Centralize SpEL security documentation in the reference manual

The Javadoc for EvaluationContext, StandardEvaluationContext, and
SimpleEvaluationContext previously repeated the same detailed
explanation of trusted sources and best-effort restrictions in four
places, making it hard to maintain and to digest.

This commit condenses each class-level warning to a succinct summary
that links to the new "Security Considerations" section of the SpEL
reference documentation, which remains the single, detailed source of
truth introduced in 9b42a40a2a.

See gh-36997

spring-expression/src/main/java/org/springframework/expression/EvaluationContext.java

@@ -38,23 +38,13 @@ import org.jspecify.annotations.Nullable;
  * manually.</li>
  * </ul>
  *
- * <p><strong>WARNING</strong>: Evaluating a SpEL expression obtained from an
- * untrusted source is inherently dangerous, since doing so can effectively
- * grant that source the ability to execute arbitrary code within the
- * application. See the class-level documentation for
- * {@code StandardEvaluationContext} and {@code SimpleEvaluationContext} for
- * details on the trust model applicable to each implementation. Regardless of
- * which {@code EvaluationContext} implementation is in use, any restrictions
- * that an implementation imposes on the SpEL language are provided on a
- * best-effort basis and do not, by themselves, guarantee that expression
- * evaluation is safe. An expression can potentially invoke any property, method,
- * or function reachable via the configured root object, property accessors,
- * index accessors, resolvers, variables, and functions. It is therefore the
- * responsibility of the code that configures an {@code EvaluationContext}
- * &mdash; for example, by supplying a root object or by registering property
- * accessors, resolvers, variables, or functions &mdash; to ensure that none of
- * those reachable objects expose operations that would be dangerous if invoked
- * by an expression from an untrusted source.
+ * <p><strong>WARNING</strong>: Special security considerations apply to the
+ * evaluation of SpEL expressions, particularly those obtained from an untrusted
+ * source. See the
+ * <a href="https://docs.spring.io/spring-framework/reference/core/expressions/evaluation.html#expressions-evaluation-context-security"
+ * >Security Considerations</a> section of the Spring Framework reference
+ * documentation for details, as well as the class-level documentation for
+ * {@code StandardEvaluationContext} and {@code SimpleEvaluationContext}.
  *
  * <p>An {@code EvaluationContext} is designed to be built once and reused across
  * many evaluations, potentially of many different {@link Expression} instances. A

spring-expression/src/main/java/org/springframework/expression/spel/support/SimpleEvaluationContext.java

@@ -88,22 +88,14 @@ import org.springframework.expression.spel.SpelMessage;
  * <p>For more power and flexibility, in particular for internal configuration
  * scenarios, consider using {@link StandardEvaluationContext} instead.
  *
- * <p><strong>WARNING</strong>: {@code SimpleEvaluationContext} takes a
- * best-effort approach to restricting the SpEL language to a subset of its
- * features; however, it cannot guarantee that evaluation of an expression is
- * safe. Evaluating a SpEL expression obtained from an untrusted source is inherently
- * dangerous and should generally be avoided, since doing so can effectively grant
- * that source the ability to execute arbitrary code within the application. Even
- * within the restricted language subset supported by {@code SimpleEvaluationContext},
- * an expression can potentially invoke any property, method, or function reachable
- * via the configured root object, property accessors, method resolvers, variables,
- * and functions. It is therefore the responsibility of the code that configures a
- * {@code SimpleEvaluationContext} &mdash; for example, by supplying a root
- * object or by registering property accessors, method resolvers, variables, or
- * functions &mdash; to ensure that none of those reachable objects expose
- * operations that would be dangerous if invoked by an expression from an
- * untrusted source. For details on what qualifies as a "trusted" source, see
- * {@link StandardEvaluationContext}.
+ * <p><strong>WARNING</strong>: Although {@code SimpleEvaluationContext} restricts
+ * the SpEL language to a subset of its features, that restriction is provided on
+ * a best-effort basis and does not guarantee that evaluation of an expression is
+ * safe; {@code SimpleEvaluationContext} must not be considered safe for evaluating
+ * a SpEL expression obtained from an untrusted source. See the
+ * <a href="https://docs.spring.io/spring-framework/reference/core/expressions/evaluation.html#expressions-evaluation-context-security"
+ * >Security Considerations</a> section of the Spring Framework reference
+ * documentation for details.
  *
  * <p>Because a parsed {@code Expression} may cache accessor and executor state
  * resolved against a particular {@code EvaluationContext} configuration, a parsed

spring-expression/src/main/java/org/springframework/expression/spel/support/StandardEvaluationContext.java

@@ -65,20 +65,12 @@ import org.springframework.util.Assert;
  * opting into several SpEL features as needed by specific use cases.
  *
  * <p><strong>WARNING</strong>: {@code StandardEvaluationContext} exposes the
- * complete SpEL language, including the ability to invoke arbitrary constructors
- * and methods and to read and write arbitrary properties and fields &mdash; all
- * backed by reflection &mdash; as well as the ability to reference beans in an
- * {@code ApplicationContext} via a configured {@link BeanResolver}. For that
- * reason, a {@code StandardEvaluationContext} must <strong>never</strong> be used
- * to evaluate a SpEL expression obtained from an untrusted source. In this context,
- * a "trusted" source is limited to a developer of the application or an
- * administrator who is responsible for configuring or operating the
- * application. Any other source of a SpEL expression &mdash; for example, an
- * expression supplied by an end user or received from an external system
- * &mdash; must be treated as untrusted. Note, however, that evaluating a SpEL
- * expression obtained from an untrusted source is inherently dangerous
- * regardless of the {@code EvaluationContext} implementation in use; see
- * {@link SimpleEvaluationContext} for further details.
+ * complete SpEL language, backed by reflection, and must <strong>never</strong>
+ * be used to evaluate a SpEL expression obtained from an untrusted source. See
+ * the
+ * <a href="https://docs.spring.io/spring-framework/reference/core/expressions/evaluation.html#expressions-evaluation-context-security"
+ * >Security Considerations</a> section of the Spring Framework reference
+ * documentation for details, including the definition of a "trusted" source.
  *
  * <p>Because a parsed {@code Expression} may cache accessor and executor state
  * resolved against a particular {@code EvaluationContext} configuration, a parsed