cqlalchemy 0.0.10

library for constructing cql2-json queries and for generating libraries with multiple STAC extensions"

cqlalchemy

Library to help make CQL2-json queries a little easier!

STAC is a terrific specification for cataloging temporal/spatial data with an emphasis on providing queryable fields for searching that data. One of the ways to make complex queries is to use cql2-json.

This project provides two different functionalities. One is the cqlalchemy.stac.query module which provides query construction class (QueryBuilder) with the most popular extensions (eo, sar, sat, view, mlm).

The other functionality is a script that allows the user to build their own QueryBuilder class from extensions of their choosing, and allowing the opportunity to restrict the fields that can be queried (in the case where it isn't a required field and it's existence in the class might mislead the user).

cqlalchemy QueryBuilder

query by spatial extent

Either a geojson dict or a shapely geometry can be passed

Expand Python Code Sample

import requests
from shapely.geometry import shape
from shapely.validation import make_valid
from cqlalchemy.stac.query import QueryBuilder

planetary_search = "https://planetarycomputer.microsoft.com/api/stac/v1/search"
# request the geojson footprint of King County, Washington
url = "http://raw.githubusercontent.com/johan/world.geo.json/master/countries/USA/WA/King.geo.json"
r = requests.get(url)
geom_dict = r.json()['features'][0]['geometry']
geom = shape(geom_dict)
# fix missing vertices
geom = make_valid(geom)
q = QueryBuilder()
# planetary computer requires defining the constellation
q.collection.equals("landsat-c2-l2")
# define the spatial intersection
q.geometry.intersects(geom)
response = requests.post(planetary_search, q.query_dump_json(limit=2))
for feature in response.json()["features"]:
    print(feature["properties"]["datetime"])
    print(feature["properties"]["eo:cloud_cover"])
    print(feature["geometry"])

query by date

querying using a python date object will query the 24 hour period of that day

Expand Python Code Sample

import requests
from datetime import date
from cqlalchemy.stac.query import QueryBuilder

q = QueryBuilder()
# planetary computer requires defining the constellation
q.collection.equals("landsat-c2-l2")
# search entire utc 24 hour period for December 1st, 2023
q.datetime.equals(date(2023, 12, 1))
planetary_search = "https://planetarycomputer.microsoft.com/api/stac/v1/search"
response = requests.post(planetary_search, q.query_dump_json(limit=2))
for feature in response.json()["features"]:
    print(feature["properties"]["datetime"])

results in

2023-12-01T23:59:27.570403Z
2023-12-01T23:59:03.607352Z

query using an extension

We'll utilize the above query and request data from that date that's less than 30 percent cloud cover by using the Electro-Optical cloud cover field

Expand Python Code Sample

import requests
from datetime import date
from cqlalchemy.stac.query import QueryBuilder

q = QueryBuilder()
# planetary computer requires defining the constellation
q.collection.equals("landsat-c2-l2")
# search entire utc 24 hour period for December 1st, 2023
q.datetime.equals(date(2023, 12, 1))
# either use the lt or lte methods
q.eo.cloud_cover.lt(30)

planetary_search = "https://planetarycomputer.microsoft.com/api/stac/v1/search"
response = requests.post(planetary_search, q.query_dump_json(limit=2))
for feature in response.json()["features"]:
    print(feature["properties"]["datetime"])
    print(feature["properties"]["eo:cloud_cover"])

2023-12-01T23:56:15.912583Z
21.82
2023-12-01T23:54:16.177807Z
28.06

We continue to expand on the above extension utilizing the Landsat extension cloud_cover_land field.

Expand Python Code Sample

import requests
from datetime import date
from cqlalchemy.stac.query import QueryBuilder

q = QueryBuilder()
# planetary computer requires defining the constellation
q.collection.equals("landsat-c2-l2")
# search entire utc 24 hour period for December 1st, 2023
q.datetime.equals(date(2023, 12, 1))
# either use the lt or lte methods
q.eo.cloud_cover.lt(30)

q.landsat.cloud_cover_land.lt(20)

planetary_search = "https://planetarycomputer.microsoft.com/api/stac/v1/search"
response = requests.post(planetary_search, q.query_dump_json(limit=2))
for feature in response.json()["features"]:
    print(feature["properties"]["datetime"])
    print(feature["properties"]["eo:cloud_cover"])
    print(feature["properties"]["landsat:cloud_cover_land"])

The results reveal that some data may not have the cloud_cover_land field defined (this might be that they're not coastal data).

2023-12-01T23:52:40.414555Z
8.28
-1.0
2023-12-01T23:52:16.472683Z
5.39
-1.0

We can try again by forcing our search to be gt -1 and lt 20:

Expand Python Code Sample

import requests
from datetime import date
from cqlalchemy.stac.query import QueryBuilder

q = QueryBuilder()
# planetary computer requires defining the constellation
q.collection.equals("landsat-c2-l2")
# search entire utc 24 hour period for December 1st, 2023
q.datetime.equals(date(2023, 12, 1))
# either use the lt or lte methods
q.eo.cloud_cover.lt(30)

q.landsat.cloud_cover_land.lt(20)
q.landsat.cloud_cover_land.gt(-1)

planetary_search = "https://planetarycomputer.microsoft.com/api/stac/v1/search"
response = requests.post(planetary_search, q.query_dump_json(limit=2))
for feature in response.json()["features"]:
    print(feature["properties"]["datetime"])
    print(feature["properties"]["eo:cloud_cover"])
    print(feature["properties"]["landsat:cloud_cover_land"])
    print(feature["properties"]["platform"])

Now we're getting low land and overall cloud cover values. But it's only landsat-7. We can keep restricting the query by using the q.platform.equals query.

2023-12-01T23:32:54.374649Z
2.0
2.0
landsat-7
2023-12-01T23:32:30.478026Z
0.0
0.0
landsat-7

cqlbuild

The cqlbuild is an interactive cli that allows for creating your own STAC cql2 query class.

Interactive cqlbuild

Add various STAC extensions to the builder. Leave blank to complete adding extensions and move to next step.

Add extension schema by extension name

In some cases the extension schema can be guessed from an extension name. In the below example we use the view extension name:

 % cqlbuild --interactive
Enter extensions, either the path to a local file, a url or the extension json-ld name (sar, sat, etc):
STAC extension, raw schema url, local json extension schema file, local list of extensions or urls : view
treating input view like extension json-ld code and querying https://raw.githubusercontent.com/stac-extensions/view/refs/heads/main/json-schema/schema.json
STAC extension, raw schema url, local json extension schema file, local list of extensions or urls :

Add extension schema with local schema file

 % cqlbuild --interactive
Enter extensions, either the path to a local file, a url or the extension json-ld name (sar, sat, etc):
STAC extension, raw schema url, local json extension schema file, local list of extensions or urls : ./tests/test_data/mlm.schema.json
STAC extension, raw schema url, local json extension schema file, local list of extensions or urls :

Add extension schema by raw schema endpoint

 % cqlbuild --interactive
Enter extensions, either the path to a local file, a url or the extension json-ld name (sar, sat, etc):
STAC extension, raw schema url, local json extension schema file, local list of extensions or urls : https://stac-extensions.github.io/projection/v2.0.0/schema.json
STAC extension, raw schema url, local json extension schema file, local list of extensions or urls :

Omitting fields from the query class interface

Omit fields from the query class interface by adding a field to ignore or a file with a list of fields to ignore.

Enter stac fields to omit from api or a path with a list of fields to omit:
Field to ignore : eo:snow_cover
Field to ignore : created
Field to ignore :

To prevent fields from being queryable through the generated STAC query interface.

cqlbuild from definition file

Below is an example of a definition file for defining what extensions to use and what fields to ignore:

{
  "extensions": [
    "sat",
    "sar",
    "eo",
    "view",
    "landsat",
    "./tests/test_data/mlm.schema.json",
    "https://stac-extensions.github.io/projection/v2.0.0/schema.json"
  ],
  "stac_fields_to_ignore": [
    "view:sun_azimuth",
    "view:sun_elevation",
    "constellation"
  ]
}

It can be used in the cli as follows:

% cqlbuild --definition ./tests/test_data/sample_definition.json --output ./tests/test_data/fixed_up_class.py