kupries | |
| Closed on: | 2012-02-23 17:37:05 | |||
| Description: |
The handling of message/* types in ::mime::parsepart currently ignores the Content-Transfer-Encoding (tcllib 1.13 / mime 1.5.4) . Example Content-Type: message/delivery-status; name="deliverystatus.txt" Content-Disposition: attachment; filename="deliverystatus.txt"; size=138; creation-date="Thu, 02 Feb 2012 13:50:05 GMT"; modification-date="Thu, 02 Feb 2012 13:50:05 GMT" Content-Description: deliverystatus.txt Content-Transfer-Encoding: base64 T3JpZ2luYWwtUmVjaXBpZW50OiA8L2ZheD1ibHViYkBndW1taS5ib290PgpBY3Rpb246IGZhaWxl ZApEaWFnbm9zdGljLUNvZGU6IHNtdHA7IDU1MCAjNS4xLjAgQWRkcmVzcyByZWplY3RlZC4KUmVt b3RlLU1UQTogNTMuMjQuMjgyLjE1MA== The attached patch fixes this problem for base64 encoded messages. | |||
| User Comments: |
andreas_kupries added on 2012-02-24 00:37:05:
allow_comments - 1 andreas_kupries added on 2012-02-24 00:37:04: Thank you very much. The test has been added now. Confirmed that with your patch disabled an error is thrown, and with it the data is properly decoded. I agree regarding the reasoning about accepting the base64 encoded form. cnassau added on 2012-02-23 16:13:42: Here is a test that verifies that the base64-encoded "message/*" is parsed. Previously that code raised an error.
test mime-12.0 {Bug 3483716} {
set token [mime::initialize -string {Content-Type: message/delivery-status; name="deliverystatus.txt"
Content-Disposition: attachment; filename="deliverystatus.txt"; size=138;
creation-date="Thu, 02 Feb 2012 13:50:05 GMT";
modification-date="Thu, 02 Feb 2012 13:50:05 GMT"
Content-Description: deliverystatus.txt
Content-Transfer-Encoding: base64
T3JpZ2luYWwtUmVjaXBpZW50OiA8L2ZheD1ibHViYkBndW1taS5ib290PgpBY3Rpb246IGZhaWxl
ZApEaWFnbm9zdGljLUNvZGU6IHNtdHA7IDU1MCAjNS4xLjAgQWRkcmVzcyByZWplY3RlZC4KUmVt
b3RlLU1UQTogNTMuMjQuMjgyLjE1MA==
}]
set parts [mime::getproperty $token parts]
mime::getheader [lindex $parts end] Remote-MTA
} 53.24.282.150
Side note: I believe the test message does not follow the respective RFC 1894 because that states that 7bit encoding "MUST" be used for message/delivery-status types. The base64-encoded version that broke the previous package was observed in the wild, however, so it seems wise to accept it noentheless.
andreas_kupries added on 2012-02-23 01:54:34: Accepted, and committed to CVS head. However, please provide me with a small script demonstrating the before/after behaviour, so that I can extend the testsuite Side note: In the future, please provide patches in the 'uified diff' format (generated by 'diff -u'). This easier to read, and easier to apply. cnassau added on 2012-02-03 17:21:16: File Added - 434791: patch | |||
Attachments:
- patch [download] added by cnassau on 2012-02-03 17:21:15. [details]