Convert --setattr values for attributes marked no_update

Attribute Patrams marked no_update never get cloned to Update commands,
and thus never receive the `attribute` flag. This makes their `encode`
method a no-op, which meant they don't get properly encoded when used
with --setattr, making the --setattr fail.

Introduce a `force` argument to encode, which overrides checking
for the attribute flag. Use this in set/add/delattr normalization,
where we know we are dealing with attributes.

https://fedorahosted.org/freeipa/ticket/2616

1 files changed, 6 insertions, 1 deletions

diff --git a/ipalib/plugins/baseldap.py b/ipalib/plugins/baseldap.py
index daf1b07fb..3e7923479 100644
--- a/ipalib/plugins/baseldap.py
+++ b/ipalib/plugins/baseldap.py
@@ -937,7 +937,12 @@ last, after all sets and adds."""),
                     raise errors.ValidationError(name=attr, error=err.error)
                 except errors.ConversionError, err:
                     raise errors.ConversionError(name=attr, error=err.error)
-                value = param.encode(value)
+                # FIXME: We use `force` when encoding because we know this is
+                # an attribute, even if it does not have the `attribute` flag
+                # set. This happens with no_update attributes, which are
+                # not cloned to Update commands. This cloning is where the flag
+                # gets set.
+                value = param.encode(value, force=True)
                 entry_attrs[attr] = value
             else:
                 # unknown attribute: remove duplicite and invalid values