Support a data_provenance entry in auspice v2 configs

Adds support for a `data_provenance` field in the auspice v2 config and
exported auspice v2 JSONs through additions of schema definitions for
`data_provenance` and inclusion of an example provenance entry in the
Zika build's auspice config.

Fixes #691

augur/data/schema-auspice-config-v2.json

@@ -167,6 +167,27 @@
             "type": "object",
             "description": "UNUSED v1 syntax for defining vaccine choices",
             "$comment": "This is unused in `augur export v2` which gets vaccine info vis a node-data JSON file. It remains in the schema so that v1 config files can be used by `augur export v2`"
+        },
+        "data_provenance": {
+            "description": "Specify provenance of data included in this analysis",
+            "type": "array",
+            "minItems": 1,
+            "items": {
+                "type": "object",
+                "description": "An individual data source",
+                "additionalProperties": false,
+                "required": ["name"],
+                "properties": {
+                    "name": {
+                        "description": "Name of the data source",
+                        "type": "string"
+                    },
+                    "url": {
+                        "description": "URL to use in link to data source",
+                        "type": "string"
+                    }
+                }
+            }
         }
     }
-}
+}

augur/data/schema-export-v2.json

@@ -243,6 +243,27 @@
                     "$comment": "Frequencies could be specified here if desired",
                     "$comment": "If not specified, and frequencies are asked for in #/panels, then Auspice will attempt to fetch a seperate JSON",
                     "$comment": "cc John / Trevor"
+                },
+                "data_provenance": {
+                    "description": "Specify provenance of data included in this analysis",
+                    "type": "array",
+                    "minItems": 1,
+                    "items": {
+                        "type": "object",
+                        "description": "An individual data source",
+                        "additionalProperties": false,
+                        "required": ["name"],
+                        "properties": {
+                            "name": {
+                                "description": "Name of the data source",
+                                "type": "string"
+                            },
+                            "url": {
+                                "description": "URL to use in link to data source",
+                                "type": "string"
+                            }
+                        }
+                    }
                 }
             }
         },

augur/export_v2.py

@@ -433,6 +433,27 @@ def set_panels(data_json, config, cmd_line_panels):
     data_json['meta']["panels"] = panels
 
 
+def set_data_provenance(data_json, config):
+    """Set the data provenance from the given config file to the given data JSON.
+
+    Parameters
+    ----------
+    data_json : dict
+        auspice JSON to be updated
+    config : dict
+        config JSON with an expected ``data_provenance`` key
+
+    >>> config = {"data_provenance": [{"name": "GISAID"}, {"name": "INSDC"}]}
+    >>> data_json = {"meta": {}}
+    >>> set_data_provenance(data_json, config)
+    >>> data_json["meta"]["data_provenance"][0]["name"]
+    'GISAID'
+
+    """
+    if "data_provenance" in config:
+        data_json["meta"]["data_provenance"] = config["data_provenance"]
+
+
 def counter_to_disambiguation_suffix(count):
     """Given a numeric count of author papers, return a distinct alphabetical
     disambiguation suffix.
@@ -914,6 +935,7 @@ def run_v2(args):
     set_node_attrs_on_tree(data_json, node_attrs)
     set_geo_resolutions(data_json, config, args.geo_resolutions, read_lat_longs(args.lat_longs), node_attrs)
     set_panels(data_json, config, args.panels)
+    set_data_provenance(data_json, config)
 
     # Write outputs - the (unified) dataset JSON intended for auspice & perhaps the ref root-sequence JSON
     indent = {"indent": None} if args.minify_json else {}

tests/builds/zika/config/auspice_config_v2.json

@@ -47,5 +47,14 @@
     "map",
     "entropy",
     "frequencies"
+  ],
+  "data_provenance": [
+    {
+      "name": "GISAID"
+    },
+    {
+      "name": "COG UK",
+      "url": "https://www.cogconsortium.uk"
+    }
   ]
 }