udata-hydra-csvapi

This connects to udata-hydra and serves the converted CSVs as an API.

Run locally

Start udata-hydra via docker compose.

Launch this project:

docker compose up

You can now access the raw postgrest API on http://localhost:8080.

Now you can launch the proxy (ie the app):

poetry install
poetry run adev runserver -p8005 udata_hydra_csvapi/app.py

And query postgrest via the proxy using a resource_id, cf below. Test resource_id is 27d469ff-9908-4b7e-a2e0-9439bb38a395

API

Meta informations on resource

curl http://localhost:8005/api/resources/27d469ff-9908-4b7e-a2e0-9439bb38a395/

{
  "created_at": "2023-02-11T11:44:03.875615+00:00",
  "url": "https://data.toulouse-metropole.fr//explore/dataset/boulodromes/download?format=csv&timezone=Europe/Berlin&use_labels_for_header=false",
  "links": [
    {
      "href": "/api/resources/27d469ff-9908-4b7e-a2e0-9439bb38a395/profile/",
      "type": "GET",
      "rel": "profile"
    },
    {
      "href": "/api/resources/27d469ff-9908-4b7e-a2e0-9439bb38a395/data/",
      "type": "GET",
      "rel": "data"
    }
  ]
}

Profile (csv-detective output) for a resource

curl http://localhost:8005/api/resources/27d469ff-9908-4b7e-a2e0-9439bb38a395/profile/

{
  "profile": {
    "header": [
        "geo_point_2d",
        "geo_shape",
        "ins_nom",
        "ins_complexe_nom_cplmt",
        "ins_codepostal",
        "secteur",
        "..."
    ]
  },
  "...": "..."
}

Data for a resource (ie resource API)

curl http://localhost:8005/api/resources/27d469ff-9908-4b7e-a2e0-9439bb38a395/data/?limit=1

[
  {
    "__id": 1,
    "geo_point_2d": "43.58061543292057,1.401751073689455",
    "geo_shape": {
      "coordinates": [
        [
          1.401751073689455,
          43.58061543292057
        ]
      ],
      "type": "MultiPoint"
    },
    "ins_nom": "BOULODROME LOU BOSC",
    "ins_complexe_nom_cplmt": "COMPLEXE SPORTIF DU MIRAIL",
    "ins_codepostal": 31100,
    "secteur": "Toulouse Ouest",
    "quartier": 6.3,
    "acces_libre": null,
    "ins_nb_equ": 1,
    "ins_detail_equ": "",
    "ins_complexe_nom": "",
    "ins_adresse": "",
    "ins_commune": "",
    "acces_public_horaires": null,
    "acces_club_scol": null,
    "ins_nom_cplmt": "",
    "ins_id_install": ""
  }
]

On this endpoint you can use every neat stuff postgrest provides. Here we only want the ins_nom column where it icontains "maurice":

curl "http://localhost:8005/api/resources/27d469ff-9908-4b7e-a2e0-9439bb38a395/data/?select=ins_nom&ins_nom=ilike.*maurice*"

[
  {
    "ins_nom": "BOULODROME MAURICE BECANNE"
  }
]