WeedCOCO category names are now a combination of role and species (#84)

* WeedCOCO category names are now a combination of role and species

Fixes #83

Fixes #57

Helps towards specifying categories in importers (#33, #34)

* Pattern in quotes

* Allow space in species name

* Avoid changing timestamp.

* Try tweak pattern. NFI why it's failing to match 'weed: chinee apple'

* use scientific species names

* debugging

* More attempts to make CI happy

search/src/App.js

@@ -29,7 +29,7 @@ class App extends Component {
 				queryFormat: "or",
 				URLParams: true,
 				react: {
-					and: ["searchbox", "resslider", "agcontextfilter", "rolefilter", "speciesfilter", "grainstextfilter"]
+					and: ["searchbox", "resslider", "agcontextfilter", "categoryfilter", "grainstextfilter"]
 				},
 				...multilistFacetProps
 			}
@@ -60,24 +60,14 @@ class App extends Component {
 			>
 				<div style={{ position: "fixed", width: "20rem", overflow: "scroll", height: "100%" }}>
 					<MultiList
-						componentId="rolefilter"
-						title="Filter by role"
-						dataField="annotation__category__role.keyword"
-						sortBy="asc"
-						selectAllLabel="All roles"
-						placeholder="Search Role"
-						filterLabel="Role"
-						{...makeProps("rolefilter", true)}
-					/>
-					<MultiList
-						componentId="speciesfilter"
+						componentId="categoryfilter"
 						title="Filter by species"
-						dataField="annotation__category__species.keyword"
+						dataField="annotation__category__name.keyword"
 						sortBy="asc"
 						selectAllLabel="All species"
 						placeholder="Search Species"
 						filterLabel="Species"
-						{...makeProps("speciesfilter", true)}
+						{...makeProps("categoryfilter", true)}
 					/>
 					<MultiList
 						componentId="agcontextfilter"

weedcoco/cwfid_to_json/cwfid_imageinfo.json

@@ -17269,6 +17269,7 @@
     ],
     "categories": [
         {
+            "name": "crop: daugus carota",
             "common_name": "carrot",
             "species": "daugus carota",
             "eppo_taxon_code": "DAUCS",
@@ -17277,6 +17278,7 @@
             "id": 0
         },
         {
+            "name": "weed: UNSPECIFIED",
             "species": "UNSPECIFIED",
             "role": "weed",
             "id": 1
@@ -17415,4 +17417,4 @@
             "collection_id": 0
         }
     ]
-}
+}

weedcoco/deepweeds_to_json/deepweeds_imageinfo.json

@@ -262644,61 +262644,70 @@
             "id": 0,
             "role": "weed",
             "species": "ziziphus mauritiana",
-            "eppo_taxon_code": "ZIPMA"
+            "eppo_taxon_code": "ZIPMA",
+            "name": "weed: ziziphus mauritiana"
         },
         {
             "common_name": "lantana",
             "id": 1,
             "role": "weed",
             "species": "lantana camara",
-            "eppo_taxon_code": "LANCA"
+            "eppo_taxon_code": "LANCA",
+            "name": "weed: lantana camara"
         },
         {
             "common_name": "negative",
             "id": 8,
+            "name": "none",
             "role": "na"
         },
         {
             "common_name": "snake weed",
             "id": 7,
             "role": "weed",
             "species": "gutierrezia sarothrae",
-            "eppo_taxon_code": "GUESA"
+            "eppo_taxon_code": "GUESA",
+            "name": "weed: gutierrezia sarothrae"
         },
         {
             "common_name": "siam weed",
             "id": 6,
             "role": "weed",
             "species": "chromolaena odorata",
-            "eppo_taxon_code": "EUPOD"
+            "eppo_taxon_code": "EUPOD",
+            "name": "weed: chromolaena odorata"
         },
         {
             "common_name": "prickly acacia",
             "id": 4,
             "role": "weed",
             "species": "vachellia nilotica",
-            "eppo_taxon_code": "ACANL"
+            "eppo_taxon_code": "ACANL",
+            "name": "weed: vachellia nilotica"
         },
         {
             "common_name": "parthenium",
             "id": 3,
             "role": "weed",
             "species": "parthenium hysterophorus",
-            "eppo_taxon_code": "PTNHY"
+            "eppo_taxon_code": "PTNHY",
+            "name": "weed: parthenium hysterophorus"
         },
         {
             "common_name": "rubber vine",
             "id": 5,
             "role": "weed",
             "species": "cryptostegia grandiflora",
-            "eppo_taxon_code": "CVRGR"
+            "eppo_taxon_code": "CVRGR",
+            "name": "weed: cryptostegia grandiflora"
         },
         {
             "common_name": "parkinsonia",
             "id": 2,
             "role": "weed",
             "species": "parkinsonia aculeata",
-            "eppo_taxon_code": "PAKAC"
+            "eppo_taxon_code": "PAKAC",
+            "name": "weed: parkinsonia aculeata"
         }
     ],
     "info": [
@@ -262759,4 +262768,4 @@
             "id": 0
         }
     ]
-}
+}

weedcoco/importers/cwfid.py

@@ -27,14 +27,20 @@
 CATEGORY_MAP = {
     # TODO: we need a spec for species unspecified, and we need a spec for species specified
     "crop": {
+        "name": "crop: daugus carota",
         "common_name": "carrot",
         "species": "daugus carota",
         "eppo_taxon_code": "DAUCS",
         "eppo_nontaxon_code": "3UMRC",
         "role": "crop",
         "id": 0,
     },
-    "weed": {"species": "UNSPECIFIED", "role": "weed", "id": 1},
+    "weed": {
+        "name": "weed: UNSPECIFIED",
+        "species": "UNSPECIFIED",
+        "role": "weed",
+        "id": 1,
+    },
 }
 
 

weedcoco/importers/deepweeds.py

@@ -113,6 +113,7 @@
         category["id"] = row["Label"]
         categoryIDToCategories[categoryID] = category
         if category["common_name"] == "negative":
+            category["name"] = "none"
             category["role"] = "na"
         else:
             category["role"] = "weed"
@@ -140,6 +141,8 @@
         if category["common_name"] == "parkinsonia":
             category["species"] = "parkinsonia aculeata"
             category["eppo_taxon_code"] = "PAKAC"
+        if "name" not in category:
+            category["name"] = f"{category['role']}: {category['species']}"
 
     # Create an annotation
     ann = {}

weedcoco/schema/Category.yaml

@@ -0,0 +1,29 @@
+$id: https://weedid.sydney.edu.au/schema/Category.json
+type: object
+required:
+  - id
+  - name
+
+properties:
+  id:
+    type: number
+    description: |-
+      A number identifying the image.
+
+  name:
+    type: string
+    pattern: "^((crop|weed): ([a-z][a-z ]+|UNSPECIFIED)|crop|weed|none)$"
+    description: |-
+      A category for an annotation, consisting of an agricultural role and
+      optionally a biological name, separated by ": ". Currently allowed
+      agricultural roles are "weed" and "crop".  The "none" category may be
+      used to indicate that no weed or crop is identified in the image.
+
+      Annotations should be labelled with the most specific category possible,
+      usually "weed: <species name>", but super categories may be more general.
+      Biological names should be lowercase.
+      For an unspecified species of weed, "weed: UNSPECIFIED" may be used.
+
+      For example, to label an oat crop in an image, the category name would be
+      "crop: avena sativa". Its super categories may include "crop: grasses"
+      and "crop". Oats as a volunteer crop would be "weed: avena sativa".

weedcoco/schema/main.yaml

@@ -7,9 +7,10 @@ description: |
 type: object
 required:
   - agcontexts
+  - annotations
+  - categories
   - collections
   - images
-  - annotations
 properties:
   images:
     type: array
@@ -19,6 +20,10 @@ properties:
     type: array
     items:
       $ref: https://weedid.sydney.edu.au/schema/Annotation.json#/
+  categories:
+    type: array
+    items:
+      $ref: https://weedid.sydney.edu.au/schema/Category.json#/
   agcontexts:
     type: array
     items: