Last active
March 2, 2022 16:08
-
-
Save ThomasG77/e6f4c2723682e0180ff0bbf59f417beb to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
curl -s "https://www.barrages-cfbr.eu/spip.php?page=gis_json&id_rubrique=114&objets=articles_branche&limit=1000" >| /tmp/barrages.geojson | |
cat /tmp/barrages.geojson | jq -r .features[].properties.title | cut -d "'" -f2 | |
cat /tmp/barrages.geojson | iconv -c -f utf-8 -t ascii | sed 's#([a-zA-Z])*##g' | sed "s#<a href='\|'>[)<\/A-Za-z0-9> -]*##g" >| /tmp/barrages_cleaned.geojson | |
jq -r .features[].properties.title /tmp/barrages_cleaned.geojson | sort | uniq >| /tmp/files.txt | |
# curl -s "https://www.barrages-cfbr.eu/spip.php?page=gis_json&id_rubrique=114&objets=articles_branche&limit=1000" | jq -r .features[].properties.title | cut -d "'" -f2 >| /tmp/files.txt | |
true >| /tmp/urls.txt; | |
for i in $(cat /tmp/files.txt); | |
do echo "https://www.barrages-cfbr.eu/"$i >> /tmp/urls.txt; | |
done; | |
true >| /tmp/content_invalidated.ndjson | |
for i in $(cat /tmp/urls.txt); | |
do echo "$i"; | |
echo $(curl -s $i | sed 's# ##g' | pup --charset utf8 '.chapo' | sed 's#<p>\|</p>\|<br class="autobr">\|</div>\|<div class="chapo surlignable">##g' | tr '\n' ' ' | sed 's#<strong>#","#g' | sed 's#</strong>##g' | sed 's#:#":"#g' | sed 's#^[ ]*","#{"#g')'", "html": "'$i'"}' | jq -c '.' >> /tmp/content_invalidated.ndjson; | |
done; | |
sed -i 's#https://www.barrages-cfbr.eu/##g' /tmp/content_invalidated.ndjson; | |
ndjson-join 'd.properties.title' 'd.html' --left <(jq -c '.features[]' /tmp/barrages_cleaned.geojson) <(cat /tmp/content_invalidated.ndjson) | ndjson-map 'd[0].properties=d[1],d[0]' >| /tmp/barrages_infos.ndgeojson | |
cat /tmp/barrages_infos.ndgeojson | jq '.properties'| jq -n '[inputs | keys[]] | unique | sort' | |
sed -i 's#" Altitude de la crête"#"alt_crete_m"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Année de mise en service"#"annee_mes"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Commune "#"commune"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Département "#"dept"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Hauteur"#"haut_m"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Latitude"#"lat"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Latitude "#"lat"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Longitude"#"lon"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Longitude "#"lon"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Longueur"#"longueur_m"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Propriétaire / Exploitant"#"prop_expl"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Propriétaire / Exploitant "#"prop_expl"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Propriétaire/Exploitant"#"prop_expl"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Rivière"#"riviere"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Surface de la retenue"#"surf_ret_ha"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Surface du bassin versant"#"surf_bv_km2"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Type Barrage"#"type_b"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Usage(s)"#"usages"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" Volume de la retenue"#"vol_ret_m3"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" latitude"#"lat1"#g' /tmp/barrages_infos.ndgeojson | |
sed -i 's#" longitude"#"lon1"#g' /tmp/barrages_infos.ndgeojson | |
jq --slurp . /tmp/barrages_infos.ndgeojson | jq '{"type": "FeatureCollection","features": .}' >| /tmp/barrages_infos.geojson | |
ogrinfo -dialect SQLite -sql "SELECT id,trim(prop_expl) AS prop_expl,cast(replace(replace(haut_m, ' m', ''), ',', '.') AS real) AS haut_m,cast(replace(replace(vol_ret_m3, 'milliers de m³', ''), ' ', '') AS integer) AS vol_ret_m3,trim(riviere) AS riviere,trim(type_b) AS type_b,cast(replace(replace(replace(longueur_m, ' m3', ''), ' m', ''), ',', '.') AS integer) AS longueur_m,cast(trim(replace(annee_mes, 'm3', '')) AS integer) AS annee_mes,cast(trim(replace(replace(surf_ret_ha, 'ha', ''), ',', '.')) AS real) AS surf_ret_ha,cast(trim(replace(replace(surf_bv_km2, 'km²', ''), ',', '.')) AS real) AS surf_bv_km2,cast(trim(replace(replace(alt_crete_m, 'm', ''), ',', '.')) AS real) AS alt_crete_m,trim(usages) AS usages,trim(commune) AS commune,trim(dept) AS dept,cast(lat AS real) AS lat,cast(lon AS real) AS lon, 'https://www.barrages-cfbr.eu/' || html AS url,trim(lat1) AS lat1,trim(lon1) AS lon1,geometry FROM barrages_infos" /tmp/barrages_infos.geojson | |
ogr2ogr -f GeoJSON -dialect SQLite -sql "SELECT id,trim(prop_expl) AS prop_expl,cast(replace(replace(haut_m, ' m', ''), ',', '.') AS real) AS haut_m,cast(replace(replace(vol_ret_m3, 'milliers de m³', ''), ' ', '') AS integer) AS vol_ret_m3,trim(riviere) AS riviere,trim(type_b) AS type_b,cast(replace(replace(replace(longueur_m, ' m3', ''), ' m', ''), ',', '.') AS integer) AS longueur_m,cast(trim(replace(annee_mes, 'm3', '')) AS integer) AS annee_mes,cast(trim(replace(replace(surf_ret_ha, 'ha', ''), ',', '.')) AS real) AS surf_ret_ha,cast(trim(replace(replace(surf_bv_km2, 'km²', ''), ',', '.')) AS real) AS surf_bv_km2,cast(trim(replace(replace(alt_crete_m, 'm', ''), ',', '.')) AS real) AS alt_crete_m,trim(usages) AS usages,trim(commune) AS commune,trim(dept) AS dept,cast(lat AS real) AS lat,cast(lon AS real) AS lon, 'https://www.barrages-cfbr.eu/' || html AS url,trim(lat1) AS lat1,trim(lon1) AS lon1,geometry FROM barrages_infos" /tmp/barrages_infos_cleaned.geojson /tmp/barrages_infos.geojson -nln barrages_infos_cleaned -lco RFC7946=YES |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment