babelfish-for-postgresql / babelfish_extensions / 7711121619
76%
BABEL_5_X_DEV: 77%

LAST BUILD BRANCH: BABEL-5577
DEFAULT BRANCH: BABEL_5_X_DEV
Ran
Jobs 1
Files 113
Run time 10s
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 72.057%. Remained the same
7711121619

push

github

web-flow 
Add oid column to information_schema_tsql.columns for internal use only (#2314)

* Add oid column to information_schema_tsql.columns for internal use only (#2265)

This PR fixes two things

    Fix JOIN conditions which used TABLE_NAME by replacing it with OID instead.
    INFORMATION_SCHEMA_TSQL.COLUMNS should return original name for TABLE_NAME & COLUMN_NAME

(Use case from DMS) Currently if we create a table with mixed case name(quoted) from PG endpoint and use same name to create another table from babelfish endpoint with certain combination of datatypes, system procedure sp_columns_100 starts failing with case not found error for this table. Root cause is the join condition used on information_schema_tsql.columns in sp_columns_100_view. We ended up picking datatype of babelfish endpoint table but typmod of PG created table.
JOIN information_schema_tsql.columns t4 ON (pg_class.relname::sys.nvarchar(128) = t4."TABLE_NAME" AND ext.orig_name = t4."TABLE_SCHEMA")
Here pg_class.relname can be a different table (mixed case) but will get through since we are using server default collation for comparison which is case insensitive (usually).

As a fix,We create a new internal catalog information_schema_tsql.columns_internal which will contain the OID of table as well. The actual information_schema_tsql.columns is just a wrapper view on this. (To mask the OID column).

Issues Resolved: BABEL-4553

Signed-off-by: Tanzeel Khan tzlkhan@amazon.com

40187 of 55771 relevant lines covered (72.06%)

60072.12 hits per line

Jobs
ID Job ID Ran Files Coverage
1 7711121619.1 113
72.06
 GitHub Action Run
Source Files on build 7711121619