Skip to content

Implement Spark-compatible xpath_int #285

Open

Description

@andygrove

Part of #271 and apache/datafusion-comet#7300.

What the function does

xpath_int(xml, path) evaluates an XPath 1.0 expression, converts the result with XPath number() into a
Java Double, and narrows it with Double.intValue(), giving an INT. Both arguments are strings
(collation-aware from 4.0), and path must be foldable. The result is always nullable. Available since Spark
2.0.0.

The number() conversion is in the xpath_double issue. Parsing, DTD and entity handling, namespaces, NULL
handling and errors are in the xpath issue. This issue covers only the double-to-int step, which is Java's
narrowing conversion (int) d:

  • NaN becomes 0;
  • values are truncated toward zero;
  • out-of-range values saturate at Integer.MAX_VALUE / Integer.MIN_VALUE, and so do +/-Infinity.

This matches Rust's d as i32. Nothing raises an error, and ANSI mode has no effect.

Examples

SELECT xpath_int('<a><b>1</b><b>2</b></a>', 'sum(a/b)');
-- 3
SELECT xpath_int('<a>this is not a number</a>', 'a');
-- 0
SELECT xpath_int('<a>this 2 is not a number</a>', 'a');
-- 0
SELECT xpath_int('<a><b>2000000000</b><c>40000000000</c></a>', 'a/b * a/c');
-- 2147483647
SELECT xpath_int('<a>try a boolean</a>', 'a = 10');
-- 0
SELECT xpath_int('<a><b class="odd">1</b><b class="even">2</b><b class="odd">4</b><c>8</c></a>', 'a/b');
-- 1
SELECT xpath_int('<a><b class="odd">1</b><b class="even">2</b><b class="odd">4</b><c>8</c></a>', 'sum(a/*)');
-- 15
SELECT xpath_int('<a><b class="odd">100000</b><b class="even">2</b><b class="odd">4</b><c>8</c></a>', 'sum(a/b[@class="odd"])');
-- 100004

These come from the ExpressionDescription, the Hive udf_xpath_int golden files run by
HiveCompatibilitySuite, and XPathExpressionSuite.

Edge cases to test

  • Saturation at both ends: 8.0E19 gives 2147483647 (above). By the same rule, -1 * a/b * a/c gives
    -2147483648, 1 div 0 gives 2147483647, and -1 div 0 gives -2147483648.
  • Truncation: <a>2.7</a> gives 2 and <a>-2.7</a> gives -2, by code reading. Note that XPath round()
    and floor() give different results, so test xpath_int(x, 'round(a)') as well.
  • NaN gives 0 (non-numeric text, 0 div 0, and the empty node-set, which is NaN by XPath 1.0 rules).
    The usage string says "zero if no match is found, or a match is found but the value is non-numeric".
  • Values just at and beyond 2^31: 2147483647, 2147483648 and -2147483649.
  • NULL or '' arguments give NULL. A non-literal path gives DATATYPE_MISMATCH.NON_FOLDABLE_INPUT.
    Malformed XML and an invalid path raise runtime errors (see the xpath issue).

Differences across Spark versions

None found beyond the shared ones in the xpath issue.

Implementation notes

  • Spark: XPathInt in sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/xml/xpath.scala.
    It calls XPathIntEvaluator (ret.intValue()) in .../xml/XmlExpressionEvalUtils.scala.
  • DataFusion core has no equivalent.
  • Reuse the shared number() evaluator from the xpath and xpath_double issues in
    crates/spark/src/function/xml/, and narrow with as i32. Rust's saturating float-to-int cast matches
    Java's here.
  • Tests: add crates/sqllogictest/test_files/xml/xpath_int.slt and document the function with #[user_doc].

This issue was generated by an LLM (Claude) from the Apache Spark source code and tests. Examples come from Spark's ExpressionDescription examples, test suites and .sql.out golden files, but please verify them against Spark before relying on them.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions