Now tested for all entity types. Closes #7.

icatdump.py

@@ -17,18 +17,14 @@
 #  + The partition of the data into chunks ist static.  It should
 #    rather be dynamic, e.g. chunks should be splitted if the number
 #    of objects in them grows too large.
-#  + The serialization of the following entity types has not yet been
-#    tested: DataCollectionParameter, DatafileParameter,
-#    DatasetParameter, FacilityCycle, InvestigationParameter,
-#    ParameterType, PermissibleStringValue, Publication,
-#    RelatedDatafile, SampleParameter, Shift, Study,
-#    StudyInvestigation.
 #  + The data in the ICAT server must not be modified while this
 #    script is retrieving it.  Otherwise the script may fail or the
 #    dumpfile be inconsistent.  There is not too much that can be done
 #    about this.  A database dump is a snapshot after all.  The
 #    picture will be blurred if the subject is moving while we take
 #    it.
+#  + icatdump fails for Study.  This is a bug in icat.server, see
+#    Issue icatproject/icat.server#155.
 #
 
 import logging

icatingest.py

@@ -17,8 +17,6 @@
 #    id, createId, createTime, modId, and modTime of any objects.
 #    This is by design and cannot be fixed.  As a consequence, access
 #    rules that are based on object ids will not work after a restore.
-#  + Restoring of several entity types has not yet been
-#    tested.  See icatdump.py for a list.
 #  + Dealing with duplicates (option --duplicate) is only supported
 #    for single objects.  If the object contains related objects in
 #    one to many relationships that are to be created at once, the