Search first

• I searched and no similar issues were found

Description

Context expressions using grouped comma-separated values like (a,b,c) are parsed incorrectly. parseContextString splits on all commas including those inside parentheses, causing changesets to be executed in the wrong context.

For example a changelog with context="a,b,c" and a changeset with
context="a" will incorrectly execute when running with --context-filter="b",
because (a,b,c) is split into ["(a", "b", "c)"] and b matches the
fragment "b" directly.

Steps To Reproduce

<databaseChangeLog 
  xmlns="http://www.liquibase.org/xml/ns/dbchangelog"
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  context="a,b,c">

    <changeSet id="1" author="example" context="a">
        ...
    </changeSet>

</databaseChangeLog>

1. liquibase update --context-filter="b"
2. the changeset is applied even though the changeset has
   context="a" and we are running with --context-filter="b".

Expected/Desired Behavior

The changeset is not applied because the changeset has context="a" and Liquibase is running with --context-filter="b".

Liquibase Version

5.03

Database Vendor & Version

any

Liquibase Integration

No response

Liquibase Extensions

No response

OS and/or Infrastructure Type/Provider

No response

Additional Context

The bug is in ContextExpression.parseContextString() which uses
StringUtil.splitAndTrim(contexts, ",") to split the expression. This splits
on all commas, including those inside parentheses, corrupting grouped
expressions like (a,b,c) into ["(a", "b", "c)"].

Are you willing to submit a PR?

• I'm willing to submit a PR (Thank you!)