SQL has been already extended as a programming language, within many engines that can handle stored procedures (e.g. PL*SQL in Oracle; Transac-SQL in Sybase reused later by MSSQL; similar in Informix, IBM DB2). All this is not new, and all these are true programming languages (i.e. Turing Complete) where "--" is the line comment prefix. This was also the case for precursors of SQL (even if SQL was very long to standardize with a very basic common syntax, but with many dialects).
Before there was Pascal using "{ ... }" for programmes written and encoded with charsets compatible with US-ASCII, or "(* ... *)" digrams for legacy 7/8-bit charsets that did not have curly braces (in C there was a similar system using trigrams, that no one have used since very long, but it was a temporary workaround to add the support for C on very old systems that were still active in the 1970's but were maintained active for much longer than expected, notably systems built for COBOL or FORTRAN that did not need such separator but use a specific "do-nothing" keyword to start a statement, where you could not place multiple statements per line so all the extra "parameters" of the statement were simply ignored and not parsed).COBOL and FORTRAN are still widely used today due to their stability, performance, and security (even if there are much less programmers knowing them, or that will never want to learn about them, even if these languages are not much more profitable today than they were in the past for long term stable jobs).