This is a pymongo helper package, that let you make more complicated actions with your collections. It gives you the possibility to populate between packages without needs of creating an aggregation by yourself. Gives you the basic actions such as find, update and delete, with more simplicity.
Pymongoose came with a Schema class that let you work with more efficiency and organization
pip install pymongoose pymongo
models/role.py
import datetime
from pymongoose import methods
from pymongoose.mongo_types import Types, Schema, MongoException, MongoError
from bson import json_util
from bson.objectid import ObjectId
class Role(Schema):
schema_name = "roles" # Name of the schema that mongo uses
# Attributes
id = None
name = None
action = None
def __init__(self, **kwargs):
self.schema = {
"name": {
"type": Types.String,
"required": True
},
"actions": [{
"type": Types.String,
"required": True
}]
}
super().__init__(self.schema_name, self.schema, kwargs)
def __str__(self):
return f"Role: {self.name}, Actions: {self.action}"
Is higly recommended to follow this model for any schema
This mongo configuration is required for pymongoose to work correctly
db.py:
import os, traceback
import signal
from pymongo import MongoClient
from pymongoose.methods import set_schemas, get_cursor_length
from models.role import Role
MONGO_URI = os.environ.get("MONGO_URI")
mongo_db = None
def mongo_init ():
global mongo_db
client = MongoClient(MONGO_URI)
db = client.test
try:
# Now schemas can be set in two ways
#-> In a dict mode
schemas = {
"roles": Role(empty=True).schema
}
set_schemas(db, schemas)
#-> Or in a list mode
set_schemas_from_list (db, [Role (empty=True)])
print("MongoDB Connected!")
except:
traceback.print_exc()
print("Error initializing database")
exit(1)
if __name__ == "__main__":
mongo_init()
- For more model examples check examples/
user = User(
name="TestA",
password="test",
role=role.id
)
id = user.save()
user.save() will throw an exception if a field marked as required in schema is None.
users = User.find({})
for user in users:
user = User.parse(user)
This will return a cursor of elements, which can be parsed into User model for a better management.
user = User.find({}, one=True)
# OR
user = User.find_one ({})
This will return a User element with fields obtained by database, if you prefer a json just add
parse = False
At last of the find statement.
# this Function will search for "_id" field
# so id must be a hex24 string or ObjectId
user = User.find_by_id(id)
# for a complete user
user = User.find_by_id(id, select={"name": 1, "username": 1})
# For a custom user
This will return a User element with fields obtained by database, if you prefer a json just add
parse = False
At last of the find statement.
user = User.find_by_id(id, populate=[
{
"path": "role",
"select": ["name", "actions"],
"options": ["actions"]
}
])
# For an extensive populate
# or
user = User.find_by_id(id, populate=["role"])
# For a simple populate
This will return a User element with fields obtained by database
Populate is a really useful tool when we work with difficult or complex models, with pymongoose you wont need to create an extensive aggregation to lookup for elements.
Populate works with simple ids like:
schema = {
"name": {
"type": Types.String,
"required": True
},
"role": { # <- Simple field
"type": Types.ObjectId,
"ref": "roles"
}
}
lists:
schema = {
"name": {
"type": Types.String,
"required": True
},
"logs": [{ # <- List field
"type": Types.ObjectId,
"ref": "logs"
}]
}
and complex models
schema = {
"name": {
"type": Types.String,
"required": True
},
"friends": [{
"friend": {
"type": Types.ObjectId, # <- Complex list field
"ref": "friends"
},
"frequent": {
"type": Types.Boolean,
"default": False
}
}]
}
Pymongoose can recursively populate fields like this:
#menu schema:
menu_schema = {
# ...
"items":[
{
"type":Types.ObjectId,
"ref":"items"
}
],
# ...
}
#item schema:
item_schema: {
# ...
"drink":{
"type":Types.ObjectId,
"ref":"drinks"
},
"dish":{
"type":Types.ObjectId,
"ref":"dishes"
}
# ...
}
#Populate
menus = Menu.find({}, populate=[{
"path": "items",
"options": ["dish", "drink"]
}])
for menu in menus:
print(menu)
This will return a CursorCommand element
In this example all menus populate each item of their array, at same time element dish and drink are populated, returning a complete array of populated menus with populated items.
pymongoose has the ability to parse the return cursor as different types: Available types are - IF AS_DEFAULT(0), will return cursor if one == False, else a dict if parse == False - IF AS_DICT(1), will return a list of dicts if one == False, else as serializable dict - IF AS_STRING(2), will return a parsed list of dicts as str if one == False, else as a str parsed serializable dict
*Note: In case you don't need a serializable dict, left cursor AS_DEFAULT
count = User.update({},
{
"$set": {
"password": "0"
}
}, many=True
) #many is set to False by default
count = User.delete({}, many=True) #many is set to False by default
If you are working with vscode I have already created a model snippet to save you a lots of time: https://github.com/Djcharles26/pymongoose/blob/master/pymongoose-snippets.json
For using this you must copy them to your python snippets:
- CTRL + SHIFT + P
- Configure User snippets
- python.json
- Copy snippets in blob
- Paste them in your user python.json