Keep track of esprima.js and escodegen.js

[NOTE]: # ( ^^ Provide a general summary of the request in the title above. ^^ )

## Summary

[NOTE]: # ( Provide a brief overview of what the new feature is all about. )
esprima.js and escodegen.js are introduced to support spidermonkey 6.0 porting in CouchDB, especially supporting the traditional CouchDB-style function declaration.
```
"map": "function(doc) {emit(doc._id, doc._id);}"
```
They are being maintained under [1] https://github.com/jquery/esprima and [2] https://github.com/estools/escodegen

In planned CouchDB 3.0, we are using the variant from [3] https://github.com/dmunch/couch-chakra.

## Desired Behaviour

[NOTE]: # ( Tell us how the new feature should work. Be specific. )
[TIP]: # ( Do NOT give us access or passwords to your actual CouchDB! )

We need to better keep track of evolution of esprima and escodegen.

## Possible Solution

[NOTE]: # ( Not required. Suggest how to implement the addition or change. )
- [x] create couch-esprima and couch-escodegen repositories so that there are safe copy
- [ ] compare the difference between [1] [2] and [3]
- [ ] find tools/scripts build workable version from couch-esprima and couch-escodegen
- [ ] figure out the way how to apply patch for CouchDB 3.0
- [ ] define and implement how to continuously apply fixes and improvement from [1] and [2] for CouchDB.

## Additional context

[TIP]: # ( Why does this feature matter to you? What unique circumstances do you have? )

@janl @garrensmith @davisp @wohali

Start by comparing the upstream esprima and escodegen repositories with the couch-chakra variants linked in the issue. Define a reproducible way to build workable CouchDB versions, apply patches, and continuously track upstream fixes; done means those update and patch steps are documented and repeatable.

Written by the indexing model from the issue text.