Overview improvements about runbox7

[castaway]

• Message dates already exist in the search message data so these should be straight forward to add

• For proper mailing list support we will need to so some header parsing: either in the indexer, adding more flags (type of mail), in javascript (using the message cache), or in the API. The proper way would be in the indexer .. this one will be more involved

• We're showing "number of messages" now. I assume "identical subjects" is for "different conversations" (per sender)? We should be able to do this with existing conversation info in the index (I hope!)

• For the diagrams/charts, will need some examples for input/outputs :)

[gtandersen]

• Message dates already exist in the search message data so these should be straight forward to add

Excellent!

* For proper mailing list support we will need to so some header parsing: either in the indexer, adding more flags (type of mail), in javascript (using the message cache), or in the API. The proper way would be in the indexer .. this one will be more involved

We should do it the proper way by expanding the index with another flag/field for type of message, which might be useful in other contexts as well.

* We're showing "number of messages" now. I assume "identical subjects" is for "different conversations" (per sender)? We should be able to do this with existing conversation info in the index (I hope!)

The idea is to aid users in reviewing their email more efficiently by grouping senders and then subject lines. While the former is already (mostly) supported, the latter would need some additional intelligence since subject lines can (especially in the case of automated messages) be identical or almost identical.

Examples of this are receipts, notifications about social media events, and various service alerts -- all of which are often identical except for a number or keyword that might differ.

Since the indices can be searched manually by using single terms or quoted phrases, could this be employed in the reverse by starting with a list of all subject lines and first grouping the identical ones, then gradually removing one word string from the end (and separately from the beginning) of each line in order to group subject lines that are identical except for one (or two) word strings?

* For the diagrams/charts, will need some examples for input/outputs :)

The bars indicating the percentage of the maximum number of messages per sender can be shown in each sender's entry. Attached screenshot visualizes a possible implementation.