[PHOENIX-611] regexp_substr(), is displaying wrong results, if starting and end character is same and vice versa also.

• Type: Task
• Status: Resolved
• Resolution: Fixed
• Labels: invalid
• old issue number: 480
• Assignee: Unassigned
• Reporter: Y. SREENIVASULU REDDY
• Created: 10/Mar/14 23:47
• Updated: 16/Mar/14 07:19
• Resolved: 16/Mar/14 07:19

Description

1. create a table and insert some amount of data into the table.
2. query to test regexp_substr(). in below manner.

but expected result is not displaying.

0: jdbc:phoenix:10.18.40.69,10.18.40.130> select *,regexp_substr(name,'[^s]+',1) from sqltable limit 5;
-----------------------------------------------------------

NO

NAME

ADDR

DEPT

REGEXP_SUBSTR(NAME,'[^s]+',1)

-----------------------------------------------------------

0	jruls	HYD	30	jrul
1	hrjcu	MUM	30	hrjcu
2	yqtrv	BLR	30	yqtrv
3	satax	HYD	30	atax
4	jjcvw	BLR	40	jjcvw

-----------------------------------------------------------

here for record 3 expected the value is null. but it is displaying "atax"

similarly in reverse order
0: jdbc:phoenix:10.18.40.69,10.18.40.130> select *,regexp_substr(name,'[^s]+',-1) from sqltable limit 5;
---------------------------------------------------------------------------------------------------------------

NO

NAME

ADDR

DEPT

REGEXP_SUBSTR(NAME,'[^s]+',com.salesforce.phoenix.parse.MultiplyParseNode[1, -1])

---------------------------------------------------------------------------------------------------------------

0	jruls	HYD	30	null
1	hrjcu	MUM	30	u
2	yqtrv	BLR	30	v
3	satax	HYD	30	x
4	jjcvw	BLR	40	w

---------------------------------------------------------------------------------------------------------------
here in reverse order value for 3.