Add the oldest changests by affiliation study

This allows to have the first 10 oldest changesets ordered by
affiliation. The difference with the rest of the studies is that
this one includes affiliation information.

vizgrimoire/analysis/gerrit_studies.py

@@ -21,6 +21,7 @@
 #     Daniel Izquierdo Cortazar <dizquierdo@bitergia.com>
 
 from sets import Set
+import pandas
 
 from vizgrimoire.analysis.analyses import Analyses
 
@@ -99,7 +100,7 @@ class MostActiveChangesetsWaiting4Reviewer(Analyses):
     id = "list_most_active_changesets_waiting4reviewer"
     name = "List of the most active changesets waiting for a reviewer action"
     desc = "List of the most active changesets waiting for a reviewer action"
-    
+
     def create_report(self, data_source, destdir):
         if data_source != SCR: return
         self.result(data_source, destdir)
@@ -209,6 +210,80 @@ def result(self, data_source = None, destdir = None):
 
         return data
 
+
+class OldestChangesetsByAffiliation(Analyses):
+
+    """ This class provides the oldest changesets without activity
+        ordered by affiliation and by the last upload.
+    """
+
+    id = "oldest_changesets_by_affiliation"
+    name = "The oldest changesets by affiliation"
+    desc = "The oldest changesets by affiliation and by the last upload"
+
+    def create_report(self, data_source, destdir):
+        if data_source != SCR: return
+        self.result(data_source, destdir)
+
+    def result(self, data_source = None, destdir = None):
+
+        fields = Set([])
+        tables = Set([])
+        filters = Set([])
+
+        fields.add("tr.url as project_name")
+        fields.add("pro.name as author_name")
+        fields.add("org.name as organization")
+        fields.add("i.issue as gerrit_issue_id")
+        fields.add("i.summary as summary")
+        fields.add("i.submitted_on as first_upload")
+        fields.add("t.last_upload as last_upload")
+
+        tables.add("issues i")
+        tables.add("trackers tr")
+        tables.add("people_uidentities puid")
+        tables.add(self.db.identities_db + ".enrollments enr")
+        tables.add(self.db.identities_db + ".organizations org")
+        tables.add(self.db.identities_db + ".profiles pro")
+        tables.add("(select issue_id, max(changed_on) as last_upload from changes where field='status' and new_value='UPLOADED' group by issue_id) t")
+
+        filters.add("t.issue_id = i.id")
+        filters.add("i.id not in (select distinct(issue_id) from changes where field='Code-Review')")
+        filters.add("i.status<>'Abandoned'")
+        filters.add("i.status<>'Merged'")
+        filters.add("tr.id=i.tracker_id")
+        filters.add("i.submitted_by=puid.people_id")
+        filters.add("puid.uuid = enr.uuid")
+        filters.add("i.submitted_on >= enr.start")
+        filters.add("i.submitted_on < enr.end")
+        filters.add("enr.organization_id = org.id")
+        filters.add("puid.uuid = pro.uuid")
+        filters.add("i.summary not like '%WIP%'")
+
+        query = " select " + self.db._get_fields_query(fields)
+        query = query + " from " + self.db._get_tables_query(tables)
+        query = query + " where " + self.db._get_filters_query(filters)
+        query = query + " order by org.name, t.last_upload"
+
+        data = self.db.ExecuteQuery(query)
+
+        # TODO: Hardcoded creation of file
+        createJSON(data, destdir + "/scr-oldest_changesets_by_affiliation.json")
+
+        #Filtering the data to have only 10 entries per organization at most
+        data_df = pandas.DataFrame(data, columns=["gerrit_issue_id", "project_name", "organization", "last_upload", "first_upload", "summary", "author_name"])
+        organizations = pandas.unique(data_df.organization)
+        dataframes = []
+        for organization in organizations:
+            dataframes.append(data_df[data_df["organization"]==organization][1:10])
+
+        filter_orgs = pandas.concat(dataframes)
+        filter_orgs = filter_orgs.to_dict(orient="list")
+        createJSON(filter_orgs, destdir + "/scr-oldest_changesets_by_affiliation.json")
+
+        return filter_orgs
+
+
 if __name__ == '__main__':
 
     filters = MetricFilters("month", "'2011-04-01'", "'2016-01-01'")
@@ -217,6 +292,9 @@ def result(self, data_source = None, destdir = None):
     oldest_changesets = OldestChangesets(dbcon, filters)
     print oldest_changesets.result()
 
+    oldest_changesets = OldestChangesetsByAffiliation(dbcon, filters)
+    oldest_changesets.result(destdir = "./")
+
     most_active = MostActiveChangesetsWaiting4Reviewer(dbcon, filters)
     most_active.result()