Postgres UPSERT on any constraint

Question

In MySQL we can do the following on any constraint violation

INSERT INTO table {setters} ON DUPLICATE KEY UPDATE {setters}

Is there anyway to do this in Postgres ?

INSERT INTO table {setters} ON CONFLICT(*) DO UPDATE {setters}

Note: * = Any Possible Key

Why is this an important feature?

• I want to make this code generic across multiple tables. So, I can't specify the keys as per current Postgres 9.5 syntax (Someone might argue, I could possibly store the keys on all tables and change the keys dynamically. First of all its not elegant and there is the following problem)
• While the application is live, some new unique constraints could be added at run time. So, you can't specify this in your code at Compile time
• Another Alternative is: In your app, you first try to insert and when there is a Duplicate key error, you catch catch it and try to do an update. This is really bad, as this is
  1. Not an atomic operation
  2. The latency between app server and the DB server doubles.
  3. Bulk insert is not possible

Answer 1

Sure. PostgreSQL 9.5+ allows you to do this, in this manner:

The MySQL query:

INSERT INTO tablename (a,b,c) VALUES (1,2,3)
  ON DUPLICATE KEY UPDATE c=c+1;

In PostgreSQL becomes:

INSERT INTO tablename (a, b, c) values (1, 2, 10)
ON CONFLICT (a) DO UPDATE SET c = tablename.c + 1;