Rev 3196: Include revisions past searched ones rather than doing very small replies. in http://people.ubuntu.com/~robertc/baz2.0/search-results

[Robert Collins]

At http://people.ubuntu.com/~robertc/baz2.0/search-results

------------------------------------------------------------
revno: 3196
revision-id:robertc at robertcollins.net-20080118040529-7uc3xpwd6yovctoz
parent: robertc at robertcollins.net-20080118034319-tp2b9j8u4h1uxspx
committer: Robert Collins <robertc at robertcollins.net>
branch nick: Repository.get_parent_map
timestamp: Fri 2008-01-18 15:05:29 +1100
message:
  Include revisions past searched ones rather than doing very small replies.
modified:
  bzrlib/smart/repository.py     repository.py-20061128022038-vr5wy5bubyb8xttk-1
=== modified file 'bzrlib/smart/repository.py'
--- a/bzrlib/smart/repository.py	2008-01-18 03:43:19 +0000
+++ b/bzrlib/smart/repository.py	2008-01-18 04:05:29 +0000
@@ -125,10 +125,13 @@
             return error
         # TODO might be nice to start up the search again; but thats not
         # written or tested yet.
-        queried_revs = set(search.get_result().get_keys())
+        client_seen_revs = set(search.get_result().get_keys())
+        # Always include the requested ids.
+        client_seen_revs.difference_update(revision_ids)
         lines = []
         repo_graph = repository.get_graph()
         result = {}
+        queried_revs = set()
         size_so_far = 0
         next_revs = revision_ids
         first_loop_done = False
@@ -140,17 +143,19 @@
                 # adjust for the wire
                 if parents == (_mod_revision.NULL_REVISION,):
                     parents = ()
-                # add parents to the result
-                result[revision_id] = parents
                 # prepare the next query
                 next_revs.update(parents)
-                # Approximate the serialized cost of this revision_id.
-                size_so_far += 2 + len(revision_id) + sum(map(len, parents))
-                # get all the directly asked for parents, and then flesh out to
-                # 64K or so.
-                if first_loop_done and size_so_far > 65000:
-                    next_revs = set()
-                    break
+                if revision_id not in client_seen_revs:
+                    # Client does not have this revision, give it to it.
+                    # add parents to the result
+                    result[revision_id] = parents
+                    # Approximate the serialized cost of this revision_id.
+                    size_so_far += 2 + len(revision_id) + sum(map(len, parents))
+                    # get all the directly asked for parents, and then flesh out to
+                    # 64K or so.
+                    if first_loop_done and size_so_far > 65000:
+                        next_revs = set()
+                        break
             # don't query things we've already queried
             next_revs.difference_update(queried_revs)
             first_loop_done = True