What happens when you create an object in beeline ? do you get the same error ?

I got syntax error in beeline, Please refer screenshot

@umesh: can you try 1 more time after you fix the error - you have entered "storad" as textfile instead of stored.

Fix that syntax and then try again in beeline

Sorry, create table syntax was wrong.
No exception in the beeline if we create table with numeric characters -  screenshot  

Observations for : Hortonworks , Apache hive, Amazon EMR, MapR

Table is created with numeric characters but when we expand the columns folder in tree node we get below error - screenshot

Error while compiling statement: FAILED: ParseException line 1:19 cannot recognize input near 'FORMATTED' '123' '' in describe statement

Exception : 

Reference Links : 

Hive : https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-Create/Drop/TruncateTable

Impala : http://www.cloudera.com/documentation/enterprise/latest/topics/impala_identifiers.html#identifiers

I found this in the above mention hive link : 

In Hive 0.13 and later, column names can contain any Unicode character (see HIVE-6013). Any column name that is specified within backticks (`) is treated literally. Within a backtick string, use double backticks (``) to represent a backtick character. Backtick quotation also enables the use of reserved keywords for table and column identifiers.

ADS observation : when we create a table with numeric characters, In Preview SQL tab query is generated with backticks but when we expand the  columns we get error as mention in the above comment. 

Also observe the SQL log in the screenshot

Pls provide a regression analysis matrix. Discuss w/ @shilpa or @tariq if you have any questions on how to generate such a matrix.

Y : Issue Observed

For Other Distributions : 

Hi Sachin, This will take a couple of hours. Thanks, Tom

@tom: pls explain the nature of the fix. Is there a bug in our scripting / extraction that needs to be fixed?

Hi Sachin,

The problem is that different distributions allow different file name characters. Currently for all distributions, we backtic our creates and alters which allow names like 123 or _tom to be used. When we display columns in the schema tree we don't use backtics with our describe's and this is causing exceptions with certain names. I added backtics to the describes and that fixes the problem across all distributions. 

return "DESCRIBE FORMATTED " + cp.withQuotedIdentifier(tablename);

Thanks, Tom

Patch in email.

svn #53630 - Added code to backtic the describe statement.

Hi QA Team,

Please also test with the distribution Apache drivers.

Thanks, Tom

Verified in ADS 19.0.0-dev-20. Verified for this comment also.

Verified in ADS v19.0.0-alpha-8