entries in the dataset now store what build system

they use

dataset.py

@@ -13,6 +13,7 @@ class Metadata:
     pr_number: int
     merge_commit_sha: str # to checkout for the tests
     successful: bool = True
+    build_system: str = ""
     reason_for_failure: str = ""
     last_cmd_error_msg: str = ""
 

handlers.py

@@ -124,6 +124,10 @@ class BuildHandler(ABC):
     def clean_repo(self) -> None:
         self.container.exec_run(self.clean_cmd())
 
+    @abstractmethod
+    def get_type(self) -> str:
+        pass
+
     @abstractmethod
     def compile_cmd(self) -> str:
         pass
@@ -160,6 +164,9 @@ class MavenHandler(BuildHandler):
         # -Dstyle.color=never: Disables ANSI colors.
         # -Dartifact.download.skip=true: Prevents Maven from printing download logs (but still downloads dependencies when needed).
 
+    def get_type(self) -> str:
+        return "maven"
+
     def compile_cmd(self) -> str:
         return f"{self.base_cmd} clean compile"
 
@@ -205,6 +212,9 @@ class GradleHandler(BuildHandler):
         super().__init__(repo_path, build_file, updates)
         self.base_cmd = "gradle --no-daemon --console=plain"
 
+    def get_type(self) -> str:
+        return "gradle"
+
     def compile_cmd(self) -> str:
         return f"{self.base_cmd} compileJava"
 

pull_requests.py

@@ -121,6 +121,7 @@ def process_pull(repo: Repository, pr: PullRequest, dataset: Dataset, repos_dir:
         entry.metadata.reason_for_failure = "Couldn't get the build handler"
         entry.metadata.successful = False
         return
+    entry.metadata.build_system = build_handler.get_type()
     build_handler.set_client(docker_client)
         
     def _check_coverage(files: list[str]):