Make tapes human readable based on content-type and content-enconding

[ijpiantanida]

Proposal for #2.

There's a whitelist of content-types that are considered human readable.
If the content-type is human readable and the request body is in plain text
(no content-enconding header or content-enconding == identity) then the
body is saved in UTF-8. Otherwise, keep using base64.

More work can be done on this feature (like decompressing the body based on the content-enconding used and then re-compressing when serving the tape) but I'd first like to check what do you think about this approach.

[mwolson]

Hey all, wanted to bump this since we're getting some further feedback from other folks inside Ticcketmaster who would like to see this in order to increase adoption.

[eadmundo]

this would be great to get into an npm release - is this project still being maintained by flickr though? last commit to master was Aug 2016. @jeremyruppel ?

[Monte9]

Can we please get this merged in? Trying to decode every request is pretty cumbersome

If not, is there a fork out there that has these changes? I'd be happy to switch over as this repo doesn't seem to be maintained anymore.

[yahoocla]

Thank you for submitting this pull request, however I do not see a valid CLA on file for you. Before we can merge this request please visit https://yahoocla.herokuapp.com/ and agree to the terms. Thanks! 😄

[Monte9]

Update: Ended up using a replacement library called Talkback.

More on it here: #49