point-server

point-server is a server use to extract pointcloud data.

install

Assume that you have installed in your system :

• Anaconda
• NodeJs

conda env create -f environment.yml
npm install

run

Create a .env file copied from .example.env with your specific secret values. Then start the server.

conda activate point-server
npm start

test

Unit test, in local :

./script/test/local_test.sh

configuration file

You must set environment varible CONFIG_FILE; pointing to a configuration file.

This JSON example show a config file with 2 possible usages :

{
    "local": {
        "EPT_JSON": "./data/micro/EPT_4978/ept.json",
        "PIVOT_THREEJS": "./data/micro/metadata/pivotTHREE.json",
        "SURFACE_MAX": "50000",
        "RETURN_URL": false
    }
    "configWithS3": {
        "EPT_JSON": "ept://http://url/to/ept.json",
        "PIVOT_THREEJS": "http://url/to/pivotTHREE.json",
        "SURFACE_MAX": 100000,
        "RETURN_URL": true,
        "S3_BUCKET": "lidarhd",
        "S3_TMP_FOLDER": "tmp",
        "REDIRECT_URL": "http://lidarhd.pocgpf.ovh/tmp"
    }
}

EPT_JSON : PointCloud store in EPT format (Entwine)

PIVOT_THREEJS : Special file used to reproject the pointcloud. It's design to be used with a pointcload created by PointTools project : https://github.com/iTowns/PointsTools

SURFACE_MAX : Max area to be extracted in square meter. Bigger area will lead to error 400 ( with id BAD_REQUEST_AREA). 0 means no limit.

RETURN_URL : if true : use S3 store to store the file created, and and send redirection as response. Also, when the request comes, send 202 to tell user we accept the request, so the user will have to request the API again to have the responses.

if false : Send the file in the response. The request hangs while the file is being created.

If RETURN_URL is set to true, you must set these 3 other config elements.

S3_BUCKET: S3 Bucket used to store created files.

S3_TMP_FOLDER: folder in the S3 store used to store created files.

REDIRECT_URL: Visible URL used to redirect the response

use S3 store

You must set environment variables :

AWS_SECRET_ACCESS_KEY : secret access key

AWS_ACCESS_KEY_ID : access key id

AWS_ENDPOINT : endpoint (default value: s3.gra.cloud.ovh.net)

http_proxy : If you are behind a proxy (ex: http://proxy.example.com:3128)

run inside screen (terminal multiplexer)

Start new screen session

screen -S point_server_running

Then run the server

conda activate point-server
npm run dev

Detach the session, so you can exit the terminal, and the server is still running.

Ctr + a, d

Verify your session is alive, listing all sessions with :

screen -ls

Re attach your terminal to a session

screen -r point_server_running

request on the server

Download a pointcloud, with curl :

curl -o output.las localhost:3000/points\?source=micro\&poly\=791940.71_6272225.14,791941.60_6272227.37,791935.63_6272228.62,791935.72_6272225.76

And open it with cloudCompare :

cloudcompare.CloudCompare output.las

source refers to one of the config block in the configuration file

poly is the polygon to extract; defined like this : x1_y1,x2_y2,x3_y3