extending to more datasets

Author: wenting-zhao

commit0/cli.py

@@ -257,18 +257,19 @@ def test(
     if reference:
         branch = "reference"
     else:
-        if "humaneval" not in commit0_config["dataset_name"].split("/")[-1].lower():
+        dataset_name = commit0_config["dataset_name"].lower()
+        if "humaneval" in dataset_name or "mbpp" in dataset_name or "bigcodebench" in dataset_name or "codecontests" in dataset_name:
+            branch = repo_or_repo_path
+        else:
             if branch is None and not reference:
                 git_path = os.path.join(
                     commit0_config["base_dir"], repo_or_repo_path.split("/")[-1]
                 )
                 branch = get_active_branch(git_path)
-        else:
-            branch = test_ids
 
     if stdin:
         # Read test names from stdin
-        test_ids = sys.stdin.read().strip()
+        test_ids = sys.stdin.read()
     elif test_ids is None:
         typer.echo("Error: test_ids must be provided or use --stdin option", err=True)
         raise typer.Exit(code=1)

commit0/harness/build.py

@@ -25,14 +25,15 @@ def main(
         dataset_name, split=dataset_split
     )  # type: ignore
     specs = []
-    if "swe" in dataset_name.lower():
+    dataset_name = dataset_name.lower()
+    if "swe" in dataset_name:
         dataset_type = "swebench"
-    elif "humaneval" in dataset_name.lower():
+    elif "humaneval" in dataset_name or "mbpp" in dataset_name or "bigcodebench" in dataset_name or "codecontests" in dataset_name:
         dataset_type = "simple"
     else:
         dataset_type = "commit0"
     for example in dataset:
-        if "swe" in dataset_name.lower() or dataset_type == "simple":
+        if "swe" in dataset_name or dataset_type == "simple":
             if split != "all" and split not in example["instance_id"]:
                 continue
         else:

commit0/harness/run_pytest_ids.py

@@ -51,17 +51,18 @@ def main(
     dataset: Iterator[Union[RepoInstance, SimpleInstance]] = load_dataset(
         dataset_name, split=dataset_split
     )  # type: ignore
+    dataset_name = dataset_name.lower()
     spec = None
     example = None
     repo_name = None
     dataset_type = None
     for example in dataset:
         if repo_or_repo_dir.endswith("/"):
             repo_or_repo_dir = repo_or_repo_dir[:-1]
-        if "swe" in dataset_name.lower():
+        if "swe" in dataset_name:
             repo_name = example["instance_id"]
             dataset_type = "swebench"
-        elif "humaneval" in dataset_name.lower():
+        elif "humaneval" in dataset_name or "mbpp" in dataset_name or "bigcodebench" in dataset_name or "codecontests" in dataset_name:
             repo_name = example["instance_id"]
             dataset_type = "simple"
         else:
@@ -130,7 +131,7 @@ def main(
                     )
 
         # make patch file
-        if "swe" in dataset_name.lower():
+        if "swe" in dataset_name:
             if branch == "reference":
                 patch = (
                     example["test"]["patch"] + "\n\n" + example["test"]["test_patch"]
@@ -164,7 +165,7 @@ def main(
                 + example["test"]
             )
         else:
-            solution = open(test_ids).read()
+            solution = test_ids
             prompt = example["prompt"] if "prompt" in example.keys() else ""
             matches = extract_code_blocks(solution)
             if len(matches) > 0:

commit0/harness/setup.py

@@ -23,12 +23,13 @@ def main(
     base_dir: str,
 ) -> None:
     dataset: Iterator[RepoInstance] = load_dataset(dataset_name, split=dataset_split)  # type: ignore
-    if "humaneval" in dataset_name.lower():
+    dataset_name = dataset_name.lower()
+    if "humaneval" in dataset_name or "mbpp" in dataset_name or "bigcodebench" in dataset_name or "codecontests" in dataset_name:
         return
     for example in dataset:
         repo_name = example["repo"].split("/")[-1]
         clone_url = f"https://github.com/{example['repo']}.git"
-        if "swe" in dataset_name.lower():
+        if "swe" in dataset_name:
             if repo_split != "all" and repo_split not in example["instance_id"]:
                 continue
             clone_dir = os.path.abspath(os.path.join(base_dir, example["instance_id"]))