* docker: upgrade ubuntu version
Upgrade rhe ubuntu version used in the main Dockerfile to latest stable version `20.04`
* Update Dockerfile
amended ubuntu version 20 -> 22
but then the following error:
```
0 added, 0 removed; done.
Running hooks in /etc/ca-certificates/update.d...
done.
Processing triggers for dbus (1.12.20-2ubuntu4.1) ...
gpg: error running '/usr/bin/gpg-agent': probably not installed
gpg: failed to start agent '/usr/bin/gpg-agent': Configuration error
gpg: can't connect to the agent: Configuration error
Traceback (most recent call last):
File "/usr/lib/python3/dist-packages/softwareproperties/shortcuthandler.py", line 423, in add_key
PPA publishes dbgsym, you may need to include 'main/debug' component
Repository: 'deb https://ppa.launchpadcontent.net/ubuntugis/ubuntugis-unstable/ubuntu/ jammy main'
Description:
Unstable releases of Ubuntu GIS packages. These releases are more bleeding edge and while generally they should work well, they dont receive the same amount of quality assurance as our stable releases do.
More info: https://launchpad.net/~ubuntugis/+archive/ubuntu/ubuntugis-unstable
Adding repository.
Adding deb entry to /etc/apt/sources.list.d/ubuntugis-ubuntu-ubuntugis-unstable-jammy.list
Adding disabled deb-src entry to /etc/apt/sources.list.d/ubuntugis-ubuntu-ubuntugis-unstable-jammy.list
Adding key to /etc/apt/trusted.gpg.d/ubuntugis-ubuntu-ubuntugis-unstable.gpg with fingerprint 6B827C12C2D425E227EDCA75089EBE08314DF160
subprocess.run(cmd.split(), check=True, input=keys)
File "/usr/lib/python3.10/subprocess.py", line 524, in run
raise CalledProcessError(retcode, process.args,
subprocess.CalledProcessError: Command '['gpg', '-q', '--no-options', '--no-default-keyring', '--batch', '--keyring', '/etc/apt/trusted.gpg.d/ubuntugis-ubuntu-ubuntugis-unstable.gpg', '--homedir', '/tmp/tmpjlr98185', '--import']' returned non-zero exit status 2.
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/usr/bin/add-apt-repository", line 364, in <module>
sys.exit(0 if addaptrepo.main() else 1)
File "/usr/bin/add-apt-repository", line 357, in main
shortcut.add()
File "/usr/lib/python3/dist-packages/softwareproperties/shortcuthandler.py", line 222, in add
self.add_key()
File "/usr/lib/python3/dist-packages/softwareproperties/shortcuthandler.py", line 425, in add_key
raise ShortcutException(e)
softwareproperties.shortcuthandler.ShortcutException: Command '['gpg', '-q', '--no-options', '--no-default-keyring', '--batch', '--keyring', '/etc/apt/trusted.gpg.d/ubuntugis-ubuntu-ubuntugis-unstable.gpg', '--homedir', '/tmp/tmpjlr98185', '--import']' returned non-zero exit status 2.
````
Removing the ubuntugis PPA build without error - is ubuntugis a requirements? probably Jammy has up-to-date packages for pygeoapi dependencies
* Update Dockerfile
removing ubuntugis PPA
* Update Dockerfile
adding `python3-pytest` and `python3-pyld` dependencies
* Update Dockerfile
* Update Dockerfile
* fixing docker environment with suggested changes
* replace ubuntu version with codename
* Update .dockerignore
* Update Dockerfile
* Update Dockerfile
* Update Dockerfile
* Update entrypoint.sh
skip the SensorThings and other tests that require a local server
Co-authored-by: Just van den Broecke <just@justobjects.nl>
* Adding metadata to matrix endpoint
* Adding Tileset metadata with mandatory fields
* Making metadata.json not mandatory, except when it is
* Adding pydantic models
Using models in place of json text
* Adding metadata to matrix endpoint
* Adding Tileset metadata with mandatory fields
* Making metadata.json not mandatory, except when it is
* Adding pydantic models
Using models in place of json text
* Fix tilesets template
Co-authored-by: Francesco Bartoli <francesco.bartoli@geobeyond.it>
* - Added support to read from a generic url template as vector tile backend.
* - Add support for not rendering the tileset metadata
* - added support for z/y/x vector tile layers
* - fixed formatting issues
* - refactored code to use get_layer function, everywhere we need to parse username
- added a couple of debug statements, to make sure we are getting the layer name
correctly
- added an error for url templates that follow a schema, which is not supported yet
* - Added documentation for tiles provider, to show how to read a generic url
* - fixed formatting issues
* - added example of ES vector tiles in the docker-config of elasticsearch
Co-authored-by: doublebyte1 <info@doublebyte.net>
* Json serialize numpy types
This commit fixes an issue with the current xarray provider. When a
NetCDF file is loaded into an xarray dataset, attributes of data
variables that are numbers will be loaded as numpy.float64 or
numpy.int64. These attributes are passed directly to metadata objects
that are json serialized, creating the issue.
This commit thereby fixes#895
* Avoid numpy import in util module
* Add pygeofilter, unpin psycopg2
* Add pygeofilter SQLAlchemy query demo
* Add note about primary key
* Add todos to script
* Make script generic
* Add (commented out) PostgreSQL provider hotosm_bdi_waterways to config
* Update to use local test database
* Return result as dictionary
* Pass GEOM_FIELD as variable
* Add offset and limt, fix two names
* Implement order by
* Use function for cql query
* Add cql query block to query and test
* Link up to db engine
* Convert WKB geometry to GeoJSON
* Add shapely as explicit dependency
* Add tests for CQL queries
* Add test using CROSSES with LINESTRING
* Uncomment Docker PostGIS test layer
* First pass at wiring up API
The CQL is read from the request but the parser fails to convert to
AST.
* Turn logging to DEBUG
Example query
http://localhost:5000/collections/hot_osm_waterways/items?cql=%27osm_id%20BETWEEN%2080800000%20AND%2080900000%20AND%20name%20IS%20NULL%27
* Create dedicated pygeoapi-config.yml for testing PostgreSQL
* Add test for PostgreSQL CQL
* Return 400 for bad CQL
* Tidy up old files
* Bring API error type tests together
* Only reflect the table of interest
* Add pygeofilter to requirements
* Remove local setup_env.sh from .gitignore
* Remove lark exception handling
* Remove lark codes from tests, docstring change
* Split cql test into good and error
* Rename cql parameter to filter
* Add optional filter-lang parameter with test
* Initial changes
* Rename cql text parser
* Use CQL-JSON POST for PostgreSQL query
* Refactor to get data early and handle errors
* Add error tests for CQL-JSON PostgreSQL
* Add tests for prev/next and provider instantiation
* As filter-lang is set the block in not needed
* Clarify test fixtures
* Use filterq only in all CQL calls
* Make engine and table_model attributes of provider
* Handle different instantiation failures
* Use SQLAlchemy for get_fields()
* Minor doc edits
* Use SQLAlchemy for get()
* Simplify _sqlalchemy_to_feature
* Store database parameters in dedicated function
* Strip out non-SQLAlchemy code
* Add draft query() based on CQL filters
* Add property filtering to query()
* Implement select_properties and skip_geometry
* Implement bbox filter
* Implement properties subset defined in config
* Clean up materialized view test
We only need to test the fields that are present to be sure that the
view has been accessed, so other checks have been removed.
* Flake8 fixes and tidy up
* Add combined test for CQL with bbox and properties
* Create sessions directly where required
* Add Engine store to allow connection reuse
* Replace cql_ast with filterq
* Tidy up tests by using fixture for api
* Add authors
* Use Session() as context manager
A session requires a connection to the database, as provided by the
SQLAlchemy engine. When the session goes out of scope, e.g. at exit
from the function when it was used, it should be garbage collected and
the connection freed. However, during load testing it was found that
this wasn't happening. This commit uses a context manager pattern to
start the session, which ensures that the session is closed at the end
of the `with` block.
* Return fields type as {'type': ...}
* Cache the table_model with the engine
Generating the table_model by "reflecting" the database is expensive but
the resulting model doesn't change. If we cache the table_model with
the engine we save on a round-trip to the database and speed up response
time.
* Use separate stores for Engine and table_models
Using a combined store based on both engine and table_model meant that
a pygeoapi serving multiple tables on the same database would maintain
a connection pool for each table. This could eventually saturate the
number of available connections. This commit separates the store into
two parts.
* Update documentation with example
* Manage or post items based on POST content type
* ghcr build action
* only ghcr job
* run criteria edit
* Update README.md
* job = main
* remove success requirememt
* add tag
* tags: bgs-cql
* Update main.yml
* Update README.md
* remove binary
* update
* Update actions to latest versions
* Dockerfile refactor
* Dockerfile refactor
* restore dockerhub
* fix pygeoif version
* # 3.6 test & dockerhub ref
* Pin pygeoif version to less than 1
* Replace psycopg-binary with psycopg
* Split Docker build into sections
* Pin click version >7,<=8
* Add further cql examples
* Update workflows to match GeoPython repo
* update flake python version
* Unquote PostgreSQL column names
* Update container workflow uses versions
* flake8 use python version 3.7
* Use pre ping to get db connection
* Test that properties are set in provider
* Amendments as requested by @tomkralidis
* Correction
* Fix reviewed changes in api code
Co-authored-by: Colin Blackburn <colb@bgs.ac.uk>
Co-authored-by: KoalaGeo <eddlewis85@gmail.com>
Co-authored-by: Edd <edlew@bgs.ac.uk>
Co-authored-by: Colin Blackburn <ximenesuk@users.noreply.github.com>
Just van den Broecke