filter-repo: track exported and imported refs

We previously nuked all refs not seen in the import using _seen_refs, by
comparing to a full list of original refs.  That works okay when doing a
full repository rewrite, but fails for partial history rewrites.
Further, external rewriting tools that wants to implement a tweak of
this behavior would have had to access the internal _seen_refs field,
but might not be able to rely on _orig_refs if they were doing a partial
history rewrite.  Fix both by tracking both which refs were exported
from the source repository, and which were ultimately imported into the
target repository (they may differ due to pruned commits, renamed
branches or tags, etc.).  Make both available via a new public API,
get_exported_and_imported_refs().

Signed-off-by: Elijah Newren <newren@gmail.com>

git-filter-repo

@@ -852,6 +852,12 @@ class FastExportParser(object):
     self._checkpoint_callback = checkpoint_callback
     self._done_callback       = done_callback
 
+    # Keep track of which refs appear from the export, and which make it to
+    # the import (pruning of empty commits, renaming of refs, and creating
+    # new manual objects and inserting them can cause these to differ).
+    self._exported_refs = set()
+    self._imported_refs = set()
+
     # A list of the branches we've seen, plus the last known commit they
     # pointed to.  An entry in latest_*commit will be deleted if we get a
     # reset for that branch.  These are used because of fast-import's weird
@@ -1076,6 +1082,7 @@ class FastExportParser(object):
     """
     # Parse the Reset
     ref = self._parse_ref_line(b'reset')
+    self._exported_refs.add(ref)
     ignoreme, from_ref = self._parse_optional_parent_ref(b'from')
     if self._currentline == b'\n':
       self._advance_currentline()
@@ -1102,6 +1109,7 @@ class FastExportParser(object):
 
     # Now print the resulting reset
     if not reset.dumped:
+      self._imported_refs.add(reset.ref)
       reset.dump(self._output)
 
   def _parse_commit(self):
@@ -1115,6 +1123,7 @@ class FastExportParser(object):
     # Parse the Commit. This may look involved, but it's pretty simple; it only
     # looks bad because a commit object contains many pieces of data.
     branch = self._parse_ref_line(b'commit')
+    self._exported_refs.add(branch)
     id_ = self._parse_optional_mark()
 
     original_id = None
@@ -1195,6 +1204,7 @@ class FastExportParser(object):
     if not (commit.old_id or commit.id) in _SKIPPED_COMMITS:
       self._latest_commit[branch] = commit.id
     if not commit.dumped:
+      self._imported_refs.add(commit.branch)
       commit.dump(self._output)
 
   def _parse_tag(self):
@@ -1207,6 +1217,7 @@ class FastExportParser(object):
     """
     # Parse the Tag
     tag = self._parse_ref_line(b'tag')
+    self._exported_refs.add(b'refs/tags/'+tag)
     ignoreme, from_ref = self._parse_optional_parent_ref(b'from')
 
     original_id = None
@@ -1235,6 +1246,7 @@ class FastExportParser(object):
     if tag.from_ref:
       # Print out this tag's information
       if not tag.dumped:
+        self._imported_refs.add(b'refs/tags/'+tag.ref)
         tag.dump(self._output)
 
   def _parse_progress(self):
@@ -1304,6 +1316,10 @@ class FastExportParser(object):
   def insert(self, obj):
     assert not obj.dumped
     obj.dump(self._output)
+    if type(obj) == Commit:
+      self._imported_refs.add(obj.branch)
+    elif type(obj) in (Reset, Tag):
+      self._imported_refs.add(obj.ref)
 
   def run(self, input, output):
     """