From 1d032ce80c7bd3da78ea173ded35133b3f7ac4ae Mon Sep 17 00:00:00 2001
From: Bruce Weirdan <weirdan@gmail.com>
Date: Fri, 20 Jun 2025 00:58:26 +0200
Subject: [PATCH] Support namespaces in `query xml` (#16008)

Refs #15992
Refs #14457

# Description

This PR introduces a new switch for `query xml`, `--namespaces`,
and thus allows people to use namespace prefixes in the XPath query
to query namespaced XML.

Example:
```nushell
r#'
   <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
      <rdf:Description rdf:about=""
            xmlns:dc="http://purl.org/dc/elements/1.1/"
         <dc:title>Black-breasted buzzard_AEB_IMG_7158</dc:title>
      </rdf:Description>
   </rdf:RDF>
'# | query xml --namespaces {dublincore: "http://purl.org/dc/elements/1.1/"} "//dublincore:title/text()"
```

# User-Facing Changes

New switch added to `query xml`: `query xml --namespaces {....}`

# Tests + Formatting

Pass.

# After Submitting

IIRC the commands docs on the website are automatically generated, so
nothing to do here.
---
 crates/nu_plugin_query/src/query_xml.rs | 60 +++++++++++++++++++++++--
 1 file changed, 56 insertions(+), 4 deletions(-)

diff --git a/crates/nu_plugin_query/src/query_xml.rs b/crates/nu_plugin_query/src/query_xml.rs
index ae45859ba4..11ed39299e 100644
--- a/crates/nu_plugin_query/src/query_xml.rs
+++ b/crates/nu_plugin_query/src/query_xml.rs
@@ -22,6 +22,12 @@ impl SimplePluginCommand for QueryXml {
     fn signature(&self) -> Signature {
         Signature::build(self.name())
             .required("query", SyntaxShape::String, "xpath query")
+            .named(
+                "namespaces",
+                SyntaxShape::Record(vec![]),
+                "map of prefixes to namespace URIs",
+                Some('n'),
+            )
             .category(Category::Filters)
     }
 
@@ -33,8 +39,9 @@ impl SimplePluginCommand for QueryXml {
         input: &Value,
     ) -> Result<Value, LabeledError> {
         let query: Option<Spanned<String>> = call.opt(0)?;
+        let namespaces: Option<Record> = call.get_flag::<Record>("namespaces")?;
 
-        execute_xpath_query(call, input, query)
+        execute_xpath_query(call, input, query, namespaces)
     }
 }
 
@@ -42,6 +49,7 @@ pub fn execute_xpath_query(
     call: &EvaluatedCall,
     input: &Value,
     query: Option<Spanned<String>>,
+    namespaces: Option<Record>,
 ) -> Result<Value, LabeledError> {
     let (query_string, span) = match &query {
         Some(v) => (&v.item, v.span),
@@ -65,7 +73,11 @@ pub fn execute_xpath_query(
     let package = package.expect("invalid xml document");
 
     let document = package.as_document();
-    let context = Context::new();
+    let mut context = Context::new();
+
+    for (prefix, uri) in namespaces.unwrap_or_default().into_iter() {
+        context.set_namespace(prefix.as_str(), uri.into_string()?.as_str());
+    }
 
     // leaving this here for augmentation at some point
     // build_variables(&arguments, &mut context);
@@ -152,7 +164,8 @@ mod tests {
             span: Span::test_data(),
         };
 
-        let actual = query(&call, &text, Some(spanned_str)).expect("test should not fail");
+        let actual = query(&call, &text, Some(spanned_str), None).expect("test should not fail");
+
         let expected = Value::list(
             vec![Value::test_record(record! {
                 "count(//a/*[posit..." => Value::test_float(1.0),
@@ -181,7 +194,8 @@ mod tests {
             span: Span::test_data(),
         };
 
-        let actual = query(&call, &text, Some(spanned_str)).expect("test should not fail");
+        let actual = query(&call, &text, Some(spanned_str), None).expect("test should not fail");
+
         let expected = Value::list(
             vec![Value::test_record(record! {
                 "count(//*[contain..." => Value::test_float(1.0),
@@ -191,4 +205,42 @@ mod tests {
 
         assert_eq!(actual, expected);
     }
+
+    #[test]
+    fn namespaces_are_used() {
+        let call = EvaluatedCall {
+            head: Span::test_data(),
+            positional: vec![],
+            named: vec![],
+        };
+
+        // document uses `dp` ("document prefix") as a prefix
+        let text = Value::string(
+            r#"<?xml version="1.0" encoding="UTF-8"?><a xmlns:dp="http://example.com/ns"><dp:b>yay</dp:b></a>"#,
+            Span::test_data(),
+        );
+
+        // but query uses `qp` ("query prefix") as a prefix
+        let namespaces = record! {
+            "qp" => Value::string("http://example.com/ns", Span::test_data()),
+        };
+
+        let spanned_str: Spanned<String> = Spanned {
+            item: "//qp:b/text()".to_string(),
+            span: Span::test_data(),
+        };
+
+        let actual =
+            query(&call, &text, Some(spanned_str), Some(namespaces)).expect("test should not fail");
+
+        let expected = Value::list(
+            vec![Value::test_record(record! {
+                "//qp:b/text()" => Value::string("yay", Span::test_data()),
+            })],
+            Span::test_data(),
+        );
+
+        // and yet it should work regardless
+        assert_eq!(actual, expected);
+    }
 }