将loadBlob和loadProject中的函数转移到了utils文件中，并增加了另外两个通用函数用来抽取commit和blob的对应关系，增加了一个对象类用来存储Blob和Commit的对应关系

.isort.cfg

@@ -1,4 +1,4 @@
 [settings]
 line_length = 79
 multi_line_output = 3
-known_third_party =dulwich
+known_third_party =dulwich,loadProjects

loadBlob.py

@@ -1,54 +0,0 @@
-import os
-
-from dulwich.repo import Repo
-
-
-def loadBlobObjects(dir):
-    """
-    function: Find all blob objects in a given repo and make a list
-    input: the directory of a repo
-    output: a list of blob object
-    """
-    repo = Repo(dir)
-    objectlist = list(repo.object_store)
-    bloblist = []
-    for object in objectlist:
-        if (
-            repo.object_store[object].type == 3
-        ):  # type == 1 -> commit; type == 2 -> tree ;type == 3 -> blob.
-            bloblist.append(repo.object_store[object].data)
-    repo.close()
-    return bloblist
-
-
-def writeSourceCode(dir, code):
-    """
-    function: Write source code to a new txt file
-    input: the directory of a new text file, the code in type byte
-    output: a file with source code
-    """
-    file = open(dir, "wb")
-    file.write(code)
-    file.close()
-
-
-def writeSourceFiles(dir, list):
-    """
-    function: Write all source code to a new folder
-    input: the directory of a new folder, the code list in type byte
-    output: a folder with source files
-    """
-    folderpath = os.path.join(dir, "sourceFiles")
-    os.mkdir(folderpath)
-    i = 0
-    for code in list:
-        filepath = os.path.join(folderpath, str(i) + ".java")
-        writeSourceCode(filepath, code)
-        i = i + 1
-
-
-# a test for local repo
-# list = loadBlobObjects("C:/Users/Administrator/redis")
-# print(list[0])
-# writeSourceCode("C:/Users/Administrator/redis/1.txt", list[0])
-# writeSourceFiles("C:/Users/Administrator/redis", list)

loadProjects.py

@@ -1,13 +0,0 @@
-import sys
-
-
-def loadProjectList(dir):
-    """
-    Fuction:load project list
-    input:a txt file including all the address
-    output: a list including all the address of files in the input file
-    """
-    file = open(dir, "r")
-    list = file.read().splitlines()
-    file.close()
-    return list

models/BlobInfo.py

@@ -0,0 +1,20 @@
+from dulwich.objects import Blob, Commit
+from dulwich.repo import Repo
+
+
+class BlobInfo(object):
+    def __init__(
+        self,
+        repo: Repo = None,
+        commit: Commit = None,
+        filepath: str = None,
+        blob: Blob = None,
+    ) -> None:
+        self.repo = repo
+        self.commit = commit
+        self.filepath = filepath
+        self.blob = blob
+
+
+if __name__ == "__main__":
+    BlobInfo()

utils.py

@@ -0,0 +1,142 @@
+# aim: this is the utility file for other scripts
+# author: zhongyan, zhangxunhui
+
+import os
+from typing import List
+
+from dulwich.objects import Commit, Tree
+from dulwich.repo import Repo
+
+from models.BlobInfo import BlobInfo
+
+
+def loadBlobObjects(dir):
+    """
+    function: Find all blob objects in a given repo and make a list
+    input: the directory of a repo
+    output: a list of blob object
+    """
+    repo = Repo(dir)
+    objectlist = list(repo.object_store)
+    bloblist = []
+    for object in objectlist:
+        if (
+            repo.object_store[object].type == 3
+        ):  # type == 1 -> commit; type == 2 -> tree ;type == 3 -> blob.
+            bloblist.append(repo.object_store[object].data)
+    repo.close()
+    return bloblist
+
+
+def writeSourceCode(dir, code):
+    """
+    function: Write source code to a new txt file
+    input: the directory of a new text file, the code in type byte
+    output: a file with source code
+    """
+    file = open(dir, "wb")
+    file.write(code)
+    file.close()
+
+
+def writeSourceFiles(dir, list):
+    """
+    function: Write all source code to a new folder
+    input: the directory of a new folder, the code list in type byte
+    output: a folder with source files
+    """
+    folderpath = os.path.join(dir, "sourceFiles")
+    os.mkdir(folderpath)
+    i = 0
+    for code in list:
+        filepath = os.path.join(folderpath, str(i) + ".java")
+        writeSourceCode(filepath, code)
+        i = i + 1
+
+
+def loadProjectList(dir):
+    """
+    Fuction:load project list
+    input:a txt file including all the address
+    output: a list including all the address of files in the input file
+    """
+    file = open(dir, "r")
+    list = file.read().splitlines()
+    file.close()
+    return list
+
+
+def find_blobs_in_tree(
+    repo: Repo, commit: Commit, tree: Tree, relpath: bytes = b""
+) -> List[BlobInfo]:
+    """
+    Function: find all the blobs in a Tree directory
+    input:
+        - repo: Repo object, which indicates the repository that the Tree belongs to.
+        - commit: Commit object, which indicates the commit that the Tree belongs to.
+        - tree: Tree object, which is the target directory.
+        - relpath: the relative path for the Tree object.
+    return:
+        - A list of BlobInfo objects
+    """
+    result = []
+    for entry in tree.items():
+        obj = repo.object_store[entry.sha]
+        new_relpath = os.path.join(relpath, entry.path)
+        if obj.type_name == b"blob":
+            result.append(
+                BlobInfo(repo=repo, commit=commit, filepath=new_relpath, blob=obj)
+            )
+        else:
+            new_tree = obj
+            result.extend(
+                find_blobs_in_tree(
+                    repo=repo, commit=commit, tree=new_tree, relpath=new_relpath
+                )
+            )
+    return result
+
+
+def extract_commit_blob_relation(repo_path: str) -> List[dict]:
+    """
+    Function: extract the commit and blob relationship from a repository
+    input:
+        - repo_path: the string path of a git repository
+    return:
+        - A list of dict: {
+            key: Commit object
+            value: [BlobInfo object]
+        }
+    """
+    result = {}
+    commits: Commit = []
+    repo = Repo(repo_path)
+    object_shas = list(repo.object_store)
+    for object_sha in object_shas:
+        obj = repo.object_store[object_sha]
+        if obj.type_name == b"commit":
+            commits.append(obj)
+    for commit in commits:
+        blobs = find_blobs_in_tree(
+            repo=repo, commit=commit, tree=repo.object_store[commit.tree]
+        )
+        result[commit] = blobs
+    return result
+
+
+"""
+test extract_commit_blob_relation function
+1. firstly clone a repository to the root directory of this project, e.g., https://gitlink.org.cn/MillerEvan/bad_clone_prediction.git
+2. invoke function extract_commit_blob_relation(repo_path="bad_clone_prediction")
+"""
+result = extract_commit_blob_relation(repo_path="bad_clone_prediction")
+print(result)
+
+
+# a test for local repo
+# list = loadBlobObjects("C:/Users/Administrator/redis")
+# print(list[0])
+# writeSourceCode("C:/Users/Administrator/redis/1.txt", list[0])
+# writeSourceFiles("C:/Users/Administrator/redis", list)
+
+# print(loadProjectList("testForLoadProjects.txt"))