[CALCITE-3933] Incorrect SQL Emitted for Unicode for Several Dialects #3474
Conversation
There was a problem hiding this comment.
Can you also take a look at https://issues.apache.org/jira/browse/CALCITE-6006?
Does this affect RelToSql?
Is it necessary to update the corresponding tests in #3429?
| @@ -537,7 +537,19 @@ public interface SqlConformance { | |||
| */ | |||
| boolean isValueAllowed(); | |||
|
|
|||
| /** | |||
| /** | |||
There was a problem hiding this comment.
indentation seems to be off here
| @@ -1486,6 +1486,9 @@ private static String toSql(RelNode root, SqlDialect dialect, | |||
| assertThat(toSql(root), isLinux(expectedSql)); | |||
| } | |||
|
|
|||
|
|
|||
| @@ -7133,6 +7136,19 @@ private void checkLiteral2(String expression, String expected) { | |||
| sql(query).ok(expected); | |||
| } | |||
|
|
|||
| @Test void testBigQueryUnicode() { | |||
There was a problem hiding this comment.
would this need a JavaDoc comment indicating the solved issue?
| @Test void testParsingNonIsoCharacter() { | ||
| String sql = "select 'ק' "; | ||
| sql(sql).ok("SELECT u&'\\05e7'"); |
There was a problem hiding this comment.
why remove this test? The u& is valid standard SQL and is still valid Calcite SQL
There was a problem hiding this comment.
the CALCITE-6001 mentioned
'This means that when converting a query like:
select 'ק' as result;
you will get the following the error: Failed to encode 'ק' in character set 'ISO-8859-1'.'
I justed it according to this behavior
|
i see you based this on a couple of my commits, there's some more updated work on #3425 which might be a bit better. Let me know what you think and we can work together. Feel free to review that also^ |
No description provided.