wip: parse-genbank-location

ingest/defaults/config.yaml

@@ -66,6 +66,8 @@ curate:
   strain_backup_fields: ["accession"]
   # List of date fields to standardize to ISO format YYYY-MM-DD
   date_fields: ["date", "date_released", "date_updated"]
+  # The field containing the location, in the format `<geo_loc_name>[:<region>][, <locality>]` (default: geo_loc_name)
+  location_field: "location"
   # List of expected date formats that are present in the date fields provided above
   # These date formats should use directives expected by datetime
   # See https://docs.python.org/3.9/library/datetime.html#strftime-and-strptime-format-codes

ingest/rules/curate.smk

@@ -72,6 +72,7 @@ rule curate:
         strain_regex=config["curate"]["strain_regex"],
         strain_backup_fields=config["curate"]["strain_backup_fields"],
         date_fields=config["curate"]["date_fields"],
+        location_field=config["curate"]["location_field"],
         expected_date_formats=config["curate"]["expected_date_formats"],
         articles=config["curate"]["titlecase"]["articles"],
         abbreviations=config["curate"]["titlecase"]["abbreviations"],
@@ -95,6 +96,7 @@ rule curate:
                 --date-fields {params.date_fields} \
                 --expected-date-formats {params.expected_date_formats} \
             | augur curate parse-genbank-location \
+                --location-field {params.location_field} \
             | augur curate titlecase \
                 --titlecase-fields {params.titlecase_fields} \
                 --articles {params.articles} \