v0.9.0-beta1 - True laziness & CSV support

This version adds true laziness (#14) and CSV/TSV file database support (#11).

This is the first beta version of the new v0.9.0, that has a lot of refactoring behind the scenes, but keeps the same public API.

True laziness

To use the true lazy lazy-ids and lazy-docs you'll have to either delete your pre-v0.9.0 .nddbmeta files, or upgrade them.
You may prefer to upgrade if the databases they represent are really big (because then indexing might take a while).

To upgrade the .nddbmeta files, you simply call nd-db.convert/upgrade-nddbmeta! from a repl. The function takes a db value parameter and takes care of the rest.

CSV Support

CSV databases are just as simple to use, and takes up less space than the other data formats, because they don't replicate the keys for every document.

You need an additional parameter to create a database value based on CSV (or TSV): :col-separator:

(nd-db.core/db :filename "some-data.csv" :col-separator ";" :id-path :id)

Parsing defaults to a parser that simply parses column data as numbers or strings - nothing else. But you can pass your own column parser like this:

(nd-db.core/db :filename "some-data.csv" :col-separator ";" :id-path :id :col-parser my-col-parser-fn)

Refer to unit tests for more info.

To see what's needed before the final release of v0.9.0, check out this pull request: #15

v0.9.0-alpha3

The new .nddbmeta format for version v0.9.0 and forward, will not only make a lazy seq of documents available. Now even the index can be lazily read.

Also the indexes are now generated in parallel - meaning 2/3 faster than before (on a mbp m1 pro).

Did lots of refactoring so far - more is needed before the final release.

Refer to #14 and #11 for status on v0.9.0.

Introducing lazy-docs

This non-breaking release introduces nd-db.core/lazy-docs - a lazy representation of the database contents.

Maintenance update

Updated nippy

v0.7.1 - bugfix release

Downgraded nippy.
Projects using nd-db 0.7.0 couldn't compile nd-db with nippy version 3.2.0.

v0.7.0 - portable databases - and more

This releases makes it possible to move the database (and serialized index) to another folder, without the need for generating a new index (which might be a bit time consuming for huge databases of multiple GBs).

Furthermore you can now use the new optional keyword parameter syntax, as well as the old well known map parameter.

See the README and unit tests for examples.

Changes:

• Make serialized databases portable (not bound to a specific filesystem path)
• Support for clojure 1.11 keyword function parameters: https://clojure.org/news/2021/03/18/apis-serving-people-and-programs
• Update dependencies

v0.6.3 - upgrade to Clojure 1.11

• Upgrade Clojure 1.10.3 -> 1.11.1
• Upgraded other dependencies
• Minor optimizations

v0.6.2 - fix issue w/indexing

Fix issue when creating index for ndjson/ndedn

v0.6.1 - avoid a reflective call

Eliminate a reflective call when serializing the database.

v0.6.0 - introducing .ndnippy!

Now you can use .ndnippy as database format. It's MUCH faster to load than
.ndjson and .ndedn, meaning better query times. Particularly when querying multiple documents at once.

Also a new util namespace lets you convert from .ndjson and .ndedn to .ndnippy.

.ndnippy - like .ndedn isn't really a standard. But it probably should be. I implemented the encoding for
.ndnippy myself, it's somewhat naive, but really fast anyhow. If you have ideas on how to make it even
fast, let me know. Because version 0.6.0 introduces the .ndnippy format, it may change several times in the
future, possibly making old .ndnippy files incompatible with new versions. Now you're warned. Thankfully the
generation of new .ndnippy files is quite fast.

NB: .ndnippy isn't widely used, and probably isn't a good format for distribution, unless you can distribute
this nd-db library with it!