Archive original KML/GPX files

[who8mycakes]

This PR adds steps to retain/archive the original KML or GPX file uploaded by a user.

Additionally, I added tests which assert:

• that the archived file is created in the outdirectory
• that the contents of the original infile match the contents of the archived file

@GretaCB Two questions I have are:

• Should digest be run against the infile before writing the contents to archived.* or does that create a file different from the original file?
• Should a random string of integers be added to the archive.* filename to differentiate between multiple archived files, or are individual outdirectories created per upload?

cc @mapbox/unpacker

[GretaCB]

Should a random string of integers be added to the archive.* filename to differentiate between multiple archived files, or are individual outdirectories created per upload?

No need in this case, since each upload is its own separate job on the EC2. Even if more jobs were being run on the EC2 at a time, the dirname created is randomized.

[who8mycakes]

@GretaCB Got the control flow issue sorted by nesting the functions:

• https://github.com/mapbox/preprocessorcerer/blob/archive-original/preprocessors/togeojson-kml.preprocessor.js#L105
• https://github.com/mapbox/preprocessorcerer/blob/archive-original/preprocessors/togeojson-gpx.preprocessor.js#L95

I attempted to test this on staging today, however, testing was unsuccessful due to the lack of some environment credentials. I will take another stab at it tomorrow.

[who8mycakes]

I was able to test this branch with the related bundle-fairy PR (https://github.com/mapbox/bundle-fairy/pull/9), and individual kml and gpx uploads were 🍏 , and the resulting zipped bundle included an archived file. I will run through some invalid bundles, kml, and gpx file test cases tomorrow.

[who8mycakes]

@mapsam Can you please take a look here before I start finishing steps?