Kivi.py 🥝

from kivi import Kivi

db = Kivi()
db.kv_create("mydb", {"greeting-1": "Good morning 🌅"})

A simple key-value database that uses JSON to store data

Supported data types

• String
• Boolean
• Integer
• Float
• Lists
• Dict
• Datetime

Why?

I just wanted a simple database to play around, why everything is so complex

Install

pip install kivi.py

Usage

• Init
• Create
• Load
• Get
• Set / Update
• Merge
• Index
• Query
• Search

---

Initialize database

from kivi import Kivi

db = Kivi(
    src="Where the database files should be stored. Defaults to `os.getcwd()/.kivi_db`",
    to_load=["path/to/file.json"])

Note

to_load only takes json file with default kivi format. If you want to store regular json data use create or merge

---

Create

Creates a new database and load it to the memory

• Arguments:

  • name: Name of the database
  • data: Dict consist of data

• Example:

  db.kv_create("wow", {"user-1": 123456})

---

Load

Note

It is recommended to load databases when initializing the database with Kivi(to_load=[...])

Loads database into the memory

• Arguments:

  • data: Path to a json file or python dict

• Returns: Index of the database

• Example:

  db.kv_load("/home/test/data/stuff.json")

---

Get

Get value from a database

• Arguments:

  • index: Index of the database (returned in kv_load or when creating the instance)
  • key: Key!

• Example:

  db.kv_get(0, "greeting")

---

Set / Update

Add or update key in a database

• Arguments:

  • index: Index of the database (returned in kv_load or when creating the instance)
  • key: Key!

• Example:

  db.kv_set(0, "gtr1", "greeting")

---

Merge

Merge data into the database

• Arguments:

  • index: Index of the database (returned in kv_load or when creating the instance)
  • tomerge: Dict consist of data that needs to be merged

• Example:

  dt = {
      "thing-1": [123, True, {'user1': 'gibberish?#45'}],
      "thing-2": [123, True, {'user2': 'gibberish?#45', 'nuh': True}]
  }
  db.kv_merge(index=0, tomerge=dt, to_std=True)

Note

If the tomerge dict is not in the kivi formatting, set to_std = True

---

Index

Index data to perform queries much faster

• Arguments:

  • data: Data to index (a dict or list of dicts)
  • fields: Json fields that needs to be indexed
  • chars: Amount of characters that indexed key can have (Ex: "Spider" will be sliced to "spi" to create the key for index)

• Example:

  data = [
      {"title": "Spooder man", "extract": "Marvel spooder man!"},
      {"title": "Iron deficiency man", "extract": "Marvel no Fe?"}
  ]
  db.kv_index(data, ["title", "extract"], 3)

---

Query

Search for string in an indexed database

• Arguments:

  • index: Index of the database (returned in kv_load or when creating the instance)
  • query: String to search for in the database

• Example:

  db.kv_query(0, "Spider man")

---

Search

Search for string in a database

• Arguments:

  • index: Index of the database (returned in kv_load or when creating the instance)
  • query: String to search for in the database
  • strict: Pass 'False' to get more results

• Example:

  db.kv_search(0, "Spider man")

How it works

Data is stored inside json files in the following format

{
    "name": "lot",
    "items": {...}
}

items contains the inserted data as a string alongside it's original data type like this,

{
    "name": "lot",
    "items": {
        "thing-1": [
            "{'user': 'gibberish?#45'}",
            "dict"
        ],
        "thing-2": [
            "{'user': 'gibberish?#45'}",
            "dict"
        ]
    }
}

When you request data using kv_get function, data will be automatically converted back to it's original type (limitations)

Requirements

• Python 3.5+
  • dependecies

Limitations

The data convertion has lots of limits as of now (I don't plan of updating this either).

For example:

• Complex json structures will fail to convert