478176 - multipart/signed does not generate JS emitter events for its immediate payload child

Status: RESOLVED FIXED

(MailNews Core :: MIME, defect)

Description

[Andrew Sutherland [:asuth] (he/him)]

Attachment: v1 emit jsemitter events for the payload child of multipart/signed parts

Because multipart/signed (MimeMultipartSigned) defers processing of its payload child, the super-class antics of multipart/* (MimeMultipart) never generate the events that the JS emitter (added in bug 447842) is expecting.  However, when the payload child is eventually processed, if it has any children, the JS emitter event chain does happen.

Example.  Actual MIME structure:
1 multipart/mixed
  1.1 multipart/signed
    1.1.1 multipart/alternative
      1.1.1.1 text/plain
      1.1.1.2 text/html
    1.1.2 application/x-pkcs7-signature
  1.2 text/plain

MIME structure perceived by JS emitter
1 multipart/mixed
  1.1 multipart/signed
    g-g-ghost!
      1.1.1.1 text/plain
      1.1.1.2 text/html
    (signature also not seen, but no one cares)
  1.2 text/plain

The JS emitter is afraid of ghosts and freaks out when it hears about the text/plain.

The provided patch causes the expected series of events to be generated for 1.1.1.

(Note, this blocks 465113, which is itself a blocker.)

Comment 1

[David :Bienvenu]

Comment on attachment 361953 [details] [diff] [review]
v1 emit jsemitter events for the payload child of multipart/signed parts

I've managed to get this code to trigger from gloda, and nothing bad seems to happen.

If you wanted to, you could tighten this up a little -

+    if (ct) {
+      mimeEmitterAddHeaderField(obj->options, HEADER_CONTENT_TYPE, ct);
+      PR_Free(ct);
+    }
+    else {
+      // no content type means text/plain
+      mimeEmitterAddHeaderField(obj->options, HEADER_CONTENT_TYPE,
+                                "text/plain");
+    }

could simply be:

+      mimeEmitterAddHeaderField(obj->options, HEADER_CONTENT_TYPE, ct ? ct : "text/plain");
+      PR_Free(ct);

because PR_Free(null) is harmless.

Comment 2

[Andrew Sutherland [:asuth] (he/him)]

tightened up and pushed: http://hg.mozilla.org/comm-central/rev/e94f4c8a4a63