NocoDB Python Client

NocoDB is a great Airtable alternative. This client allows python developers to use NocoDB API in a simple way.

• Contributors guidelines

Installation

pip install nocodb

Usage

Client configuration

from nocodb.nocodb import NocoDBProject, APIToken, JWTAuthToken
from nocodb.filters import LikeFilter, EqFilter, And
from nocodb.infra.requests_client import NocoDBRequestsClient


# Usage with API Token
client = NocoDBRequestsClient(
        # Your API Token retrieved from NocoDB conf
        APIToken("YOUR-API-TOKEN"),
        # Your nocodb root path
        "http://localhost:8080"
)

# Usage with JWT Token
client = NocoDBRequestsClient(
        # Your API Token retrieved from NocoDB conf
        JWTAuthToken("your.jwt.token"),
        # Your nocodb root path
        "http://localhost:8080"
)

Project creation

# Example with default database
project_body = {"title": "My new project"}

# Example with Postgresql
project_body = {
    "title": "MyProject",
    "bases": [
        {
            "type": "pg",
            "config": {
                "client": "pg",
                "connection": {
                    "host": "localhost",
                    "port": "5432",
                    "user": "postgres",
                    "password": "postgres",
                    "database": "postgres"
                },
                "searchPath": [
                    "public"
                ]
            },
            "inflection_column": "camelize",
            "inflection_table": "camelize"
        }
    ],
    "external": True
}

project = client.project_create(body=project_body)

Project selection

# Be very carefull with org, project_name and table names
# weird errors from nocodb can arrive if they are wrong
# example: id is not defined...
# probably they will fix that in a future release.
project = NocoDBProject(
        "noco", # org name. noco by default
        "myproject" # project name. Case sensitive!!
)

Table rows operations

table_name = "tablename"

# Retrieve a page of rows from a table
table_rows = client.table_row_list(project, table_name)

# Retrieve the first 1000 rows
table_rows = client.table_row_list(project, table_name, params={'limit': 1000})

# Skip 100 rows
table_rows = client.table_row_list(project, table_name, params={'offset': 100})

⚠️ Seems that we can't retrieve more than 1000 rows at the same time but we can paginate to retrieve all the rows from a table

Pagination example

first_100_rows = client.table_row_list(project, table_name, params={'limit': 100})
next_100_rows = client.table_row_list(project, table_name, params={'limit': 100, 'offset': 100})
next_100_rows = client.table_row_list(project, table_name, params={'limit': 100, 'offset': 200})

More row operations

# Filter the query
table_rows = client.table_row_list(project, table_name, LikeFilter("name", "%sam%"))
table_rows = client.table_row_list(project, table_name, And(LikeFilter("name", "%sam%"), EqFilter("age", 26)))
table_rows = client.table_row_list(project, table_name, filter_obj=EqFilter("Id", 100))

# Filter and count rows
count = client.table_count(project, table_name, filter_obj=EqFilter("Id", 100))

# Find one row
table_row = client.table_find_one(project, table_name, filter_obj=EqFilter("Id", 100), params={"sort": "-created_at"})

# Retrieve a single row
row_id = 10
row = client.table_row_detail(project, table_name, row_id)

# Create a new row
row_info = {
    "name": "my thoughts",
    "content": "i'm going to buy samuel a beer 🍻 because I 💚 this module",
    "mood": ":)"
}
client.table_row_create(project, table_name, row_info)

# Update a row
row_id = 2
row_info = {
    "content": "i'm going to buy samuel a new car 🚙 because I 💚 this module",
}
client.table_row_update(project, table_name, row_id, row_info)

# Delete a row (only if you've already bought me a beer)
client.table_row_delete(project, table_name, row_id)

Available filters

• EqFilter
• EqualFilter (Alias of EqFilter)
• NotEqualFilter
• GreaterThanFilter
• GreaterOrEqualFilter
• LessThanFilter
• LessOrEqualFilter
• LikeFilter
• Or
• Not
• And

Combining filters using Logical operations

from nocodb import filters

# Basic filters...
nick_filter = filters.EqFilter("nickname", "elchicodepython")
country_filter = filters.EqFilter("country", "es")
girlfriend_code = filters.EqFilter("gfcode", "404")
current_mood_code = filters.EqFilter("moodcode", "418")

# Combining filters using logical filters
or_filter = filters.Or(nick_filter, country_filter)
and_filter = filters.And(girlfriend_code, current_mood_code)

# Negating filters with a Not filter
not_me = filters.Not(filters.EqFilter("nickname", "elchicodepython"))

# You can also combine combinations
or_combined_filter = filters.Or(or_filter, and_filter)
and_combined_filter = filters.And(or_filter, and_filter)

Using custom filters

Nocodb is evolving and new operators are coming with each release.

Most of the basic operations are inside this package but you could need some new feature that could not be added yet. For those filters you can build your own.

Example for basic filters:

from nocodb.filters.factory import basic_filter_class_factory

BasicFilter = basic_filter_class_factory('=')
table_rows = client.table_row_list(project, table_name, BasicFilter('age', '16'))

You can find the updated list of all the available nocodb operators here.

In some cases you might want to write your own filter string as described in the previous link. For that cases you can use the less-semmantic RawFilter.

from nocodb.filters.raw_filter import RawFilter

table_rows = client.table_row_list(project, table_name, RawFilter('(birthday,eq,exactDate,2023-06-01)'))

In some cases we might want to have a file with some custom raw filters already defined by us. We can easily create custom raw filter classes using raw_template_filter_class_factory.

from nocodb.filters.factory import raw_template_filter_class_factory

BirthdayDateFilter = raw_template_filter_class_factory('(birthday,eq,exactDate,{})')
ExactDateEqFilter = raw_template_filter_class_factory('({},eq,exactDate,{})')
ExactDateOpFilter = raw_template_filter_class_factory('({},{op},exactDate,{})')

table_rows = client.table_row_list(project, table_name, BirthdayDateFilter('2023-06-01'))
table_rows = client.table_row_list(project, table_name, ExactDateEqFilter('column', '2023-06-01'))
table_rows = client.table_row_list(project, table_name, ExactDateOpFilter('column', '2023-06-01', op='eq'))

Credits to @MitPitt for asking this feature.

Author notes

I created this package to bootstrap some personal projects and I hope it will help other developers from the python community. It's not completed but it has what I needed: A full CRUD with some filters.

Feel free to add new capabilities by creating a new MR.

Contributors

• Samuel López Saura @elchicodepython
• Ilya Sapunov @davert0
• Delena Malan @delenamalan
• Jan Scheiper @jangxx