filter: Keep index as a column for filtering

This allows column-based filters to use the id column.

Co-authored-by: Cornelius Roemer <[email protected]>

Author: victorlin

augur/filter/_run.py

@@ -200,6 +200,7 @@ def run(args):
         delimiters=[metadata_object.delimiter],
         columns=useful_metadata_columns,
         id_columns=[metadata_object.id_column],
+        keep_id_as_column=True,
         chunk_size=args.metadata_chunk_size,
         dtype="string",
     )

tests/functional/filter/cram/filter-metadata-id-columns.t

@@ -0,0 +1,32 @@
+Setup
+
+  $ source "$TESTDIR"/_setup.sh
+
+The id column can be used in --query, --exclude-where, and --include-where.
+
+  $ cat > metadata.tsv <<~~
+  > strain	accession	clade
+  > strain1	acc1	A
+  > strain2	acc2	A
+  > strain3	acc3	B
+  > strain4	acc4	B
+  > strain4	acc5	C
+  > ~~
+
+  $ ${AUGUR} filter \
+  >   --metadata metadata.tsv \
+  >   --metadata-id-columns accession \
+  >   --query "accession != 'acc5'" \
+  >   --exclude-where accession=acc2 clade=B \
+  >   --include-where accession=acc3 \
+  >   --output-strains filtered.txt
+  3 strains were dropped during filtering
+  	1 was dropped because of 'accession=acc2'
+  	2 were dropped because of 'clade=B'
+  	1 was filtered out by the query: "accession != 'acc5'"
+  	1 was added back because of 'accession=acc3'
+  2 strains passed all filters
+
+  $ sort filtered.txt
+  acc1
+  acc3