Added additional unicode letter points

[nvoxland]

Description

Comments in #1107 listed additional unicode points which cause problems with Japanese chars in sql

---

Dev Handoff Notes (Internal Use)

Links

• Fixed Issue: Stand alone PR / comment from fix #1106 #CORE-3269 degree sign accepted as text #1107
• Test Results: https://github.com/liquibase/liquibase/pull/2329/checks

Testing

• Steps to Reproduce: Create sql with unicode chars in the new ragnes
• Guidance:
  • Impact: Expands what chars don't cause errors when parsed with the SQL parser

Dev Verification

Ensured existing parser-based tests pass. Did not add additional automated tests since we don't check for every possible char

┆Issue is synchronized with this Jira Bug by Unito

[XDelphiGrl]

Thanks for adding this test.

[XDelphiGrl]

The parse error reproduces during update-sql or update with insert statements of strings containing unicode characters into an ntext or nvarchar column. During initial testing, I used inserts with a value containing one problematic unicode character; this did not show the error because there was nothing to 'parse' with a single character.

Special Thanks! Borja Serrano, your comment with a full SQL insert statement in CORE-3269 gave me what I needed to reproduce this bug. Thank you!

---

Example XML Changelog with Inserts

https://app.zenhub.com/files/2019791/e41ea4e8-c4df-430c-9884-0ddef2af0b41/download

• Validate update-sql does not throw a parse error when passed unicode characters. PASS

  • degree (u00b) PASS
  • British pound (u0083) PASS
  • Japanese industrial symbol (u3004) PASS

• Validate updatel does not throw a parse error when passed unicode characters. PASS

  • degree (u00b) PASS
  • British pound (u0083) PASS
  • Japanese industrial symbol (u3004) PASS

---

Test Environment
Liquibase Core: sqlgrammar-unicode/1381/3dd0fa, Pro: master/546/12ffc3
SQL Server 2017
Passing Functional Tests