|
|
@ -1,5 +1,6 @@
|
|
|
|
import glob
|
|
|
|
import glob
|
|
|
|
import time
|
|
|
|
import time
|
|
|
|
|
|
|
|
import urllib.error
|
|
|
|
import urllib.request
|
|
|
|
import urllib.request
|
|
|
|
|
|
|
|
|
|
|
|
import requests
|
|
|
|
import requests
|
|
|
@ -30,7 +31,14 @@ with open("barcode_cleaned.csv") as bcfile:
|
|
|
|
|
|
|
|
|
|
|
|
rdf_url = "https://www.botanicalcollections.be/specimen/" + barcode + "/rdf"
|
|
|
|
rdf_url = "https://www.botanicalcollections.be/specimen/" + barcode + "/rdf"
|
|
|
|
print("fetching", rdf_url)
|
|
|
|
print("fetching", rdf_url)
|
|
|
|
|
|
|
|
try:
|
|
|
|
img_url = find_image(rdf_url)
|
|
|
|
img_url = find_image(rdf_url)
|
|
|
|
|
|
|
|
except urllib.error.HTTPError as er:
|
|
|
|
|
|
|
|
print("Error on ", rdf_url, str(er))
|
|
|
|
|
|
|
|
continue
|
|
|
|
|
|
|
|
if img_url is None:
|
|
|
|
|
|
|
|
print("Error. No image for barcode.")
|
|
|
|
|
|
|
|
continue
|
|
|
|
img_ext = str(img_url).rsplit(".", 1)[1]
|
|
|
|
img_ext = str(img_url).rsplit(".", 1)[1]
|
|
|
|
print("image url:", img_url)
|
|
|
|
print("image url:", img_url)
|
|
|
|
save_path = "specimen_img_raw/" + barcode + "." + img_ext
|
|
|
|
save_path = "specimen_img_raw/" + barcode + "." + img_ext
|
|
|
|