Makes Variable::new validate the variable name

Allows to be sure that the variables name is always valid and easily serializable

CHANGELOG.md

@@ -5,6 +5,7 @@
 
 ## Changed
 - Fixes evaluation of `MONTH()` and `DAY()` functions on the `xsd:date` values.
+- `Variable::new` now validates the variable name.
 
 ## [0.1.1] - 2020-08-14
 

lib/src/sparql/csv_results.rs

@@ -172,7 +172,10 @@ mod tests {
 
     fn build_example() -> QueryResults {
         QuerySolutionIter::new(
-            Rc::new(vec![Variable::new("x"), Variable::new("literal")]),
+            Rc::new(vec![
+                Variable::new_unchecked("x"),
+                Variable::new_unchecked("literal"),
+            ]),
             Box::new(
                 vec![
                     Ok(vec![

lib/src/sparql/mod.rs

@@ -27,7 +27,7 @@ pub use crate::sparql::model::QueryResultsFormat;
 pub use crate::sparql::model::QuerySolution;
 pub use crate::sparql::model::QuerySolutionIter;
 pub use crate::sparql::model::QueryTripleIter;
-pub use crate::sparql::model::Variable;
+pub use crate::sparql::model::{Variable, VariableNameParseError};
 pub use crate::sparql::parser::ParseError;
 pub use crate::sparql::parser::{Query, Update};
 use crate::sparql::plan::{PlanNode, TripleTemplate};

lib/src/sparql/model.rs

@@ -7,6 +7,7 @@ use crate::sparql::error::EvaluationError;
 use crate::sparql::json_results::write_json_results;
 use crate::sparql::xml_results::{read_xml_results, write_xml_results};
 use rand::random;
+use std::error::Error;
 use std::io::{BufRead, Write};
 use std::rc::Rc;
 use std::{fmt, io};
@@ -248,7 +249,7 @@ impl QuerySolutionIter {
     ///
     /// let store = MemoryStore::new();
     /// if let QueryResults::Solutions(solutions) = store.query("SELECT ?s ?o WHERE { ?s ?p ?o }", QueryOptions::default())? {
-    ///     assert_eq!(solutions.variables(), &[Variable::new("s"), Variable::new("o")]);
+    ///     assert_eq!(solutions.variables(), &[Variable::new("s")?, Variable::new("o")?]);
     /// }
     /// # Result::<_,Box<dyn std::error::Error>>::Ok(())
     /// ```
@@ -409,8 +410,9 @@ impl Iterator for QueryTripleIter {
 ///
 /// assert_eq!(
 ///     "?foo",
-///     Variable::new("foo").to_string()
+///     Variable::new("foo")?.to_string()
-/// )
+/// );
+/// # Result::<_,oxigraph::sparql::VariableNameParseError>::Ok(())
 /// ```
 #[derive(Eq, PartialEq, Ord, PartialOrd, Debug, Clone, Hash)]
 pub struct Variable {
@@ -418,8 +420,23 @@ pub struct Variable {
 }
 
 impl Variable {
+    /// Creates a variable name from a unique identifier.
+    ///
+    /// The variable identifier must be valid according to the SPARQL grammar.
+    pub fn new(name: impl Into<String>) -> Result<Self, VariableNameParseError> {
+        let name = name.into();
+        validate_variable_identifier(&name)?;
+        Ok(Self::new_unchecked(name))
+    }
+
+    /// Creates a variable name from a unique identifier without validation.
+    ///
+    /// It is the caller's responsibility to ensure that `id` is a valid blank node identifier
+    /// according to the SPARQL grammar.
+    ///
+    /// [`new`](#method.new) is a safe version of this constructor and should be used for untrusted data.
     #[inline]
-    pub fn new(name: impl Into<String>) -> Self {
+    pub fn new_unchecked(name: impl Into<String>) -> Self {
         Variable { name: name.into() }
     }
 
@@ -435,7 +452,7 @@ impl Variable {
 
     #[inline]
     pub(crate) fn new_random() -> Self {
-        Self::new(format!("{:x}", random::<u128>()))
+        Self::new_unchecked(format!("{:x}", random::<u128>()))
     }
 }
 
@@ -445,3 +462,67 @@ impl fmt::Display for Variable {
         write!(f, "?{}", self.name)
     }
 }
+
+fn validate_variable_identifier(id: &str) -> Result<(), VariableNameParseError> {
+    let mut chars = id.chars();
+    let front = chars.next().ok_or(VariableNameParseError {})?;
+    match front {
+        '0'..='9'
+        | '_'
+        | ':'
+        | 'A'..='Z'
+        | 'a'..='z'
+        | '\u{00C0}'..='\u{00D6}'
+        | '\u{00D8}'..='\u{00F6}'
+        | '\u{00F8}'..='\u{02FF}'
+        | '\u{0370}'..='\u{037D}'
+        | '\u{037F}'..='\u{1FFF}'
+        | '\u{200C}'..='\u{200D}'
+        | '\u{2070}'..='\u{218F}'
+        | '\u{2C00}'..='\u{2FEF}'
+        | '\u{3001}'..='\u{D7FF}'
+        | '\u{F900}'..='\u{FDCF}'
+        | '\u{FDF0}'..='\u{FFFD}'
+        | '\u{10000}'..='\u{EFFFF}' => (),
+        _ => return Err(VariableNameParseError {}),
+    }
+    for c in chars {
+        match c {
+            '0'..='9'
+            | '\u{00B7}'
+            | '\u{00300}'..='\u{036F}'
+            | '\u{203F}'..='\u{2040}'
+            | '_'
+            | 'A'..='Z'
+            | 'a'..='z'
+            | '\u{00C0}'..='\u{00D6}'
+            | '\u{00D8}'..='\u{00F6}'
+            | '\u{00F8}'..='\u{02FF}'
+            | '\u{0370}'..='\u{037D}'
+            | '\u{037F}'..='\u{1FFF}'
+            | '\u{200C}'..='\u{200D}'
+            | '\u{2070}'..='\u{218F}'
+            | '\u{2C00}'..='\u{2FEF}'
+            | '\u{3001}'..='\u{D7FF}'
+            | '\u{F900}'..='\u{FDCF}'
+            | '\u{FDF0}'..='\u{FFFD}'
+            | '\u{10000}'..='\u{EFFFF}' => (),
+            _ => return Err(VariableNameParseError {}),
+        }
+    }
+    Ok(())
+}
+
+/// An error raised during [`Variable`](struct.Variable.html) name validation.
+#[allow(missing_copy_implementations)]
+#[derive(Debug)]
+pub struct VariableNameParseError {}
+
+impl fmt::Display for VariableNameParseError {
+    #[inline]
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        write!(f, "The variable name is invalid")
+    }
+}
+
+impl Error for VariableNameParseError {}

lib/src/sparql/parser.rs

@@ -475,17 +475,17 @@ fn copy_graph(
     to: impl Into<Option<NamedNodeOrVariable>>,
 ) -> GraphUpdateOperation {
     let bgp = GraphPattern::BGP(vec![TriplePattern::new(
-        Variable::new("s"),
+        Variable::new_unchecked("s"),
-        Variable::new("p"),
+        Variable::new_unchecked("p"),
-        Variable::new("o"),
+        Variable::new_unchecked("o"),
     )
     .into()]);
     GraphUpdateOperation::DeleteInsert {
         delete: Vec::new(),
         insert: vec![QuadPattern::new(
-            Variable::new("s"),
+            Variable::new_unchecked("s"),
-            Variable::new("p"),
+            Variable::new_unchecked("p"),
-            Variable::new("o"),
+            Variable::new_unchecked("o"),
             to.into(),
         )],
         using: DatasetSpec::default(),
@@ -984,7 +984,7 @@ parser! {
             if from == to {
                 Vec::new() // identity case
             } else {
-                let bgp = GraphPattern::BGP(vec![TriplePattern::new(Variable::new("s"), Variable::new("p"), Variable::new("o")).into()]);
+                let bgp = GraphPattern::BGP(vec![TriplePattern::new(Variable::new_unchecked("s"), Variable::new_unchecked("p"), Variable::new_unchecked("o")).into()]);
                 vec![copy_graph(from, to)]
             }
         }
@@ -995,7 +995,7 @@ parser! {
             if from == to {
                 Vec::new() // identity case
             } else {
-                let bgp = GraphPattern::BGP(vec![TriplePattern::new(Variable::new("s"), Variable::new("p"), Variable::new("o")).into()]);
+                let bgp = GraphPattern::BGP(vec![TriplePattern::new(Variable::new_unchecked("s"), Variable::new_unchecked("p"), Variable::new_unchecked("o")).into()]);
                 vec![GraphUpdateOperation::Drop { silent, graph: to.clone().into() }, copy_graph(from.clone(), to), GraphUpdateOperation::Drop { silent, graph: from.into() }]
             }
         }
@@ -1006,7 +1006,7 @@ parser! {
             if from == to {
                 Vec::new() // identity case
             } else {
-                let bgp = GraphPattern::BGP(vec![TriplePattern::new(Variable::new("s"), Variable::new("p"), Variable::new("o")).into()]);
+                let bgp = GraphPattern::BGP(vec![TriplePattern::new(Variable::new_unchecked("s"), Variable::new_unchecked("p"), Variable::new_unchecked("o")).into()]);
                 vec![GraphUpdateOperation::Drop { silent, graph: to.clone().into() }, copy_graph(from, to)]
             }
         }
@@ -1606,7 +1606,7 @@ parser! {
             i:iri() { i.into() }
 
         //[108]
-        rule Var() -> Variable = v:(VAR1() / VAR2()) { Variable::new(v) }
+        rule Var() -> Variable = v:(VAR1() / VAR2()) { Variable::new_unchecked(v) }
 
         //[109]
         rule GraphTerm() -> Term =

lib/src/sparql/plan_builder.rs

@@ -734,7 +734,10 @@ impl<E: WriteEncoder<Error = EvaluationError>> PlanBuilder<E> {
                 PatternValue::Variable(variable_key(variables, variable))
             }
             TermOrVariable::Term(Term::BlankNode(bnode)) => {
-                PatternValue::Variable(variable_key(variables, &Variable::new(bnode.as_str())))
+                PatternValue::Variable(variable_key(
+                    variables,
+                    &Variable::new_unchecked(bnode.as_str()),
+                ))
                 //TODO: very bad hack to convert bnode to variable
             }
             TermOrVariable::Term(term) => PatternValue::Constant(self.build_term(term)?),

lib/src/sparql/xml_results.rs

@@ -217,7 +217,7 @@ pub fn read_xml_results(source: impl BufRead + 'static) -> Result<QueryResults,
                             mapping.insert(var.as_bytes().to_vec(), i);
                         }
                         return Ok(QueryResults::Solutions(QuerySolutionIter::new(
-                            Rc::new(variables.into_iter().map(Variable::new).collect()),
+                            Rc::new(variables.into_iter().map(Variable::new).collect::<Result<Vec<_>,_>>().map_err(invalid_data_error)?),
                             Box::new(ResultsIterator {
                                 reader,
                                 buffer: Vec::default(),
@@ -255,7 +255,7 @@ pub fn read_xml_results(source: impl BufRead + 'static) -> Result<QueryResults,
                 State::AfterHead => {
                     return if event.name() == b"results" {
                         Ok(QueryResults::Solutions(QuerySolutionIter::new(
-                            Rc::new(variables.into_iter().map(Variable::new).collect()),
+                            Rc::new(variables.into_iter().map(Variable::new).collect::<Result<Vec<_>,_>>().map_err(invalid_data_error)?),
                             Box::new(empty()),
                         )))
                     } else {

python/src/model.rs

@@ -113,6 +113,7 @@ impl PyObjectProtocol for PyNamedNode {
 ///
 /// :param value: the `blank node ID <https://www.w3.org/TR/rdf11-concepts/#dfn-blank-node-identifier>`_ (if not present, a random blank node ID is automatically generated).
 /// :type value: str, optional
+/// :raises ValueError: if the blank node ID is invalid according to NTriples, Turtle and SPARQL grammars.
 ///
 /// The :py:func:`str` function provides a serialization compatible with NTriples, Turtle and SPARQL:
 ///
@@ -746,6 +747,7 @@ impl PyIterProtocol for PyQuad {
 ///
 /// :param value: the variable name as a string
 /// :type value: str
+/// :raises ValueError: if the variable name is invalid according to the SPARQL grammar.
 ///
 /// The :py:func:`str` function provides a serialization compatible with SPARQL:
 ///
@@ -779,8 +781,10 @@ impl<'a> From<&'a PyVariable> for &'a Variable {
 #[pymethods]
 impl PyVariable {
     #[new]
-    fn new(value: String) -> Self {
+    fn new(value: String) -> PyResult<Self> {
-        Variable::new(value).into()
+        Ok(Variable::new(value)
+            .map_err(|e| ValueError::py_err(e.to_string()))?
+            .into())
     }
 
     /// :return: the variable name

testsuite/src/sparql_evaluator.rs

@@ -466,7 +466,7 @@ impl StaticQueryResults {
                     objects_for_subject_predicate(&dataset, &result_set, rs::RESULT_VARIABLE)
                         .filter_map(|object| {
                             if let Term::Literal(l) = object {
-                                Some(Variable::new(l.value()))
+                                Some(Variable::new_unchecked(l.value()))
                             } else {
                                 None
                             }
@@ -497,7 +497,10 @@ impl StaticQueryResults {
                                                         rs::VALUE,
                                                     ),
                                                 ) {
-                                                    Some((Variable::new(variable.value()), value))
+                                                    Some((
+                                                        Variable::new_unchecked(variable.value()),
+                                                        value,
+                                                    ))
                                                 } else {
                                                     None
                                                 }