cnr-ibba / image-commondatapool Goto Github PK

FAO has changed the interface for breed data sheet links and now the respective links in IMAGE Data Portal are broken. The change is: instead of the country name, country 3 letters code is used now, e.g. NLD instead of Netherlands:

The link in Image:

https://fao-dadis-breed-detail.firebaseapp.com/?country=Netherlands&specie=Goose%20%28domestic%29&breed=Twentse%20landgans

should be:

https://fao-dadis-breed-detail.firebaseapp.com/?country=NLD&specie=Goose%20%28domestic%29&breed=Twentse%20landgans

An entry with spaces in URLField (such the dadis_url attribute at https://api.image2020genebank.eu/backend/dadis_link/1/) doesn't render correctly URLs with encoded spaces in templates. This beaviour seems depends by the urlize template tags, which is called in rest_framework/templates/rest_framework/base.html to render a string formatted like this (not a JSON but a <pre> formatted text):

{
    "url": "http://localhost:26081/backend/dadis_link/1/",
    "species": {
        "scientific_name": "Ovis aries",
        "common_name": "Sheep"
    },
    "supplied_breed": "Schoonebeek heath",
    "country": "Netherlands",
    "most_common_name": "Schoonebeeker Heideschaap",
    "transboundary_name": "Schoonebeek heath sheep",
    "other_name": [],
    "dadis_url": "https://fao-dadis-breed-detail.firebaseapp.com/?country=Netherlands&specie=Sheep&breed=Schoonebeeker%20Heideschaap",           
    "is_custom": true
}

more precisely, urlize cant represent the encoded space %20 in the href link:

from django.utils.html import urlize
urlize("https://fao-dadis-breed-detail.firebaseapp.com/?country=Netherlands&specie=Sheep&breed=Schoonebeeker%20Heideschaap")
# '<a href="https://fao-dadis-breed-detail.firebaseapp.com/?country=Netherlands&amp;specie=Sheep&amp;breed=Schoonebeeker+Heideschaap">https://fao-dadis-breed-detail.firebaseapp.com/?country=Netherlands&specie=Sheep&breed=Schoonebeeker%20Heideschaap</a>'

The returned link (in which %20 is changed in +) is not a valid dad-is permalink

Sending a query like https://api.image2020genebank.eu/backend/organism.geojson/?lat=10&lng=10&rad=10.1 cause Internal Server Error (500) since getting an integer from float string throws an exception. Fix GeoMaterialMixin to convert a string in a float number (not a int like this):

if rad:
    # express radius in km
    qs = qs.filter(distance__lte=int(rad)*1000)

Sometimes supplied_breed breed doesn't match the FAO DAD-IS breed name. Adding a record in a table is not sufficient since process_fao_metadata.py is able to operate over DAD-IS exported CSV file. Find a way to detect custom links between CDP supplied_breed and DAD-IS breed

• Add a custom link to DAD-IS table and manage table relationship through atomated scripts

Biosamples records, like SAMEA7044752 have a 'externalReferences' key in which crosslinks with other EBI databases are reported. Try to track this information into database and API