ingest: Remove use of ncov-ingest geolocation rules (#75)

Remove the use of the ncov-ingest geolocation rules since Augur
now uses the built-in geolocation rules by default.

Depends on the release of
<nextstrain/augur#1745>

Part of larger migration
nextstrain/public#17

ingest/defaults/config.yaml

@@ -32,10 +32,6 @@ ncbi_datasets_fields:
 
 # Config parameters related to the curate pipeline
 curate:
-  # URL pointed to public generalized geolocation rules
-  # For the Nextstrain team, this is currently
-  # "https://raw.githubusercontent.com/nextstrain/ncov-ingest/master/source-data/gisaid_geoLocationRules.tsv"
-  geolocation_rules_url: "https://raw.githubusercontent.com/nextstrain/ncov-ingest/master/source-data/gisaid_geoLocationRules.tsv"
   # The path to the local geolocation rules within the pathogen repo
   # The path should be relative to the ingest directory.
   local_geolocation_rules: "defaults/geolocation-rules.tsv"

ingest/rules/curate.smk

@@ -13,28 +13,6 @@ Parameters are expected to be defined in `config.curate`.
 """
 
 
-rule fetch_general_geolocation_rules:
-    output:
-        general_geolocation_rules="data/general-geolocation-rules.tsv",
-    params:
-        geolocation_rules_url=config["curate"]["geolocation_rules_url"],
-    shell:
-        """
-        curl -fsSL --output {output.general_geolocation_rules} {params.geolocation_rules_url}
-        """
-
-
-rule concat_geolocation_rules:
-    input:
-        general_geolocation_rules="data/general-geolocation-rules.tsv",
-        local_geolocation_rules=config["curate"]["local_geolocation_rules"],
-    output:
-        all_geolocation_rules="data/all-geolocation-rules.tsv",
-    shell:
-        """
-        cat {input.general_geolocation_rules} {input.local_geolocation_rules} >> {output.all_geolocation_rules}
-        """
-
 def format_field_map(field_map: dict[str, str]) -> str:
     """
     Format dict to `"key1"="value1" "key2"="value2"...` for use in shell commands.
@@ -44,7 +22,7 @@ def format_field_map(field_map: dict[str, str]) -> str:
 rule curate:
     input:
         sequences_ndjson="data/genbank.ndjson",
-        all_geolocation_rules="data/all-geolocation-rules.tsv",
+        geolocation_rules=config["curate"]["local_geolocation_rules"],
         annotations=config["curate"]["annotations"],
         manual_mapping="defaults/host_hostgenus_hosttype_map.tsv",
     output:
@@ -93,7 +71,7 @@ rule curate:
                 --default-value {params.authors_default_value:q} \
                 --abbr-authors-field {params.abbr_authors_field} \
             | augur curate apply-geolocation-rules \
-                --geolocation-rules {input.all_geolocation_rules} \
+                --geolocation-rules {input.geolocation_rules} \
             | ./scripts/transform-state-names \
             | ./scripts/post_process_metadata.py \
             | ./scripts/transform-new-fields \