Fixing Attachment Error New Fork

[Gamma169]

From PR go-gomail#97

[ivy]

It's generally not a good practice to panic unless absolutely necessary. Have a look at using t.Errorf("teardown: %s", err) instead.

[ivy]

@Gamma169 Thanks for making time for this over your weekend. On further consideration, I don't think we'll be able to benefit from this improvement without sacrificing usability/features somewhere else:

• SetCopyFunc allows replacing the underlying file-backed copier with anything that acts on an io.Writer (as you see in the mocked tests). We need something that still allows this functionality as it's useful in other areas outside of testing such as in Accept io.Reader for Attach and Embed go-gomail/gomail#78.
• Rename makes things trickier as we also can't rely on the filename matching what's on disk.

I was considering simply adding an error return value to Embed and Attach but that also breaks SetCopyFunc for non-file-backed attachments. To sum things up, we're going to have to wait on this until there's a solid plan for v3.

[pedromorgan]

@ivy why?

this is a good patch, got a lot of stuff in there..

at least create the v3 branch..

[pedromorgan]

@ivy serious do not reject.. accept everything as a possibility..

and id u do that.. then u have a whole heap behind u,,

Thing u a good "leader" in making noise..
But this is code ok...
and its differrent kinda respect.....and trust

[Gamma169]

@ivy Understood about the SetCopyFunc function. @pedromorgan The problem is with that function, a person can attach data and label it as a file that doesn't exist. Thus triggering the error on the send, when there actually is no file.

I can think about things a little more and figure out a way to get rid of the bug (because I think it's pretty egregious).