- 3,714,551 Users
- 2,242,575 Discussions
- 7,844,931 Comments
Forum Stats
Discussions
Categories
- 12 Data
- 362.2K Big Data Appliance
- 6 Data Science
- 1.5K Databases
- 440 General Database Discussions
- 3.7K Java and JavaScript in the Database
- 22 Multilingual Engine
- 487 MySQL Community Space
- 3 NoSQL Database
- 7.6K Oracle Database Express Edition (XE)
- 2.7K ORDS, SODA & JSON in the Database
- 415 SQLcl
- 42 SQL Developer Data Modeler
- 184.8K SQL & PL/SQL
- 21K SQL Developer
- 1.8K Development
- 3 Developer Projects
- 32 Programming Languages
- 135K Development Tools
- 7 DevOps
- 3K QA/Testing
- 231 Java
- 4 Java Learning Subscription
- 10 Database Connectivity
- 65 Java Community Process
- Java 25
- 8 Java APIs
- 141.1K Java Development Tools
- 5 Java EE (Java Enterprise Edition)
- 153K Java Essentials
- 134 Java 8 Questions
- 86.2K Java Programming
- 270 Java Lambda MOOC
- 65.1K New To Java
- 1.7K Training / Learning / Certification
- 13.8K Java HotSpot Virtual Machine
- 10 Java SE
- 13.8K Java Security
- 3 Java User Groups
- 22 JavaScript - Nashorn
- 18 Programs
- 118 LiveLabs
- 28 Workshops
- 9 Software
- 3 Berkeley DB Family
- 3.5K JHeadstart
- 5.7K Other Languages
- 2.3K Chinese
- 3 Deutsche Oracle Community
- 10 Español
- 1.9K Japanese
- 2 Portuguese
Add an optional argument c after a comma in the TRIM(s) function

Hi!
I compared string functions in 5 different RDBMS: PostgreSQL, MySQL, Oracle, SQL Server and SQLite, and also in the ANSI SQL Standard. There are large differences among different RDBMS, and I propose to reduce them a little bit.
Particularly, I propose to add an optional argument c
after a comma in the TRIM(s)
function (the space ' '
by default).
Explanation: We can now call the following functions in Oracle:
LTRIM(s [,c]) RTRIM(s [,c]) TRIM (s)
As you see, we can pass the argument c
after the comma in the LTRIM
and RTRIM
functions, but not in the TRIM
function, which is strange. For the TRIM
function we have to use the SQL ANSI Standard syntax: TRIM(c FROM s)
. On the other hand, in PostgreSQL and SQLite calling TRIM(s, c)
is possible. So this difference between different RDBMS and the difference in Oracle itself between the LTRIM/RTRIM(s [,c])
and TRIM(s)
functions seem to be unjustified. Or maybe somebody knows a justification?
I also proposed this change: https://community.oracle.com/tech/developers/discussion/4477919/proposal-add-the-substring-alias-for-the-substr-function/.
The changes proposed by me for other RDBMS are listed here: https://github.com/iwis/Scalar-SQL-functions/blob/main/Proposed_changes.md
Comments
-
I always found the implementation of TRIM a bit strange. Bringing it in line with LTRIM and RTRIM is a very good idea.