regexp_replace Archives

Publishing new Python Library for JSON & NoSQL

Posted on September 9, 2019September 9, 2019 by SatyakiDe in api, Azure, Data Science, function, json, member function, numpy, Pandas, partition, pattern matching, Python, regexp_replace, regexp_substr, regular expression, replace, select, snippet, String Manipulation, Technology

Hi Guys!

As discussed,

Please find the link of the PyPI package of new enhanced JSON library on Python. This is particularly very useful as I’ve accommodated the following features into it.

distinct
nvl
partition_by
regex_like
regex_replace
regex_substr

All these functions can be used over JSON payload through python. I’ll discuss this in details in my next blog post.

However, I would like to suggest this library that will be handy for NoSQL databases like Cosmos DB. Now, you can quickly implement many of these features such as distinct, partitioning & regular expressions with less effort.

Please find the library URL.

Let me know your feedback on the same.

N.B.: I’ve tested this library both on Windows 10 & Ubuntu 18. And, the python version that I’ve used are Python3.6 & Python3.7.

Till then!

Happy Avenging!

Password Validation Using Regular Expression In Teradata 14 & 15

Posted on December 6, 2014 by SatyakiDe in 14, friends, function, pattern matching, regexp_replace, regexp_similar, regexp_substr, regular expression, sql, String Manipulation, table, Teradata

Today, we’ll be checking one new area where we can implement regular expression to achieve the password validation without involving any kind of Macro, Stored-Proc.

Let’s consider the following conditions to be implemented –

1. Password should contain characters between 6 & 10.

2. One character should be digit.

3. One character should be in upper case letter.

4. There should be at least one special character.

Let’s check the Query & Output –

select seq_no,
       passwd,
       regexp_similar(passwd,'^(?=^([[:graph:]]{6,10})$)(?=.*([[:upper:]]{1,}))(?=.*([[:digit:]]{1,})).*$') as reg_test
from scott.login_det
order by 1;


SEQ_NO	PASSWD	 REG_TEST
-----   -------  --------------
1	hoti         0
2	hotimla	     0
3	hotImla	     0
4	hot@imla     0
5	hoT@imla     0
6	hoT@iml9a    1
7	hoT@iml9a66  0

Similarly, you can add condition of lower case character if you want to make it more complex.

Hope, this will give you another way – to implement the same logic. 🙂

Reverse String using Regexp Functions

Posted on April 8, 2014April 8, 2014 by SatyakiDe in function, regexp_replace

Hi,

Today, we’ll be checking one new area where we can implement regular expression to achieve the same without involving any kind of Macro, Stored-Proc.

Many occasion we may have to parse various kind of strings. Assume that, we need to parse the string in reverse order. Until TD 14.0, you don’t have any easy method to implement the same. Off course, this new method also has some limits. It can only able to reverse couple of characters only. I have checked with 9 characters. It supports that.

So, if you have specific string lengths, then you may also try this solution if you are using TD 14.0. That will be handy.

Let’s check the Query & Output –

SELECT SEQ_NO,
                 SRC_STR,
                 regexp_replace(SRC_STR,'([[:alnum:]]{1,1})([[:alnum:]]{0,1})([[:alnum:]]{0,1})([[:alnum:]]{0,1})([[:alnum:]]{0,1})([[:alnum:]]{0,1})([[:alnum:]]{0,1})([[:alnum:]]{0,1})([[:alnum:]]{0,1})','\9\8\7\6\5\4\3\2\1') AS REV_SRC_STR
FROM WM_CONCAT_TAB
ORDER BY 1;


SEQ_NO	SRC_STR	 REV_SRC_STR
-----   -------  --------------
1	BIRESWAR RAWSERIB
1	TRIDIB	 BIDIRT
1	SUJAY	 YAJUS
1	ANUPAM	 MAPUNA
1	RAM	 MAR
2	PAPU	 UPAP
2	SAYAN	 NAYAS
2	TUNKAI	 IAKNUT
2	BABU	 UBAB
3	ATBIS	 SIBTA
3	SAPMUNDA ADNUMPAS
3	IK	 KI

Hope, this will give you another way – to implement the same logic. 🙂

String Manipulation Advanced Using Teradata 14.0 Regular Expression

Posted on March 28, 2014April 8, 2014 by SatyakiDe in 14, analytic function, function, group_concat, Listagg, pattern matching, regexp_replace, regexp_similar, regexp_substr, regular expression, replace, select, snippet, sql, String Manipulation, table, Teradata, wm_concat

Today, I’ll show couple of very useful functions or logic implemented in Teradata using It’s Regular Expression.

There is two very popular demand comes from most of the Developer across different databases regarding the following two cases –

1. How to Split Comma Separated Values in each rows

2. How to bind separate values in 1 row (Just opposite of Step 1)

2nd Options are very demanding as Cross platform database professional specially Oracle Developers looking for these kind of implementation as Oracle has directly built-in functions to do the same. Those functions are Listagg, wm_concat, group_concat.

Let’s check the solution –

Case 1,

Let’s create the table & prepare some data –

CREATE MULTISET TABLE ETL_DATA.PARSE_STR
  (
     SEQ_NO       INTEGER,
     SRC_STR     VARCHAR(70)
  );
 
CREATE TABLE completed. 0 rows processed. Elapsed Time =  00:00:01.864

Let’s insert some data –

1
2
3

INSERT INTO ETL_DATA.PARSE_STR VALUES(1,'RAM,TRIDIB,ANUPAM,BIRESWAR,SUJAY')
;INSERT INTO ETL_DATA.PARSE_STR VALUES(2,'TUNKAI,SAYAN,BABU,PAPU')
;INSERT INTO ETL_DATA.PARSE_STR VALUES(3,'IK,ATBIS,SAPMUNDA');

Let’s check the value –

SEQ_NO          SRC_STR
------  ----------------------------------
    1   RAM,TRIDIB,ANUPAM,BIRESWAR,SUJAY
    2   TUNKAI,SAYAN,BABU,PAPU
    3   IK,ATBIS,SAPMUNDA

Fine, Now our objective will be split these comma separated values in each lines.

SELECT b.SEQ_NO,
       regexp_substr(b.SRC_STR,'[^,]+',1,day_of_calendar) AS SRC_STR
FROM sys_calendar.calendar ,
     PARSE_STR b
WHERE day_of_calendar BETWEEN 1 AND  (LENGTH(b.SRC_STR) - LENGTH(regexp_replace(b.SRC_STR,'[^A-Z]+','',1,0,'i'))+1 )
ORDER BY 1,2;

And, let’s check the output –

SEQ_NO  SRC_STR
-----   ----------------------
1       ANUPAM
1       BIRESWAR
1       RAM
1       SUJAY
1       TRIDIB
2       BABU
2       PAPU
2       SAYAN
2       TUNKAI
3       ATBIS
3       IK
3       SAPMUNDA

Gr8! I guess, result is coming as per my expectation. 🙂

Case 2(Subsitute Of Listagg, wm_concat, group_concat in Oracle),

This we’ve to do it in Two small Steps for better understanding & performance.

First, let us create another table –

CREATE MULTISET TABLE ETL_DATA.WM_CONCAT_TAB
   (
      SEQ_NO   INTEGER,
      SRC_STR VARCHAR(20)
   );
    
CREATE TABLE completed. 0 rows processed. Elapsed Time =  00:00:01.230

Good. Now we’ll populate some data into this table. We’ll populate data from Step 1 as this will provide the exact data that we’re expecting as input test data for Case 2.

Let’s insert those data –

INSERT INTO ETL_DATA.WM_CONCAT_TAB
SELECT b.SEQ_NO,
       regexp_substr(b.SRC_STR,'[^,]+',1,day_of_calendar) AS SRC_STR
FROM sys_calendar.calendar ,
     PARSE_STR b
WHERE day_of_calendar BETWEEN 1 AND  (LENGTH(b.SRC_STR) - LENGTH(regexp_replace(b.SRC_STR,'[^A-Z]+','',1,0,'i'))+1 );

Let’s check the data –

SEQ_NO  SRC_STR
------  --------------------
1       ANUPAM
1       BIRESWAR
1       RAM
1       SUJAY
1       TRIDIB
2       BABU
2       PAPU
2       SAYAN
2       TUNKAI
3       ATBIS
3       IK
3       SAPMUNDA

As you know in TD we’ve significant restcriction regarding Hirarchical Queries & Recursive Queries. So, In this step we’ll build one relationship like employee & manager in popular employee table. So, if we have that kind of relation then we can easily establish & fit that in TD model.

Let’s create this intermediate table. In this case we’ll go for mapping between current rows with next rows. This is also very useful process. In Oracle, they have LEAD or LAG functions to achieve the same. But, here we’ve to work a little bit more to achive the same.

CREATE MULTISET VOLATILE TABLE VT_SRC_ARRNG
AS
     (
            SELECT SEQ_NO,
                   SRC_STR,
                   MAX(SRC_STR) OVER(
                                        PARTITION BY SEQ_NO
                                        ORDER BY SEQ_NO, SRC_STR
                                        ROWS BETWEEN 1 FOLLOWING AND 1 FOLLOWING 
                                    ) AS PREV_SRC_STR,
                   COUNT(*)  OVER(
                                    PARTITION BY SEQ_NO
                                 ) AS MAX_RECUR_CNT
            FROM WM_CONCAT_TAB
      )
WITH DATA
ON COMMIT
PRESERVE ROWS;
 
CREATE TABLE completed. 0 rows processed. Elapsed Time =  00:00:01.102

Let’s look the output –

SELECT *
FROM VT_SRC_ARRNG
ORDER BY 1,2;
 
 
 
 
SEQ_NO  SRC_STR  PREV_SRC_STR    MAX_RECUR_CNT
-----   -------  --------------- ---------------------
1       ANUPAM      BIRESWAR     5
1       BIRESWAR    RAM          5
1       RAM         SUJAY        5
1       SUJAY       TRIDIB       5
1       TRIDIB      ?            5
2       BABU        PAPU         4
2       PAPU        SAYAN        4
2       SAYAN       TUNKAI       4
2       TUNKAI      ?            4
3       ATBIS       IK           3
3       IK          SAPMUNDA     3
3       SAPMUNDA    ?            3

Fine. From the above VT we can see every Source String has one Previous Source String. Also, we’ve noted down that in each window of SEQ_NO how many levels are there by MAX_RECUR_CNT. We’ll use this column later.

Let’s move to the 2nd & final part –

Let’s aggregate the values based on SEQ_NO & club them with comma –

WITH RECURSIVE WM_CONCAT(SEQ_NO, SRC_STR, PREV_SRC_STR, MAX_RECUR_CNT, LVL,  COMMA_SEP_STR)
AS
     (
        SELECT SEQ_NO,
               SRC_STR,
               PREV_SRC_STR,
               MAX_RECUR_CNT,
               1 AS LVL,
               CAST( '' AS VARCHAR(100)) AS COMMA_SEP_STR
       FROM VT_SRC_ARRNG
       WHERE  PREV_SRC_STR IS NULL
       UNION ALL
       SELECT  b.SEQ_NO,
               b.SRC_STR,
               b.PREV_SRC_STR,
               b.MAX_RECUR_CNT,
               c.LVL+1 AS LVL,
               c.COMMA_SEP_STR||b.SRC_STR||',' AS COMMA_SEP_STR
       FROM VT_SRC_ARRNG b,
               WM_CONCAT c
       WHERE c.SRC_STR =  b.PREV_SRC_STR
     )
SELECT k.SEQ_NO,
       k.AGGR_STR
FROM (               
    SELECT SEQ_NO,
           SRC_STR,
           LVL,
           MAX_RECUR_CNT,
           MIN(CASE
                 WHEN LVL = 1 THEN
                    SRC_STR
               ELSE
                  'ZZZZZ'
               END   ) OVER(
                                 PARTITION BY SEQ_NO
                                 ORDER BY LVL ASC
                           ) ROOT_SRC_STR,
           COMMA_SEP_STR||ROOT_SRC_STR AS AGGR_STR
    FROM WM_CONCAT
    )  k
WHERE k.LVL = k.MAX_RECUR_CNT
ORDER BY 1,2;

Let’s check the output –

SEQ_NO  AGGR_STR
------- ---------------------------
1       SUJAY,RAM,BIRESWAR,ANUPAM,TRIDIB
2       SAYAN,PAPU,BABU,TUNKAI
3       IK,ATBIS,SAPMUNDA

I guess, We’ve done it. 😀

So, You can achieve the same without writing any UDF.

Regular Expression on Teradata 14.0

Posted on February 23, 2014February 23, 2014 by SatyakiDe in 14, analytic function, code, data warehouse, pattern matching, regexp_replace, regexp_similar, regexp_substr, regular expression, Teradata

I’ve been working for more than 8 years in Oracle 10g, 11g & worked significant queries on Regular expressions in various scenario using SQL. It is real handy if you know how to use it & can reduce lots of pain with single SQL. And, the performance will be better compared to the total effort to achieve the same functionalists by using multiple SQL queries or PL/SQL Procedures.

Last couple of years, I’m working on Teradata. And, on some occasion – I was expecting features like these, where I can easily manipulate data with regular expression. I’m pretty excited when I heard that Teradata also introduced Regular Expression from Version 14.0.

As a result, I tried all those features that I think can be handy & useful for various scenarios & followings are the successful queries that I get. There are two occasion, where Teradata partially able to manipulate those strings. I’ve checked the latest Teradata Manual. However, unable to find those solution. So, I’m expecting other forum members can contribute here in order to make this thread useful for every one of us. And, I’ll post here as soon as I get some answers on these partial conversions.

For better understanding, I’ve provided the actual column value & after transformation value of that column in the output. That will help us to grasp it easily – I guess. 🙂

Case 1,

SELECT regexp_replace('SatyakiDe','([[:lower:]]{1,})([[:upper:]]{1,})','\1 \2') AS COL_VAL;
 
COLA             COL_VAL
---------------- ----------------------------------------
SatyakiDe        Satyaki De

Case 2,

select regexp_replace('919047242526','^([[:digit:]]{2})([[:digit:]]{10})','+\1 \2') COL_VAL;
 
COLA         COL_VAL
------------ ---------------
919047255555 +91 9047255555

Case 3,

select regexp_replace('+++C','^([[:punct:]]{2})([[:punct:]]{1})(.*)$','\1\3') COL_VAL;
 
COLA COL_VAL
---- -----
+++C ++C

Case 4,

select initcap(regexp_replace(regexp_substr(' satyaki.de@mail.com','[^@]+'),'(.*)(\.)(.*)','\1 \3')) COL_VAL;
 
COLA                             COL_VAL
-------------------------------- --------------------------------------------------
satyaki.de@mail.com              Satyaki De

Case 5,

select regexp_replace('100011001','([[:digit:]]{3})([[:digit:]]{2})([[:digit:]]{4})','XXX-XX-\3') as COL_VAL;
 
COLA             COL_VAL
---------------- --------------------
100011001        XXX-XX-1001

Case 6,

select regexp_replace('123456789','([[:digit:]]{3})([[:digit:]]{3})([[:digit:]]{3})','\3.\2.\1') as COL_VAL;
 
COLA      COL_VAL
--------- ---------------
123456789 789.456.123

Case 7,

SELECT regexp_replace('satyaki9de0loves3to8work2on2sql0and2bi6tools1','[^0-9]+','',1,0,'i') AS DER_VAL;

COLA                                            DER_VAL
---------------------------------------------   ----------
satyaki1de0loves3to8work2on2sql0and2bi4tools1   1038220241

As you can see, all the characters have filtered out from the string & only numbers are kept here. These sorts of queries are very useful in lots of different business scenarios as well.

So, any extra space may not produce desired result. And, needs to pay attention into these small details.

And, I’ve tested all these queries in the following two versions –

select * from dbcinfo;
 
    InfoKey  InfoData
    -------- ------------------------
1   VERSION  14.10.00.02
2   RELEASE  14.10.00.02
3   LANGUAGE SUPPORT MODE   Standard
 
 
select * from dbcinfo; 
 
    InfoKey  InfoData
    -------- ------------------------
1   VERSION  14.10.01.05
2   RELEASE  14.10.01.04
3   LANGUAGE SUPPORT MODE   Standard

Hope, this will give you much more clarity. 🙂

One more thing, I would like to clarify here – my intention is to describe more features about these regexp_(similar/substr/instr/replace) functions.

I’ve received one question whether these regexp functions available in TD 13 or not in Teradata forum while posting the same article over there.

And, here is my answer to that question –

Regarding version 13,

Let us check whether they have these regexp functions or not –

select * from dbcinfo;
 
    InfoKey  InfoData
    -------- ------------------------
1   VERSION  13.00.00.15
2   RELEASE  13.00.00.15
3   LANGUAGE SUPPORT MODE   Standard
 
 
select * from dbcinfo; 
 
    InfoKey  InfoData
    -------- ------------------------
1   VERSION  13.10.07.12
2   RELEASE  13.10.07.12
3   LANGUAGE SUPPORT MODE   Standard

select regexp_replace('SatyakiDe','^(.*)([[:upper:]]{1,})(.*) $','\1 \2\3') AS COL_VAL;
 
select regexp_replace('SatyakiDe','^(.*)([[:upper:]]{1,})(.*) $','\1 \2\3') AS COL_VAL;
 
select regexp_replace('SatyakiDe','^(.*)([[:upper:]]{1,})(.*) $','\1 \2\3') AS COL_VAL;
                                   $
  *** Failure 3706 Syntax error:  expected something between '(' and the string 'S' keyword.
                      Statement# 1, Info =35
 *** Total elapsed time was 1 second.

Hope this will give adequate clarity to the answer of that above question.

Now, Lets see some other functionality.

REGEXP_SIMILAR has similar functionality like REGEXP_LIKE in Oracle.

Let’s see couple of such cases –

Lets prepare the table with some dummy data –

SELECT * FROM dbc.dbcinfo;
 
    InfoKey  InfoData
    -------- -----------------------
1   VERSION  14.10.01.05
2   RELEASE  14.10.01.04
3   LANGUAGE SUPPORT MODE   Standard
 
 
CREATE MULTISET VOLATILE TABLE TEST_T1
  (
            COL1  VARCHAR(10)
  ) 
ON COMMIT 
PRESERVE ROWS;
 
  INSERT INTO TEST_T1 VALUES('456')
  ;INSERT INTO TEST_T1 VALUES('123x')
  ;INSERT INTO TEST_T1 VALUES('x123')
  ;INSERT INTO TEST_T1 VALUES('y')
  ;INSERT INTO TEST_T1 VALUES('+789')
  ;INSERT INTO TEST_T1 VALUES('-789')
  ;INSERT INTO TEST_T1 VALUES('159-')
  ;INSERT INTO TEST_T1 VALUES('-1-');

Lets check the data now –

SELECT *
FROM TEST_T1;
 
    COL1
1   123x
2   456
3   x123
4   +789
5   -789
6   y
7   159-
8   -1-

Let’s look into the various scenarios now –

Case 1 (Returns Mixed Numbers, Signed Numbers & Non Numbers),

SELECT *
 FROM TEST_T1
 WHERE REGEXP_SIMILAR(COL1,'^[0-9]+$','c')=0;
 
    COL1
    -----
1   123x
2   x123
3   +789
4   -789
5   y
6   159-
7   -1-

Case 2 (Returns Only Unsigned Positive Numbers),

SELECT *
 FROM TEST_T1
 WHERE REGEXP_SIMILAR(COL1,'^[0-9]+$','c')=1;
 
COL1
-----
456

Case 3 (Returns All Numbers including Positive, Negative & unsigned),

SELECT *
FROM TEST_T1
WHERE REGEXP_SIMILAR(COL1,'^[+-]?[0-9]+[+-]?$','c')=1;
 
COL1
-----
456
+789
-789
159-
-1-

Case 4 (Returns Only Non Numbers i.e. Characters),

SELECT *
FROM TEST_T1
WHERE REGEXP_SIMILAR(COL1,'[^0-9]+','c')=1;

COL1
----
y

Hope this will give you some additional idea. 🙂

My objective is to provide basic information to my friends. So, that they can write better SQL in TD while migrating from other popular databases or new developer in TD can get a flavor of this powerful feature & exploit them in all the positive aspect & apply them properly. 😀

Really appreciate your time to read this post.

Regards.

Satyaki De.

Regular Expression In Oracle

Posted on January 3, 2010 by SatyakiDe in OTN, regexp_replace, regexp_substr, regular expression, snippet

From 10g in Oracle there is a significant power is given to all new Oracle Programmer and that is inherited from POSIX. Yes, you guess right – it is regular expression. One of the most powerful features that is missing for quite some time and programmer no need to write big and complex functions or procedures to produce their required result.

In this post i’m not going to discuss about the syntax and all the minute details which require to use it. Rather, i’ll concentrate on some useful snippet on regular expression which might be helpful for you in your case or may require little enhancement from your part.

As i follow OTN, and find these requirements are frequently needed by many programmer. So, why not compiled them in a single platform. If you want to know the basic syntax of regular expression then you can visit this place.

One thing i would like to tell to produce the formatted output – i use CAST function here. Otherwise, it is not required in most of the case where i’ve used in this post. Hope you don’t get confused for this.

Splitting comma-separated values:

scott@ORCL>select * from v$version;

BANNER
--------------------------------------------------------------------------------
Oracle Database 11g Enterprise Edition Release 11.1.0.6.0 - Production
PL/SQL Release 11.1.0.6.0 - Production
CORE    11.1.0.6.0      Production
TNS for 32-bit Windows: Version 11.1.0.6.0 - Production
NLSRTL Version 11.1.0.6.0 - Production

Elapsed: 00:00:00.00
scott@ORCL>
scott@ORCL>
scott@ORCL>with tt
  2  as
  3    (
  4      select 1 id, 'saty,anup,john,sas' src from dual
  5      union all
  6      select 2, 'shank,rajib,papu,sanjit,amit' from dual
  7    )
  8  select k.*
  9  from (
 10             select distinct id,
 11                    regexp_substr(src,'[^,]+',1,level) cooked_src
 12             from tt
 13             connect by level <= (length(src) - length(replace(src,',',''))) + 1
 14      ) k
 15  order by k.id,k.cooked_src;

        ID COOKED_SRC
---------- ----------------------------
         1 anup
         1 john
         1 sas
         1 saty
         2 amit
         2 papu
         2 rajib
         2 sanjit
         2 shank

9 rows selected.

Elapsed: 00:00:00.38
scott@ORCL>
scott@ORCL>

Insert a space inside a joined word:

scott@ORCL>with res
  2  as
  3    (
  4      select 'SatyakiDe' cola from dual
  5      union all
  6      select 'RajibChakroborty' from dual
  7      union all
  8      select 'PranabPaul' from dual
  9    )
 10  select cola,
 11         cast(regexp_replace(cola,'^(.*)([[:upper:]]{1,})(.*)$','\1 \2\3') as varchar2(40)) cooked_cola
 12  from res;

COLA             COOKED_COLA
---------------- ----------------------------------------
SatyakiDe        Satyaki De
RajibChakroborty Rajib Chakroborty
PranabPaul       Pranab Paul

Elapsed: 00:00:00.04
scott@ORCL>

Insert a space inside a joined word:

scott@ORCL>
scott@ORCL>with fer
  2  as
  3    (
  4      select '919047242526' cola from dual
  5      union all
  6      select '919038220261' from dual
  7    )
  8  select cola,
  9         cast(regexp_replace(cola,'^([[:digit:]]{2})([[:digit:]]{10})','+\1 \2') as varchar2(15)) cooked_cola
 10  from fer;

COLA         COOKED_COLA
------------ ---------------
919047242526 +91 9047242526
919038220261 +91 9038220261

Elapsed: 00:00:00.03
scott@ORCL>

Remove one + from the source:

scott@ORCL>with hig
  2  as
  3    (
  4      select '+++C' cola from dual
  5    )
  6  select cola,
  7         cast(regexp_replace(cola,'^([[:punct:]]{2})([[:punct:]]{1})(.*)$','\1\3') as varchar2(5)) cooked_cola
  8  from hig;

COLA COOKE
---- -----
+++C ++C

Elapsed: 00:00:00.02
scott@ORCL>

Extracting number from string:

scott@ORCL>with tep
  2  as
  3    (
  4      select 'satyaki9de0is3a8studious2and2energetic0software2engineer6here1' cola from dual
  5    )
  6  select cola,
  7         cast(regexp_replace(cola,'[^0-9]','') as varchar2(12)) cooked_cola
  8  from tep;

COLA                                                           COOKED_COLA
-------------------------------------------------------------- ------------
satyaki9de0is3a8studious2and2energetic0software2engineer6here1 9038220261

Elapsed: 00:00:00.03
scott@ORCL>

Extracting names from mails:

scott@ORCL>
scott@ORCL>with reg
  2  as
  3    (
  4      select 'satyaki.de@gmail.com' cola from dual
  5      union all
  6      select 'pranab.paul@aol.in' from dual
  7      union all
  8      select 'tuhin.chakroborty@rediffmail.com' from dual
  9      union all
 10      select 'debraj.saha@yahoo.com' from dual
 11    )
 12  select cola,
 13         cast(
 14               initcap(regexp_replace(regexp_substr(cola,'[^@]+'),'(.*)(\.)(.*)','\1 \3')) as varchar2(50)
 15             ) cooked_cola
 16  from reg;

COLA                             COOKED_COLA
-------------------------------- --------------------------------------------------
satyaki.de@gmail.com             Satyaki De
pranab.paul@aol.in               Pranab Paul
tuhin.chakroborty@rediffmail.com Tuhin Chakroborty
debraj.saha@yahoo.com            Debraj Saha

Elapsed: 00:00:00.03
scott@ORCL>

Insert spaces between small & caps letter:

scott@ORCL>
scott@ORCL>with kit
  2  as
  3    (
  4      select 'AbraCaDabra' cola from dual
  5    )
  6  select cola,
  7         cast(
  8                regexp_replace(cola,'([[:lower:]])([[:upper:]])','\1 \2') as varchar2(20)
  9             ) cooked_cola
 10  from kit;

COLA        COOKED_COLA
----------- --------------------
AbraCaDabra Abra Ca Dabra

Elapsed: 00:00:00.02
scott@ORCL>

Masking credit card information:

scott@ORCL>with jol
  2  as
  3    (
  4       select '4567098723560984' cola from dual
  5    )
  6  select cola,
  7         cast(
  8               regexp_replace(cola,'([[:digit:]]{4})([[:digit:]]{4})([[:digit:]]{4})([[:digit:]]{4})','\1-XXXX-XXXX-\4')
  9               as varchar2(20)
 10             ) cooked_cola
 11  from jol;

COLA             COOKED_COLA
---------------- --------------------
4567098723560984 4567-XXXX-XXXX-0984

Elapsed: 00:00:00.02
scott@ORCL>

Convert a number series to ip address & also print this in reverse order:

scott@ORCL>
scott@ORCL>with rev
  2  as
  3    (
  4      select '123456789' cola from dual
  5    )
  6  select cola,
  7         cast(
  8               regexp_replace(cola,'([[:digit:]]{3})([[:digit:]]{3})([[:digit:]]{3})','\3.\2.\1')
  9               as varchar2(15)
 10             ) cooked_cola
 11  from rev;

COLA      COOKED_COLA
--------- ---------------
123456789 789.456.123

Elapsed: 00:00:00.02
scott@ORCL>

Hope you like this.

Very soon i’m going to post another useful snippet on some other topic. Till then best of luck and keep following this blog.

	RAG implementation o… on RAG implementation of LLMs by…
	Realtime reading fro… on Live visual reading using Conv…
	Neural prophet… on Predicting real-time Covid-19…
	Projecting real-time… on Python-based dash framework vi…
	Displaying real-time… on Creating a real-time dashboard…

Category: regexp_replace

Publishing new Python Library for JSON & NoSQL

Like this:

Password Validation Using Regular Expression In Teradata 14 & 15

Like this:

Reverse String using Regexp Functions

Like this:

String Manipulation Advanced Using Teradata 14.0 Regular Expression

Like this:

Regular Expression on Teradata 14.0

Like this:

Regular Expression In Oracle

Like this:

Share this:

Like this:

Share this:

Like this:

Share this:

Like this:

Share this:

Like this:

Share this:

Like this:

Share this:

Like this: