replicated project LT and milestones

github_api_req.py

@@ -0,0 +1,32 @@
+import requests
+import datetime as dt
+
+def main(vcs, begin_date):
+    gha_info = {}
+    #this is the entire list of Github 'milestones' grabbed from the API
+    gha_info['milestones'] = get_milestone_information(vcs)
+    #this is the count of milestones that occur after the cutoff date
+    gha_info['milestone_count'] = parse_milestones(gha_info['milestones'], begin_date)
+    return gha_info
+
+#this simple API call has been working for now but may need to be updated as more information is desired
+def get_milestone_information(vcs_path):
+    repo_uri=vcs_path[0]
+    repo_uri_list = repo_uri.split('/')
+    api_url = "https://api.github.com/repos/" + repo_uri_list[-2]  + "/" + repo_uri_list[-1] + "/milestones"
+    response = requests.get(api_url)
+    response_dict = response.json()
+    return response_dict
+
+def parse_milestones(milestones, earliest_date):
+    count_of_milestones = 0
+    for entry in milestones:
+        #if entry date is more recent than the earliest date we're looking at
+        if dt.datetime.fromisoformat(entry['created_at'][:-1]) > earliest_date:
+            count_of_milestones += 1
+    return count_of_milestones
+
+
+if __name__ == "__main__":
+    vcs = ['https://github.com/fabiangreffrath/woof']
+    main(vcs)

main.py

@@ -1,2 +1,34 @@
 import perceval
-import os
+import os
+import yaml
+import perceval_tasks as pt
+import github_api_req as gha
+import datetime as dt
+
+
+#TODO: get lists of authors -> get lists of contributors -> compute 'Mean Membership Type'
+
+#TODO: compute Formality level metric
+
+def main():
+    # we should discuss whether we're using the 93 day window that seems to be widely used or if we want a longer window
+    early_cutoff = dt.datetime(2015,3, 17)
+    print("earliest date examined: " + str(early_cutoff))
+    #placeholder for now
+    manifest = '../kaylea_dissertation/lifecycle/package_metadata/jupyter-notebook_manifest.yaml'
+    with open(manifest, 'r') as stream:
+        try:
+            config = yaml.safe_load(stream)
+            #below lines will probably need to be refactored as tasks expand
+            vcs_path = config['Upstream_VCS']
+            perceval_obj = pt.main(vcs_path, early_cutoff)
+            gha_obj = gha.main(vcs_path, early_cutoff)
+            #these are the two variables in the denominator of the formality measure
+            print(perceval_obj['age_of_project'])
+            print(gha_obj['milestone_count'])
+        except yaml.YAMLOError as err:
+            print(err)
+
+
+if __name__ == "__main__":
+    main()

perceval_tasks.py

@@ -1,24 +1,37 @@
-import perceval
+import datetime as dt
-import venv 
+from perceval.backends.core.git import Git
-import yaml
-import subprocess
 
-def main(manifest):
+#globals
-    with open(manifest, 'r') as stream:
+repo_dir = '/tmp/perceval.git'
-        #try:
-        config = yaml.safe_load(stream)
-        get_perceval_log(config['Upstream_VCS'])
-        print(config['Upstream_VCS'])
-        #except yaml.YAMLOError as err:
-        #    print(err)
 
-def get_perceval_log(vcs_path):
+#main function for all subsequent tasks using perceval 
-    print(vcs_path)
+def main(vcs_path, begin_date):
-    perceval_output = subprocess.run(['perceval', 'git', vcs_path[0], '--from-date', '2023-08-08'], capture_output=True)
+    perceval_info = {}
-    print(type(perceval_output.__str__()))
+    perceval_info['list_of_commits'] = get_perceval_log(vcs_path, begin_date)
-    with open('test_perceval_output.txt', 'w') as f:
+    perceval_info['age_of_project'] = get_repo_age(perceval_info['list_of_commits'] )
-        f.write(perceval_output.__str__())
+    return perceval_info
-    print('COMPLETE')
+
+
+# this is the primary function for getting the list of commits from perceval
+def get_perceval_log(vcs_path, begin_date):
+    repo = Git(uri=vcs_path[0], gitpath=repo_dir)
+    # this is a temporary date_from, will need to be more inclusive in the future
+    fetched_commits = repo.fetch(from_date=begin_date)
+    return list(fetched_commits)
+
+#this function is just to evaluate the repository age, as defined by Tamburri and used by van Meijel
+def get_repo_age(all_commits):
+    first_commit = all_commits[0]
+    last_commit = all_commits[-1]
+    first_date = dt.datetime.strptime(first_commit['data']["CommitDate"], '%c %z')
+    last_date = dt.datetime.strptime(last_commit['data']["CommitDate"], '%c %z')
+    print(first_date)
+    print("---------------------")
+    print(last_date)
+    #project life, as defined in YOSHI, unit is days
+    project_life = last_date - first_date
+    print(project_life)
+    return project_life
 
 
 if __name__ == "__main__":