Merge pull request #261 from 3N4N/enan

Fix: issues of 'harness' in Windows

swebench/harness/docker_build.py

@@ -52,7 +52,7 @@ def setup_logger(instance_id: str, log_file: Path, mode="w"):
     """
     log_file.parent.mkdir(parents=True, exist_ok=True)
     logger = logging.getLogger(f"{instance_id}.{log_file.name}")
-    handler = logging.FileHandler(log_file, mode=mode)
+    handler = logging.FileHandler(log_file, mode=mode, encoding='utf-8')
     formatter = logging.Formatter("%(asctime)s - %(levelname)s - %(message)s")
     handler.setFormatter(formatter)
     logger.addHandler(handler)

swebench/harness/dockerfiles.py

@@ -36,6 +36,7 @@
 _DOCKERFILE_ENV = r"""FROM --platform={platform} sweb.base.{arch}:latest
 
 COPY ./setup_env.sh /root/
+RUN sed -i -e 's/\r$//' /root/setup_env.sh
 RUN chmod +x /root/setup_env.sh
 RUN /bin/bash -c "source ~/.bashrc && /root/setup_env.sh"
 
@@ -48,6 +49,7 @@
 _DOCKERFILE_INSTANCE = r"""FROM --platform={platform} {env_image_name}
 
 COPY ./setup_repo.sh /root/
+RUN sed -i -e 's/\r$//' /root/setup_repo.sh
 RUN /bin/bash /root/setup_repo.sh
 
 WORKDIR /testbed/

swebench/harness/run_evaluation.py

@@ -2,12 +2,13 @@
 
 import docker
 import json
-import resource
+import platform
+if platform.system() == 'Linux': import resource
 import traceback
 
 from argparse import ArgumentParser
 from concurrent.futures import ThreadPoolExecutor, as_completed
-from pathlib import Path
+from pathlib import Path,PurePosixPath
 from tqdm import tqdm
 
 from swebench.harness.constants import (
@@ -121,7 +122,7 @@ def run_instance(
         logger.info(
             f"Intermediate patch for {instance_id} written to {patch_file}, now applying to container..."
         )
-        copy_to_container(container, patch_file, Path(DOCKER_PATCH))
+        copy_to_container(container, patch_file, PurePosixPath(DOCKER_PATCH))
 
         # Attempt to apply patch to container
         val = container.exec_run(
@@ -511,7 +512,8 @@ def main(
     """
     # set open file limit
     assert len(run_id) > 0, "Run ID must be provided"
-    resource.setrlimit(resource.RLIMIT_NOFILE, (open_file_limit, open_file_limit))
+    if platform.system() == 'Linux':
+        resource.setrlimit(resource.RLIMIT_NOFILE, (open_file_limit, open_file_limit))
     client = docker.from_env()
 
     # load predictions as map of instance_id to prediction

swebench/harness/utils.py

@@ -1,5 +1,6 @@
 import json
 import os
+import posixpath
 from pathlib import Path
 import re
 import requests
@@ -176,7 +177,7 @@ def get_lines_with_word(text, target_word):
 @cache
 def get_environment_yml_by_commit(repo: str, commit: str, env_name: str) -> str:
     for req_path in MAP_REPO_TO_ENV_YML_PATHS[repo]:
-        reqs_url = os.path.join(SWE_BENCH_URL_RAW, repo, commit, req_path)
+        reqs_url = posixpath.join(SWE_BENCH_URL_RAW, repo, commit, req_path)
         reqs = requests.get(reqs_url, headers=HEADERS)
         if reqs.status_code == 200:
             break
@@ -221,7 +222,7 @@ def get_environment_yml(instance: SWEbenchInstance, env_name: str) -> str:
 @cache
 def get_requirements_by_commit(repo: str, commit: str) -> str:
     for req_path in MAP_REPO_TO_REQS_PATHS[repo]:
-        reqs_url = os.path.join(SWE_BENCH_URL_RAW, repo, commit, req_path)
+        reqs_url = posixpath.join(SWE_BENCH_URL_RAW, repo, commit, req_path)
         reqs = requests.get(reqs_url, headers=HEADERS)
         if reqs.status_code == 200:
             break