Skip to main content

Query proxy that allows the usage of AtlasSearch with mongoengine specific syntax

Project description

AtlasQ

AtlasQ allows the usage of AtlasSearch keeping the MongoEngine syntax.

Structure

The package tries to follow the MongoEngine structure; the major differences reside in the transform.py and queryset.py files.

Transform

Like in MongoEngine, a step in the pipeline is the creation of a query from a Q object: we have to find a correspondence between the MongoEngine common syntax and what AtlasSearch allows. For doing this, we had to find some compromises.

Not every keyword is supported at the moment: if you have an actual use case that you would like to support, please be free to open an issue or a PR at any moment.

QuerySet

There are probably a thousand of better implementation, if you actually knew MongoEngine and above all PyMongo. Unfortunately, our knowledge is limited, so here we go. If you find a solution that works better, again, feel free to open an issue or a PR.

The main idea, is that the filter should work like an aggregation. For doing so, and with keeping the compatibility on how MongoEngine works (i.e. the filter should return a queryset of Document) we had to do some work.
Calling .aggregate instead has to work as MongoEngine expect, meaning a list of dictionaries.

Usage

Now the most important part: how do you use this package?

from mongoengine import Document, fields

from atlasq import AtlasManager, AtlasQ, AtlasQuerySet

index_name = str("my_index")

class MyDocument(Document):
    name = fields.StringField(required=True)
    surname = fields.StringField(required=True)
    atlas = AtlasManager(index_name)

obj = MyDocument.objects.create(name="value", surname="value2")

qs = MyDocument.atlas.filter(name="value")
assert isinstance(qs, AtlasQuerySet)
obj_from_atlas = qs.first()
assert obj == obj_from_atlas

obj2_from_atlas = MyDocument.atlas.get(AtlasQ(name="value") & AtlasQ(surname="value2"))
assert obj == obj2_from_atlas


obj3_from_atlas = MyDocument.atlas.get(AtlasQ(wrong_field="value"))
assert obj3_from_atlas is None

Extended Features

Validation

We also decided to have, optionally, a validation of the index. Two things are checked:

  • The index actually exists (If you query a non-existing index, Atlas as default behaviour will not raise any error).
  • The fields that you are querying are actually indexed(If you query a field that is not indexed, Atlas as default behaviour will not raise any error, and will return an empty list). To make these check, you need to call the function ensure_index on the queryset.
from mongoengine import Document, fields

from atlasq import AtlasManager, AtlasQ

index_name = str("my_index")

class MyDocument(Document):
    name = fields.StringField(required=True)
    surname = fields.StringField(required=True)
    atlas = AtlasManager(index_name)

result = MyDocument.atlas.ensure_index("user", "pwd", "group", "cluster")
assert result is True
obj1_from_atlas = MyDocument.atlas.get(AtlasQ(name="value")) 
obj2_from_atlas = MyDocument.atlas.get(AtlasQ(wrong_field="value")) # raises AtlasIndexFieldError

EmbeddedDocuments

Embedded documents are queried in two different ways, depending on how you created your Search Index. Remember to ensure the index so that AtlasQ can know how your index is defined If you used the embeddedDocuments type, AtlasQ will use the embeddedDocument query syntax. Otherwise, if you used the document type, or you did not ensure the index, a normal text search with the . syntax will be used.

Upload index

It is possible to upload directly the Search index using AtlasQ, calling the function upload_index on the queryset. Syntax checks on the index itself are performed. If the _id is not present but pk or id was specified, it will be automatically added, allowing valid text query on the primary key.

from mongoengine import Document, fields

from atlasq import AtlasManager

index_name = str("my_index")
index = {
  "analyzer": "lucene.keyword",
  "mappings": {
    "dynamic": False,
    "fields": {
      "_id": {
        "type": "objectId"
      },
        "name": {
        "type": "string"
      },
      "surname": {
        "type": "string"
      },
    }
  }
}
class MyDocument(Document):
    name = fields.StringField(required=True)
    surname = fields.StringField(required=True)
    atlas = AtlasManager(index_name)

result = MyDocument.atlas.ensure_index("user", "pwd", "group", "cluster")
assert result is False
MyDocument.atlas.upload_index(index, "user", "pwd", "group", "cluster")
result = MyDocument.atlas.ensure_index("user", "pwd", "group", "cluster")
assert result is True

Sort

On the 10th of July 2023, the Sort functionality was released for Atlas search.

AtlasQ, from version 0.12.0, will support this feature inside the order_by function. To have the old behaviour of the order_by (useful if you want to sort after aggregations and not after the search stage), you can set the kwarg as_aggregation as True.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

atlasq-0.13.2.tar.gz (38.5 kB view details)

Uploaded Source

Built Distribution

atlasq-0.13.2-py3-none-any.whl (15.7 kB view details)

Uploaded Python 3

File details

Details for the file atlasq-0.13.2.tar.gz.

File metadata

  • Download URL: atlasq-0.13.2.tar.gz
  • Upload date:
  • Size: 38.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.11.5

File hashes

Hashes for atlasq-0.13.2.tar.gz
Algorithm Hash digest
SHA256 7bfcd61267d641a90e717a895a48ea1d9aa67dedc092c38de005461d2e222042
MD5 4b463b0db1035cb98f21c40fd71871d8
BLAKE2b-256 f272c9d71fb75e0a7a195d86d2cd327b605d47f585a71acb85ba3d385d2f3636

See more details on using hashes here.

File details

Details for the file atlasq-0.13.2-py3-none-any.whl.

File metadata

  • Download URL: atlasq-0.13.2-py3-none-any.whl
  • Upload date:
  • Size: 15.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.11.5

File hashes

Hashes for atlasq-0.13.2-py3-none-any.whl
Algorithm Hash digest
SHA256 b08645555bf2db333a6acaebd24f3315f3806622a62284aec1d66a19234e415f
MD5 6d4288b72b2585ec4f0015f0111e5ada
BLAKE2b-256 2bcc23791bfba509c42ed158d8187eaaee82e818447640af8ff945535d70f02f

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page