tnakama created CALCITE-7922:
--------------------------------
Summary: Support BigQuery-style raw string literals
Key: CALCITE-7922
URL: https://issues.apache.org/jira/browse/CALCITE-7922
Project: Calcite
Issue Type: New Feature
Components: core
Reporter: tnakama
Support BigQuery-style raw string literals when the parser uses
{{Lex.BIG_QUERY}}.
GoogleSQL accepts a case-insensitive {{r}} prefix on string literals.
Backslashes and the following characters are preserved instead of being
decoded. This is useful for regular expression patterns and other strings
containing backslashes.
h3. Current behavior
The released {{calcite-core:1.42.0}} parser rejects raw string literals. For
example:
{code:java}
import org.apache.calcite.config.Lex;
import org.apache.calcite.sql.parser.SqlParser;
import org.apache.calcite.sql.validate.SqlConformanceEnum;
var config = SqlParser.config()
.withLex(Lex.BIG_QUERY)
.withConformance(SqlConformanceEnum.BIG_QUERY);
SqlParser.create("SELECT r'\\d+'", config).parseQuery();
// Actual: SqlParseException
// Expected: a character literal with value "\\d+"
{code}
h3. Examples and expected values
These examples are SQL source text, not Java string literals.
{noformat}
SELECT r'\d+'; -- backslash + d + plus
SELECT r'\n'; -- backslash + n, not a newline
SELECT R"a\b"; -- a + backslash + b
SELECT r'it\'s'; -- i + t + backslash + quote + s
SELECT r'trailing\\'; -- ends with two backslashes
{noformat}
h3. Proposed initial scope
* Accept both {{r}} and {{R}} prefixes with single- and double-quoted strings.
* Preserve backslashes, including the backslash used before an enclosed quote.
* Reject literals ending with an odd number of backslashes and other invalid
quote delimiters, according to GoogleSQL rules.
* Keep ordinary character literals and other lexical policies unchanged.
* Verify parsed values and their use as function arguments, including regular
expressions.
GoogleSQL also supports triple-quoted raw strings and raw bytes. Those
extensions, along with literal concatenation, can be scoped separately rather
than being prerequisites for the initial single-/double-quoted string support.
h3. Relation to existing issues
CALCITE-7921 tracks decoding backslash escapes in existing ordinary BigQuery
string literals. Raw literals add separate syntax and preserve their contents,
so they are tracked here as a distinct feature.
CALCITE-2406 concerns implicit timestamp conversion of ordinary MySQL strings
and is unrelated to the {{r}} prefix.
h3. Reference
https://docs.cloud.google.com/bigquery/docs/reference/standard-sql/lexical#string_and_bytes_literals
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
