#
#
# add_dir "tests/revert_drop_not_rename"
# 
# add_file "tests/revert_drop_not_rename/__driver__.lua"
#  content [ee408a43814864389f2325d62997362596675dad]
#
============================================================
--- tests/revert_drop_not_rename/__driver__.lua	ee408a43814864389f2325d62997362596675dad
+++ tests/revert_drop_not_rename/__driver__.lua	ee408a43814864389f2325d62997362596675dad
@@ -0,0 +1,26 @@
+
+--
+-- this is a bug report for a weird behaviour with revert
+-- when a dropped (previously renamed) node should be reverted
+-- and the rename has happened on one of the parents of the
+-- node
+-- see http://colabti.de/irclogger/irclogger_log/monotone?date=2008-01-18,Fri&sel=92#l174
+-- for more details
+--
+
+mtn_setup()
+
+check(mtn("mkdir", "old_dir"), 0, false, false)
+writefile("old_dir/a", "blabla")
+addfile("old_dir/a")
+commit()
+
+check(mtn("mv", "old_dir", "new_dir"), 0, false, false)
+check(mtn("drop", "new_dir/a"), 0, false, false)
+
+-- this is the actual command that misbehaves, in the way
+-- that it does not re-create a under new_dir/a, but
+-- old_dir/a, which fails, because old_dir is an old node
+-- and renamed
+check(mtn("revert", "old_dir/a"), 0, false, false)
+xfail(mtn("status"), 0, false, false)