[assembler] structure of zipped dataset (archive)

[AcckiyGerman]

As the data get will now fetch and unpack archived version of the dataset - we should fix it has a fancy and valid structure.

How to reproduce

• data push gold-prices
• wait for processing
• data get http://datahub.io/<user>/gold-prices
• check the content you've got

Expected behavior

• [x] valid datapackage.json - you can do data validate and it is OK
• [x] README.md
• [x] data files in data folder
• [x] report.json exist
• [x] no folders with 'hash' names
• [x] no unneccessary files (preview.json, etc)
• [x] original source file is presented and not changed
• [x] user can see it is the source (source-preffix or suffix)"
• [x] pretty-printed datapackage.json

[zelima]

@AcckiyGerman pushes after today should have nice structure with readme and archive folders

[AcckiyGerman]

@zelima This is a very cool fix!

AcckiyGerman/finance-vix$ tree
.
├── archive
│   └── vix-daily.csv
├── data
│   ├── validation_report.json
│   ├── vix-daily.csv
│   ├── vix-daily_csv.csv
│   └── vix-daily_json.json
├── datapackage.json
└── README.md

I have only one suggestion:

$ head data/vix-daily.csv 
Date,VIXOpen,VIXHigh,VIXLow,VIXClose
2004-01-02,17.96,18.68,17.54,18.22
2004-01-05,18.45,18.49,17.44,17.49
2004-01-06,17.66,17.67,16.19,16.73
2004-01-07,16.72,16.75,15.05,15.05
2004-01-08,15.42,15.68,15.32,15.61
2004-01-09,16.15,16.88,15.57,16.75
2004-01-12,17.32,17.46,16.79,16.82
2004-01-13,16.06,18.33,16.53,18.04
2004-01-14,17.29,17.03,16.04,16.75

$ head data/vix-daily_csv.csv 
Date,VIXOpen,VIXHigh,VIXLow,VIXClose
2004-01-02,17.96,18.68,17.54,18.22
2004-01-05,18.45,18.49,17.44,17.49
2004-01-06,17.66,17.67,16.19,16.73
2004-01-07,16.72,16.75,15.05,15.05
2004-01-08,15.42,15.68,15.32,15.61
2004-01-09,16.15,16.88,15.57,16.75
2004-01-12,17.32,17.46,16.79,16.82
2004-01-13,16.06,18.33,16.53,18.04
2004-01-14,17.29,17.03,16.04,16.75

Seems the last two are the same file - could we get rid of one of them?

[zelima]

@AcckiyGerman Yes we can, but I did this on purpose - to be aligned with original dp.json (we do have the same duplicate in original dp.json as well) Let's close this one open after we discuss that.

[AcckiyGerman]

FIXED: User get a valid dataset with pretty folders and files structure after data get