V3.0

[msiemens]

This PR prepares an overhaul of TinyDB released as version 3.0.

Notable changes:

• New Query model:
  • where('...').contains('...') has been renamed to where('...').search('...') to indicate that it works on regexes
  • Support for ORM-like usage: User = Query(); db.find(User.name == 'John')
  • where('foo') is an alias for Query().foo
  • where('foo').has('bar') is replaced by either where('foo').bar or Query().foo.bar
    • In case the key is not a valid Python identifier, array notation can be used: where('a.b.c') is now Query()['a.b.c']
  • Checking for the existence of a key has to be done explicitely: where('foo').exists()
• Migrations from v1 to v2 have been removed
• SmartCacheTable has been moved to msiemens/tinydb-smartcache
• Serialization has been moved to msiemens/tinydb-serialization

Todo:

• Revise Query concept (see Refactoring Query objects #62)
• Revise storage concept (see Reading an existing .json file written out using TinyDB #67, Fix exception AttributeError #70)
• Revise Document extension concept (storages, middlewares, custom table classes)
• Move SmartCacheTable to an extension package (depends on Release 1.0 tinydb-smartcache#1)
• Move serialization to an extension package
• List new extensions in docs
• Remove migrations (which were for v1→v2)
• Update the documentation
• Update version number & release date

Closes #62, #67, #70

[msiemens]

I've finally gotten around to finish the missing bits.

@eugene-eeo Could you have a look at this whether we're missing something for v3.0?

[eugene-eeo]

@msiemens you probably want to expose some API for clearing the cache "properly"- at the moment tinyrecord relies on the _cache dictionary (once we have that done perhaps I should look at improving the isolation and correctness of transactions in tinyrecord, perhaps by holding a DB-local lock). I especially like the new Query implementation and how it looks more like an ORM while probably performing better and having less code to maintain than the old model. 👍

I especially like the Query modifiers- so elegant! A little critique but I would probably use a frozenset to represent the __and__ and __or__ "nodes" just so that it is more correct- A AND B is the same as B AND A, and the same applies to A OR B, mathematically.

[msiemens]

you probably want to expose some API for clearing the cache "properly"

Good idea, I've added Table.clear_cache.

A little critique but I would probably use a frozenset to represent the and and or "nodes" just so that it is more correct- A AND B is the same as B AND A, and the same applies to A OR B, mathematically.

Good point! I've updated the AND and OR operations to be commutative.

As always, thanks for your insight, @eugene-eeo!

[msiemens]

TinyDB v3.0.0 is now officially released 🎉