🚧 export v2: Automatically minify output

Automatically format the output JSON file based on the number of tips,
unless either --minify-json or --no-minify-json (new option) is
specified.

augur/export_v2.py

@@ -16,6 +16,10 @@
 from .utils import read_node_data, write_json, read_config, read_lat_longs, read_colors
 from .validate import export_v2 as validate_v2, auspice_config_v2 as validate_auspice_config_v2, ValidateError
 
+
+MINIFY_THRESHOLD = 1000 # number of nodes
+
+
 # Set up warnings & exceptions
 warn = warnings.warn
 deprecationWarningsEmitted = False
@@ -878,7 +882,10 @@ def register_parser(parent_subparsers):
     optional_settings = parser.add_argument_group(
         title="OPTIONAL SETTINGS"
     )
-    optional_settings.add_argument('--minify-json', action="store_true", help="export JSONs without indentation or line returns")
+    minify = optional_settings.add_mutually_exclusive_group()
+    minify.add_argument('--minify-json', action="store_true", help="export JSONs without indentation or line returns, regardless of how many nodes are in the tree.")
+    minify.add_argument('--no-minify-json', action="store_true", help="export JSONs to be human readable, regardless of how many nodes are in the tree.")
+
     root_sequence = optional_settings.add_mutually_exclusive_group()
     root_sequence.add_argument('--include-root-sequence', action="store_true", help="Export an additional JSON containing the root sequence (reference sequence for vcf) used to identify mutations. The filename will follow the pattern of <OUTPUT>_root-sequence.json for a main auspice JSON of <OUTPUT>.json")
     root_sequence.add_argument('--include-root-sequence-inline', action="store_true", help="Export the root sequence (reference sequence for vcf) used to identify mutations as part of the main dataset JSON. This should only be used for small genomes for file size reasons.")
@@ -1165,8 +1172,22 @@ def run(args):
     if config.get("extensions"):
         data_json["meta"]["extensions"] = config["extensions"]
 
+    # Should output be minified?
+    # User-specified arguments take precedence before determining behavior based
+    # on the size of the tree.
+    if args.minify_json:
+        minify = True
+    elif args.no_minify_json:
+        minify = False
+    else:
+        num_nodes = len(node_data["nodes"])
+        if num_nodes > MINIFY_THRESHOLD:
+            minify = True
+        else:
+            minify = False
+
     # Write outputs - the (unified) dataset JSON intended for auspice & perhaps the ref root-sequence JSON
-    indent = {"indent": None} if args.minify_json else {}
+    indent = {"indent": None} if minify else {}
     if args.include_root_sequence or args.include_root_sequence_inline:
         # Note - argparse enforces that only one of these args will be true
         if 'reference' in node_data:

tests/functional/export_v2/cram/minify-output.t

@@ -0,0 +1,31 @@
+Setup
+
+  $ source "$TESTDIR"/_setup.sh
+
+Tree with a few nodes is not automatically minified.
+
+  $ ${AUGUR} export v2 \
+  >   --tree "$TESTDIR/../data/tree.nwk" \
+  >   --output output.json &>/dev/null
+
+  $ head -c 20 output.json
+  {
+    "version": "v2", (no-eol)
+
+It can be forcefully minified with an argument.
+
+  $ ${AUGUR} export v2 \
+  >   --tree "$TESTDIR/../data/tree.nwk" \
+  >   --minify-json \
+  >   --output output.json &>/dev/null
+
+  $ head -c 20 output.json
+  {"version": "v2", "m (no-eol)
+
+Tree with many nodes is automatically minified.
+
+FIXME: find a tree with >1000 nodes for testing here.
+
+It can be forcefully not minified with an argument.
+
+FIXME: add test for --no-minify-json.