reviewers: optimize diff data, and creation of PR with advanced default reviewers

rhodecode/api/views/pull_request_api.py

@@ -725,8 +725,12 @@ def create_pull_request(
 
     # recalculate reviewers logic, to make sure we can validate this
     default_reviewers_data = get_default_reviewers_data(
-        owner, source_db_repo,
-        source_commit, target_db_repo, target_commit)
+        owner,
+        source_repo,
+        Reference(source_type, source_name, source_commit_id),
+        target_repo,
+        Reference(target_type, target_name, target_commit_id)
+    )
 
     # now MERGE our given with the calculated
     reviewer_objects = default_reviewers_data['reviewers'] + reviewer_objects

rhodecode/apps/repository/views/repo_review_rules.py

@@ -72,9 +72,11 @@ class RepoReviewRulesView(RepoAppView):
         target_type = request.GET['target_ref_type']
         target_name = request.GET['target_ref_name']
 
-        source_ref = Reference(source_type, source_name, source_commit_id)
-        target_ref = Reference(target_type, target_name, target_commit_id)
-
         review_data = get_default_reviewers_data(
-            current_user, source_repo, source_ref, target_repo, target_ref)
+            current_user,
+            source_repo,
+            Reference(source_type, source_name, source_commit_id),
+            target_repo,
+            Reference(target_type, target_name, target_commit_id)
+        )
         return review_data

rhodecode/model/pull_request.py

@@ -155,12 +155,21 @@ def get_diff_info(
     commits = []
     if get_commit_authors:
         log.debug('Obtaining commit authors from set of commits')
-        commits = target_scm.compare(
+        _compare_data = target_scm.compare(
             target_ref, source_ref, source_scm, merge=True,
-            pre_load=["author", "date", "message", "branch", "parents"])
+            pre_load=["author", "date", "message"]
+        )
 
-        for commit in commits:
-            user = User.get_from_cs_author(commit.author)
+        for commit in _compare_data:
+            # NOTE(marcink): we serialize here, so we don't produce more vcsserver calls on data returned
+            # at this function which is later called via JSON serialization
+            serialized_commit = dict(
+                author=commit.author,
+                date=commit.date,
+                message=commit.message,
+            )
+            commits.append(serialized_commit)
+            user = User.get_from_cs_author(serialized_commit['author'])
             if user and user not in commit_authors:
                 commit_authors.append(user)
 
@@ -170,14 +179,29 @@ def get_diff_info(
         target_commit = source_repo.get_commit(ancestor_id)
 
         for fname, lines in changed_lines.items():
+
             try:
-                node = target_commit.get_node(fname)
+                node = target_commit.get_node(fname, pre_load=["is_binary"])
             except Exception:
+                log.exception("Failed to load node with path %s", fname)
                 continue
 
             if not isinstance(node, FileNode):
                 continue
 
+            # NOTE(marcink): for binary node we don't do annotation, just use last author
+            if node.is_binary:
+                author = node.last_commit.author
+                email = node.last_commit.author_email
+
+                user = User.get_from_cs_author(author)
+                if user:
+                    user_counts[user.user_id] = user_counts.get(user.user_id, 0) + 1
+                author_counts[author] = author_counts.get(author, 0) + 1
+                email_counts[email] = email_counts.get(email, 0) + 1
+
+                continue
+
             for annotation in node.annotate:
                 line_no, commit_id, get_commit_func, line_text = annotation
                 if line_no in lines: