shared VS insta-data-styled

Hi HN !

Before discovering JSON, I worked on projects where I had to design relational databases [1] and write SQL queries. I understood the difference between a relational database and a file format, but couldn't resist the temptation to adopt JSON to store data locally.

Ever since I adopted JSON to store data locally, I wanted a class with an intuitive interface to play with JSON files. Eventually, the class evolved into a library.

I had to make a graphical SQL editor to test a GUI framework, so I created the LiteDao package (Lite Data Access object) which offers an intuitive interface to connect to a database with SQLite. The project worked as I wanted it to.

So, I had two different packages to play with data. I wanted to unify these two packages into a single class. But I failed to do so.

I came back to the challenge, and this time I had the idea of a single package with three classes with almost same intuitive interface that meet three needs: the need to expose a configuration file so that the user can modify values, the need to store collections and binary data without worrying about how they are actually saved, and the need to work with a relational database. These classes are respectively named Document, Dossier, and Database.

I managed to implement it in Python and added a command line interface [2] that allows different applications to exchange collections and binary data.

I would like to know what you think [3] of this project which is still in beta and which I already use in several other projects which are also in beta.

[1] https://www3.ntu.edu.sg/home/ehchua/programming/sql/Relation...

[2] https://github.com/pyrustic/shared#command-line-interface

[3] https://news.ycombinator.com/item?id=12886522 (Crocker's rules)

[2] https://github.com/pyrustic/shared

Hi HN !

A few days ago I wrote Shared [1] a library to store, expose, read and edit data collections. Data collections are stored on the hard drive in separate files. I needed a mechanism to know when the content of a data collection has changed, so only the updated data collection will be stored. Thus came to me the idea of writing Probed.

This is a work in progress, I will write some documentation and tests later. Tell me what you think [2].

[1] https://github.com/pyrustic/shared

[2] http://sl4.org/crocker.html