Fix batch repo names containing github (#1406)

Co-authored-by: zhouyijin <zhouyijin@meituan.com>

sweagent/run/batch_instances.py

@@ -23,7 +23,7 @@ from sweagent.agent.problem_statement import (
 from sweagent.environment.repo import GithubRepoConfig, LocalRepoConfig, PreExistingRepoConfig, SWESmithRepoConfig
 from sweagent.environment.swe_env import EnvironmentConfig
 from sweagent.utils.files import load_file
-from sweagent.utils.github import _is_repo_private
+from sweagent.utils.github import _is_github_repo_url, _is_repo_private
 from sweagent.utils.log import get_logger
 
 logger = get_logger("swea-config", emoji="🔧")
@@ -98,7 +98,7 @@ class SimpleBatchInstance(BaseModel):
     repo_name: str = ""
     """Specifies the repository to use. If empty, no repository is used.
     If the string does not contain a slash, it is interpreted as an already existing repository at the root
-    of the docker container. If it contains the word "github", it is interpreted as a github repository.
+    of the docker container. If it is a GitHub URL, it is interpreted as a github repository.
     Else, it is interpreted as a local repository.
     """
     base_commit: str = "HEAD"
@@ -130,7 +130,7 @@ class SimpleBatchInstance(BaseModel):
 
         if not self.repo_name:
             repo = None
-        elif "github" in self.repo_name:
+        elif _is_github_repo_url(self.repo_name):
             repo = GithubRepoConfig(github_url=self.repo_name, base_commit=self.base_commit)
         elif "/" not in self.repo_name:
             repo = PreExistingRepoConfig(repo_name=self.repo_name, base_commit=self.base_commit)

tests/test_batch_instance.py

@@ -4,7 +4,7 @@ import pytest
 from swerex.deployment.config import DockerDeploymentConfig
 
 from sweagent.agent.problem_statement import TextProblemStatement
-from sweagent.environment.repo import PreExistingRepoConfig
+from sweagent.environment.repo import GithubRepoConfig, PreExistingRepoConfig
 from sweagent.run.batch_instances import BatchInstance, SimpleBatchInstance, SWEBenchInstances, _slice_spec_to_slice
 
 
@@ -22,6 +22,31 @@ def test_simple_batch_from_swe_bench_to_full_batch_instance(test_data_sources_pa
     assert instance.problem_statement.id == "pydicom__pydicom-1458"
 
 
+@pytest.mark.parametrize("repo_name", ["go-github", "github", "github-action-build-chain"])
+def test_simple_batch_treats_repo_names_containing_github_as_existing_repos(repo_name):
+    instance = SimpleBatchInstance(
+        image_name="python:3.11",
+        problem_statement="Fix the bug",
+        instance_id="repo-name-with-github",
+        repo_name=repo_name,
+    ).to_full_batch_instance(DockerDeploymentConfig(image="python:3.11"))
+
+    assert isinstance(instance.env.repo, PreExistingRepoConfig)
+    assert instance.env.repo.repo_name == repo_name
+
+
+def test_simple_batch_treats_github_urls_as_github_repos():
+    instance = SimpleBatchInstance(
+        image_name="python:3.11",
+        problem_statement="Fix the bug",
+        instance_id="github-url",
+        repo_name="https://github.com/SWE-agent/test-repo",
+    ).to_full_batch_instance(DockerDeploymentConfig(image="python:3.11"))
+
+    assert isinstance(instance.env.repo, GithubRepoConfig)
+    assert instance.env.repo.github_url == "https://github.com/SWE-agent/test-repo"
+
+
 def test_slice_spec_to_slice():
     assert _slice_spec_to_slice("10") == slice(10)
     assert _slice_spec_to_slice("10:20") == slice(10, 20)