imap, doc: documented stuffs on imapsession

[yorkie]

I documented some stuffs and now updates the TODO List:

• Documentation of each method in IMAPSession.js, parameters and examples, returned data.
• appendMessage flags and date parameters should be made more nodeJS friendly + they should be made optional through an option parameter.
• fetchMessagesByNumber / fetchMessagesByUid option parameter should also be made more nodeJS friendly.
• remove options parameter of copyMessages: use only UID.
• add tests(what i added)

LGTY?

[dinhvh]

Would be nice to document what option can contain.

[dinhvh]

before showing the example.

[dinhvh]

You should split README.md in two:

• One to describe what is the project and show a quick example.
• Documentation.md, which shows a more complete documentation.

[dinhvh]

Do you want me to merge it and make progress in an other pull request?

[yorkie]

You should split README.md in two

Ok, thanks, good idea :)

Do you want me to merge it and make progress in an other pull request?

In this Pull Request, we should fix the whole TODO listed below IMO, so may i ask u merge this once all the TODO has been strike.

And I'd like to do your comments tomorrow :)

[yorkie]

fixed something and still leaves few questions:

appendMessage flags and date parameters should be made more nodeJS friendly + they should be made optional through an option parameter.

Can I create a new class named: IMAPMessage and let appendMessage be called like:

var msg = new IMAPMessage();
msg.date = new Date();
msg.flags.push('\Seen');
// ...some sets to `msg`
session.appendMessage(folder, message);

fetchMessagesByNumber / fetchMessagesByUid option parameter should also be made more nodeJS friendly.

What i think the best way to write the option just passing the command directly, like:

session.fetchMessagesByUID(0, 1, 'HEADER.FIELDS (FROM TO SUBJECT DATE)');

and for fresh we provide a default value for options: ALL, then they don't care about the option should be, however if they need, see the RFC3501 to learn how to concat their fetch command.

How about this? I need your review again, @dinhviethoa
Thank you :)

[dinhvh]

IMAPMessage already stands for something else in mailcore2. It's the result of fetchMessage...().
I don't think it's a good idea to introduce a new data type for the purpose of appendMessage().

For fetch options, in mailcore2, there's a bunch of options in a binary mask:

        IMAPMessagesRequestKindUid           = 0, // This is the default and it's always fetched
        IMAPMessagesRequestKindFlags         = 1 << 0,
        IMAPMessagesRequestKindHeaders       = 1 << 1,
        IMAPMessagesRequestKindStructure     = 1 << 2,
        IMAPMessagesRequestKindInternalDate  = 1 << 3,
        IMAPMessagesRequestKindFullHeaders   = 1 << 4,
        IMAPMessagesRequestKindHeaderSubject = 1 << 5,
        IMAPMessagesRequestKindGmailLabels   = 1 << 6,
        IMAPMessagesRequestKindGmailMessageID = 1 << 7,
        IMAPMessagesRequestKindGmailThreadID  = 1 << 8,
        IMAPMessagesRequestKindExtraHeaders  = 1 << 9,
        IMAPMessagesRequestKindSize          = 1 << 10,

We could map them in JS to something like ['uid', 'fullheaders', 'gmail-labels'].

[yorkie]

then how we handle the PEEK in this option?

[dinhvh]

I'd suggest to do always PEEK. And let people mark explicitly as read with a storeFlags() API.

[yorkie]

How do u think the naming of the element of the option? I prefer use a-b for all, do you?

[yorkie]

Ok, updated document for fetch functions, if it good to u, i will update it to source code :)

[yorkie]

and delete(complete) 3rd task in the top of this conversation.

[dinhvh]

How do u think the naming of the element of the option? I prefer use a-b for all, do you?

Sounds good to me.

[dinhvh]

Can we make flags optional?

[yorkie]

Yep

On Wed, Jul 2, 2014 at 12:47 AM, Hoà V. DINH [email protected]
wrote:

In docs/API.md:

+* text
+
+Otherwise like RFC3501, you should assign the option to ALL, FAST or FULL to do a shortcut.
+
+#### imap.search(query, callback)
+
+* query String
+* callback Function
+
+The SEARCH command implementation.
+
+#### imap.appendMessage(folder, data, flags, callback)
+
+* folder String
+* data String, rfc822 message data
+* flags Array

Can we make flags optional?

—
Reply to this email directly or view it on GitHub
https://github.com/MailCore/mailcore.node/pull/2/files#r14415276.

Yorkie Neil
https://github.com/yorkie
+18600219033

[yorkie]

Ok, done :-)

[dinhvh]

Could you document or link to some possible values for flags?

[dinhvh]

Remove this line. We don't want people to use ALL, FAST, FULL, etc.

[dinhvh]

You still need to document the return value of fetchMessages...() in the callback.

[yorkie]

Ok, fix some comments :-)

[dinhvh]

Couldn't it be ['uid', 'flags', {'custom-headers': ['X-Mailer', 'X-Enveloper-Sender']}]?

[yorkie]

Ah, i think it is too nested :(

[dinhvh]

how about: 'uid', 'flags', 'custom-header:X-Mailer', 'custom-header:X-Mailing-List'?

[yorkie]

the custom-header and X- prefix seems act a same actor? i'm not sure if there is a custom header that doesn't contain a X- prefix?

[dinhvh]

There are custom header that contain a X-Prefix. And some headers are not listed as part of the "headers" or "full-headers".

[yorkie]

Hi, there is a problem on implementation:

fetchAtt = [{type:Constants.FetchEnvelope}, {type:Constants.FetchBodySection, param:'1.2'}]

This is a line that you write at libetpan.node, and could you provide a way of add custom header fetch item?

[yorkie]

for example, how do i add the x-mailer to the fetchAtt?

[dinhvh]

Have a look at _fetchTypeToString() in imapbase.js. You can use {type: Constants.FetchBodySection, section:'HEADER.FIELDS (x-mailer mailing-list)'}.
I think the documentation in comments is wrong. It should be section and not param.

[yorkie]

Thanks, i see

[dinhvh]

Also, you should start implementation.