test: add interesting cases for multi -d support

python/tests/test_vwconfig.py

@@ -19,8 +19,8 @@ def test_vw_config_manager():
     expected_set = {
         "--no_stdin",
         "--quiet",
-        "--loss_function logistic",
-        "--data test/train-sets/rcv1_small.dat",
+        "--loss_function=logistic",
+        "--data=test/train-sets/rcv1_small.dat",
     }
     expected_reductions = {"gd", "scorer-identity", "count_label"}
 

python/vowpalwabbit/pyvw.py

@@ -176,11 +176,10 @@ def __str__(self):
             if self.is_flag():
                 return "--{}".format(self.name)
             else:
-                # missing list case
                 if isinstance(self.value, list):
-                    return "**NOT_IMPL**"
+                    return " ".join(map(lambda x: f"--{self.name}={x}", self.value))
                 else:
-                    return "--{} {}".format(self.name, self.value)
+                    return "--{}={}".format(self.name, self.value)
         else:
             return ""
 

test/core.vwtest.json

@@ -5123,5 +5123,91 @@
     "input_files": [
       "train-sets/cb_l_namespace.txt"
     ]
+  },
+  {
+    "id": 399,
+    "desc": "test multi d -- based on test 316 to share same metrics of the complete file dsjson_cb.json (which is split into two here)",
+    "bash_command": "./negative-test.sh {VW} -d train-sets/dsjson_cb_part1.json -d train-sets/dsjson_cb_part2.json --dsjson --cb_explore_adf --epsilon 0.2 --quadratic GT --extra_metrics metrics_time.json",
+    "diff_files": {
+      "stdout": "train-sets/ref/metrics_time_fail.stdout",
+      "stderr": "train-sets/ref/multid_fail.stderr"
+    },
+    "input_files": [
+      "./negative-test.sh",
+      "train-sets/dsjson_cb_part1.json",
+      "train-sets/dsjson_cb_part2.json"
+    ]
+  },
+  {
+    "id": 400,
+    "desc": "test multi d positional -- based on test 316 to share same metrics of the complete file dsjson_cb.json (which is split into two here)",
+    "vw_command": "--dsjson --cb_explore_adf --epsilon 0.2 --quadratic GT --extra_metrics metrics_time.json train-sets/dsjson_cb_part1.json train-sets/dsjson_cb_part2.json",
+    "diff_files": {
+      "metrics_time.json": "test-sets/ref/metrics_time_one.json",
+      "stdout": "train-sets/ref/metrics_time_one.stdout",
+      "stderr": "train-sets/ref/multid_one.stderr"
+    },
+    "input_files": [
+      "train-sets/dsjson_cb_part1.json",
+      "train-sets/dsjson_cb_part2.json"
+    ]
+  },
+  {
+    "id": 401,
+    "desc": "test multi d - see test 121 cb_explore",
+    "bash_command": "./negative-test.sh {VW} -d train-sets/rcv1_raw_cb_small_p1.vw -d train-sets/rcv1_raw_cb_small_p2.vw --cb_explore 2 --ngram 2 --skips 4 -b 24 -l 0.25 -p rcv1_raw_cb_explore.preds",
+    "diff_files": {
+      "stderr": "train-sets/ref/rcv1_raw_cb_explore_pass2_split.stderr",
+      "rcv1_raw_cb_explore.preds": "pred-sets/ref/rcv1_raw_cb_explore_pass2_empty.preds",
+      "stdout": "train-sets/ref/rcv1_raw_cb_explore_pass2_fail.stdout"
+    },
+    "input_files": [
+      "./negative-test.sh",
+      "train-sets/rcv1_raw_cb_small_p1.vw",
+      "train-sets/rcv1_raw_cb_small_p2.vw"
+    ]
+  },
+  {
+    "id": 402,
+    "desc": "cb_explore - passes 2",
+    "vw_command": "-d train-sets/rcv1_raw_cb_small.vw --cb_explore 2 --ngram 2 --skips 4 -b 24 -l 0.25 -p rcv1_raw_cb_explore.preds --passes 2 -c -k",
+    "diff_files": {
+      "stderr": "train-sets/ref/rcv1_raw_cb_explore_pass2.stderr",
+      "rcv1_raw_cb_explore.preds": "pred-sets/ref/rcv1_raw_cb_explore_pass2.preds",
+      "stdout": "train-sets/ref/rcv1_raw_cb_explore_pass2.stdout"
+    },
+    "input_files": [
+      "train-sets/rcv1_raw_cb_small.vw"
+    ]
+  },
+  {
+    "id": 403,
+    "desc": "previous but with split files - cb_explore - passes 2",
+    "bash_command": "./negative-test.sh {VW} -d train-sets/rcv1_raw_cb_small_p1.vw -d train-sets/rcv1_raw_cb_small_p2.vw --cb_explore 2 --ngram 2 --skips 4 -b 24 -l 0.25 -p rcv1_raw_cb_explore.preds --passes 2 -c -k --single_cache_file",
+    "diff_files": {
+      "stderr": "train-sets/ref/rcv1_raw_cb_explore_pass2_split.stderr",
+      "rcv1_raw_cb_explore.preds": "pred-sets/ref/rcv1_raw_cb_explore_pass2_empty.preds",
+      "stdout": "train-sets/ref/rcv1_raw_cb_explore_pass2_fail.stdout"
+    },
+    "input_files": [
+      "./negative-test.sh",
+      "train-sets/rcv1_raw_cb_small_p1.vw",
+      "train-sets/rcv1_raw_cb_small_p2.vw"
+    ]
+  },
+  {
+    "id": 404,
+    "desc": "test multi d positional mixed with -d, this should fail",
+    "vw_command": "--dsjson --cb_explore_adf --epsilon 0.2 train-sets/dsjson_cb_part2.json --quadratic GT --extra_metrics metrics_time.json -d train-sets/dsjson_cb_part1.json",
+    "diff_files": {
+      "stderr": "train-sets/ref/multid_mixed.stderr",
+      "stdout": "train-sets/ref/multid_mixed.stdout"
+    },
+    "input_files": [
+      "train-sets/dsjson_cb_part1.json",
+      "train-sets/dsjson_cb_part2.json"
+    ]
   }
+
+
 ]