validate row: Oracle RAW Columns Cause Exception

[nj1973]

Oracle RAW columns hold binary data and are sometimes used to store UUID/GUID data. This is not a common pattern but certainly something I've seen in the past.

Below are two test tables, one with a GUID column and another with a GUID column as a primary key.

create table dvt_test.tab_guid (
  id number(2), col_guid raw(16)
);
ALTER TABLE dvt_test.tab_guid ADD PRIMARY KEY (id);
insert into dvt_test.tab_guid values (1,sys_guid());
insert into dvt_test.tab_guid values (2,sys_guid());
commit;

create table dvt_test.tab_guid_pk (
  id raw(16), col_dec number(5)
);
ALTER TABLE dvt_test.tab_guid_pk ADD PRIMARY KEY (id);
insert into dvt_test.tab_guid_pk values (sys_guid(),1);
insert into dvt_test.tab_guid_pk values (sys_guid(),2);
commit;

tab_guid command:

data-validation validate row -sc ora_local -tc ora_local \
-tbls=dvt_test.tab_guid --primary-keys id --concat '*' \
--filter-status fail
...
sqlalchemy.exc.DatabaseError: (cx_Oracle.DatabaseError) ORA-00904: "ENCODE": invalid identifier

tab_guid_pk command:

data-validation validate row -sc ora_local -tc ora_local \
-tbls=dvt_test.tab_guid_pk --primary-keys id --concat '*' \
--filter-status fail
...
  File "pandas/_libs/lib.pyx", line 2859, in pandas._libs.lib.map_infer
  File "/usr/local/google/home/neiljohnson/github/professional-services-data-validator/.venv/lib/python3.10/site-packages/ibis/backends/pandas/execution/strings.py", line 286, in replacer
    return pattern.sub(replacement, x)
TypeError: cannot use a string pattern on a bytes-like object

I haven't tested PostgreSQL but that has a dedicated UUID data type. When working on this we should do the equivalent PostgreSQL tests as above.

[nj1973]

Still fails after Ibis upgrade

[nj1973]

This is also true of Oracle BLOB columns.