This scripts works with Linux

You need to adapt .sh scripts to works out of Linix OS

ggl-places

What ?

This repo contain some scripts to update database market hours from google place ID.

Prerequisite

We needs :

• market datasource (with google place id field at least) from geoserver WFS
• google place /detail API KEY to request hours for each market
• PostGreSQL/PostGIS and user access (will be use to create and write a table)

Some packages will be installed by install.sh script :

https://github.com/jdev-org/ggl-places/blob/main/script/python/install.sh

Install

If the process use a root or sudo user, you just have to go to the next section because install process is includes in start.sh file (need super user right to install dependancies).

If you use a non superuser to trigger start.sh file, you have to use install.sh first to install dependancies and python virtual env :

https://github.com/jdev-org/ggl-places/blob/main/script/python/install.sh

Commands :

• cd script/python
• sudo bash install.sh

Go to the next section to start the process.

Process

First, open start.sh and change environment variables values :

ggl-places/start.sh

Lines 13 to 22 in 334ca15

	export GGL_API_KEY=""
	export EDP_GEOSERVER_USER=""
	export EDP_GEOSERVER_PASSWORD=""
	export EDP_GEOSERVER_URL=""
	export EDP_COMMERCE_LAYER=""
	export EDP_DB_NAME=""
	export EDP_DB_HOST=""
	export EDP_DB_USER=""
	export EDP_DB_PASSWORD=""

Next, you can start the process with this (Linux) command :

[CREATE_FILE] : <boolean> - true to force places file generation, false to use current and don't call API [INSTALL] : <boolean> - true to install (need superuser) or not (if already done or trigger without superuser)

. ./start.sh [CREATE_FILE] [INSTALL]

Process important information

• You will find a .venv in script/python/ directory (python3)

• Process will automatically save a new place_hours.csv into root directory

• Process will automatically create (or overwrite) a PostGIS table (name is same as CSV file)

• Process will create a new ./places.json file to allow to read API response if you don't need to request all every time

• Process will create a ./places.json.error.json file to list feature without API ID and invalid API IDs

• API URI is located in ./script/python/const.py

• Var env GEOSERVER_USER and GEOSERVER_PASSWORD are not required

And now ?

Publish this new table with your favorite GIS server (e.g GeoServer)

Data model explainations

• Hours fields (close/open) are show in minutes to ease comparison and conversion : 12h30 = 12h*60min +30 = 750 min

• A day is identify by a number :

Monday	1
wednesday	3
Tuesday	2
...	...

• For a day, table contain many open periods and many close period

1_open : [720,1140], -> Monday : 2 open/close periods
1_close: [840,1320], -> Monday : open/closed : 12h-14h and 19h-22h
2_open: [], -> Tuesday : Closed
2_close: [] -> Tuesday : Closed
3_open: [540] -> Wednesday : open at 9h during 1 period
3_close: [1020] -> Wednesday : close at 17h

How to read open/close periods ?

A restaurant is open the monday on 2 periods : 12h-14h and 19h-22h. In minutes, this will be : 720-840 and 1140-1320

In the table, we will find these values in 2 fields grouped by open or close time :

1_open : [720,1140]
1_close: [840,1320]

If we need to know if this restaurant is open at 13h00 :

• Convert 13h00 to minutes -> 780
• Identify day (for this example we use monday) -> 1
• Call server by this restaurant ID
• Read response and fields 1_open, 1_close
• Control that 780 is between 720 <-> 840 or 1140 <-> 1320
• 780 min is between first period for the monday, so the restaurant is OPEN !