[WIP] Extract missing values (to data)

[roll]

• fixes https://github.com/frictionlessdata/datapackage-pipelines/issues/177

[coveralls]

Pull Request Test Coverage Report for Build 434

• 30 of 30 (100.0%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.2%) to 85.251%

---

Totals
Change from base Build 430:	0.2%
Covered Lines:	1734
Relevant Lines:	2034

---

💛 - Coveralls

[akariv]

Looks good @roll - do you want to merge this?

[roll]

Hey @akariv,

Sorry I've forgotten about these two Pull Requests:

• https://github.com/datahq/dataflows/pull/124 - WIP for solving the missing values problem attaching it as a row column
• https://github.com/datahq/dataflows/pull/125 - WIP solving the missing values problem attaching it as a schema's field (it seems I forgot to change the readme there)

Both solutions were not good enough for BCO-DMO (still not flexible as it's needed and hard to manage with a list of arbitrary processors). So we decided to go with a global env var for tableschema-py.

If it doesn't feel like a clear improvement for dataflows in general I would close both unmerged