refactored the release_frequency data gathering to use Augur instead …

…of the GitHub API, which removed the dependence on the GitHub API entirely, so the functions making the GH API calls were also removed

health_by_repo.py

@@ -28,13 +28,10 @@
     }
 Replace the 'x's with values to connect to your Augur database
 
-api_key
-A text file containing one GitHub personal access token
-
 Usage
 ----- 
 
-health_by_repo.py [-h] -o ORG_NAME -r REPO_NAME [-y YEARS] -c AUGUR_CONFIG -t API_KEY
+health_by_repo.py [-h] -o ORG_NAME -r REPO_NAME [-y YEARS] -c AUGUR_CONFIG 
 
 options:
   -h, --help            show this help message and exit
@@ -46,8 +43,6 @@
                         The number of years of data to collect (default to 1)
   -c AUGUR_CONFIG, --configfile AUGUR_CONFIG
                         The full file path to an Augur config.json file (required)
-  -t API_KEY, --token API_KEY
-                        The file where your GitHub personal access token can be found (required)
 
 Output
 ------
@@ -75,14 +70,12 @@
 parser.add_argument("-r", "--repo", required=True, dest = "repo_name", help="The name of a GitHub repository in that org where your PRs can be found (required)")
 parser.add_argument("-y", "--years", required=False, dest = "years", type=int, default=1, help="The number of years of data to collect (default to 1)")
 parser.add_argument("-c", "--configfile", required=True, dest = "augur_config", help="The full file path to an Augur config.json file (required)")
-parser.add_argument("-t", "--token", required=True, dest = "api_key", help="The file where your GitHub personal access token can be found (required)")
 
 args = parser.parse_args()
 org_name = args.org_name
 repo_name = args.repo_name
 years = args.years
 augur_config = args.augur_config
-api_key = args.api_key
 
 # Get the dates for the analysis using the years argument if provided
 days = 365 * years
@@ -99,23 +92,10 @@
 is_forked, is_archived = fork_archive(repo_name, org_name, engine)
 print('Forked:', str(is_forked), '\nArchived:', str(is_archived))
 
-# Get the GitHub API repository object used to gather release data.
-repo_api = repo_api_call(repo_name, org_name, api_key)
-
-# This section compares the Augur org / repo and renames them for repos that have been redirected
-# using the GH API as the canonical source of data for the org and repo, rather than what's in Augur
-# This fixes the problem of repos being renamed, but not updated in Augur.
-full_name = org_name + '/' + repo_name
-api_name = repo_api.full_name
-
-if full_name.lower() != api_name.lower():
-    org_name = api_name.split("/")[0]
-    repo_name = api_name.split("/")[1]
-
 # This section collects all of the data using the functions for each graph
 # found in common_functions.py and creates the graphs for each metric
 
-activity_release_graph(repo_name, org_name, start_date, end_date, repo_api)
+activity_release_graph(repo_id, repo_name, org_name, start_date, end_date, engine)
 
 sustain_prs_by_repo_graph(repo_id, repo_name, org_name, start_date, end_date, engine)
 

metrics/release_frequency.py

@@ -4,38 +4,49 @@
 """ Contains functions used to gather data and graph the Release Frequency metric
 """
 
-def get_release_data(repo_api):
-    """ Uses the GitHub REST API repository object to collect release data
+def get_release_data(repo_id, start_date, end_date, engine):
+    """ Get release data from the Augur database
 
     Parameters
     ----------
-    repo_api : GitHub repository object
+    repo_id : str
+    start_date : str
+    end_date : str
+    engine : sqlalchemy object
 
     Returns
     -------
     releases_df : dataframe
     """
     import pandas as pd
 
-    releases = repo_api.get_releases()
+    releases_df = pd.DataFrame()
 
-    releases_df = pd.DataFrame(
-        [x, x.tag_name, x.published_at, 1] for x in releases
-    )
-    releases_df.columns = ['release', 'name', 'date', 'releases']
+    release_query = f"""
+                    SELECT
+                        release_published_at as date
+                    FROM
+                        releases
+                    WHERE 
+                        repo_id = {repo_id}
+                        AND release_published_at > {start_date}
+                        AND release_published_at <= {end_date}
+                        """
+    releases_df = pd.read_sql_query(release_query, con=engine)
 
     return releases_df
 
-def activity_release_data(repo_name, org_name, start_date, end_date, repo_api):
-    """ Gathers release data from the GitHub API
+def activity_release_data(repo_id, repo_name, org_name, start_date, end_date, engine):
+    """ Takes release data and does some reformatting before graphing
 
     Parameters
     ----------
+    repo_id : str
     repo_name : str
     org_name : str
     start_date : str
     end_date : str
-    repo_api : GitHub repository object
+    engine : sqlalchemy object
 
     Returns
     -------
@@ -53,11 +64,11 @@ def activity_release_data(repo_name, org_name, start_date, end_date, repo_api):
     from utils.date_calcs import convert_dates
 
     try:
-        releases_df = get_release_data(repo_api)
+        releases_df = get_release_data(repo_id, start_date, end_date, engine)
         error_num = 0
         error_text = None
     except:
-        return -1, 'NO DATA', None, None, None, None, None, None, None, None
+        return -1, 'NO DATA', None, None, None, None, None, None
 
     start_dt, end_dt = convert_dates(start_date, end_date)
     six_mos_dt = end_dt - datetime.timedelta(days=180)
@@ -69,24 +80,25 @@ def activity_release_data(repo_name, org_name, start_date, end_date, repo_api):
 
     # return before creating plots if no release data in past 6 months
     if release_num == 0:
-        return -1, 'NO DATA', None, None, None, None, None, None, None, None
+        return -1, 'NO DATA', None, None, None, None, None, None
 
     title = org_name + "/" + repo_name + "\n" + str(release_num) + " releases in the past 6 months."
 
     interpretation = 'Interpretation: Healthy projects will have frequent releases with security updates, bug fixes, and features.'
 
     return error_num, error_text, releases_df, start_dt, end_dt, title, interpretation, release_num
 
-def activity_release_graph(repo_name, org_name, start_date, end_date, repo_api):
+def activity_release_graph(repo_id, repo_name, org_name, start_date, end_date, engine):
     """ Graphs the release data returned from the activity_release_data function
 
     Parameters
     ----------
+    repo_id : str
     repo_name : str
     org_name : str
     start_date : str
     end_date : str
-    repo_api : GitHub repository object
+    engine : sqlalchemy object
 
     Output
     ------
@@ -99,7 +111,7 @@ def activity_release_graph(repo_name, org_name, start_date, end_date, repo_api):
     import matplotlib.ticker as ticker
     from utils.file_operations import output_filename
 
-    error_num, error_text, releases_df, start_dt, end_dt, title, interpretation, release_num = activity_release_data(repo_name, org_name, start_date, end_date, repo_api)
+    error_num, error_text, releases_df, start_dt, end_dt, title, interpretation, release_num = activity_release_data(repo_id, repo_name, org_name, start_date, end_date, engine)
 
     if error_num == -1:
         return -1, 'NO DATA'
@@ -117,7 +129,7 @@ def activity_release_graph(repo_name, org_name, start_date, end_date, repo_api):
     ax.yaxis.set_major_locator(ticker.MultipleLocator(1))
     ax.set(yticklabels=[])
 
-    plottermonth = sns.lineplot(y='releases', x='date', data=releases_df, marker="X", linewidth=0, markersize=20).set_title(title, fontsize=30)
+    plottermonth = sns.lineplot(y=1, x='date', data=releases_df, marker="X", linewidth=0, markersize=20).set_title(title, fontsize=30)
     xlabel_str = 'Year Month\n\n' + interpretation
     plottermonthlabels = ax.set_xlabel(xlabel_str)