(camel) branch main updated: CAMEL-24608: camel-langchain4j-agent-api - enforce LanguageGuardrail minLanguageRatio

[fmariani]

This is an automated email from the ASF dual-hosted git repository.

Croway pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/camel.git


The following commit(s) were added to refs/heads/main by this push:
     new d591c2406135 CAMEL-24608: camel-langchain4j-agent-api - enforce 
LanguageGuardrail minLanguageRatio
d591c2406135 is described below

commit d591c24061352149ed34b578312d098332282146
Author: Andrea Cosentino <anco...@gmail.com>
AuthorDate: Thu Sep 3 10:15:08 2026 +0200

    CAMEL-24608: camel-langchain4j-agent-api - enforce LanguageGuardrail 
minLanguageRatio
    
    The minLanguageRatio option was set from the constructor and the builder 
but never
    read in validate(), so it silently did nothing: text that is mostly a 
foreign
    script but contains a single allowed-language character still passed. 
Enforce it by
    computing the fraction of non-whitespace characters that belong to an 
allowed
    language and failing when it is below the configured ratio; also add the 
missing
    getMinLanguageRatio() getter. The check is a no-op at the default 0.0.
    
    Co-Authored-By: Claude Opus 4.8 <nore...@anthropic.com>
    Claude-Session: https://claude.ai/code/session_01Ka4dAcJMpxahMfk3kmG5Ls
    Signed-off-by: Andrea Cosentino <anco...@gmail.com>
---
 .../agent/api/guardrails/LanguageGuardrail.java    | 36 ++++++++++++++++++++++
 .../api/guardrails/LanguageGuardrailTest.java      | 13 ++++++++
 2 files changed, 49 insertions(+)

diff --git 
a/components/camel-ai/camel-langchain4j-agent-api/src/main/java/org/apache/camel/component/langchain4j/agent/api/guardrails/LanguageGuardrail.java
 
b/components/camel-ai/camel-langchain4j-agent-api/src/main/java/org/apache/camel/component/langchain4j/agent/api/guardrails/LanguageGuardrail.java
index bf8eea8c7b34..adc862100cdc 100644
--- 
a/components/camel-ai/camel-langchain4j-agent-api/src/main/java/org/apache/camel/component/langchain4j/agent/api/guardrails/LanguageGuardrail.java
+++ 
b/components/camel-ai/camel-langchain4j-agent-api/src/main/java/org/apache/camel/component/langchain4j/agent/api/guardrails/LanguageGuardrail.java
@@ -212,9 +212,38 @@ public class LanguageGuardrail implements InputGuardrail {
             }
         }
 
+        // Enforce the minimum ratio of allowed-language characters, when 
configured
+        if (minLanguageRatio > 0.0) {
+            double ratio = allowedLanguageRatio(text);
+            if (ratio < minLanguageRatio) {
+                return failure(String.format(
+                        "Message does not meet the minimum allowed-language 
ratio (required %.2f, found %.2f).",
+                        minLanguageRatio, ratio));
+            }
+        }
+
         return success();
     }
 
+    /**
+     * Fraction of the non-whitespace characters that belong to at least one 
allowed language/script. Returns 1.0 when
+     * there are no non-whitespace characters.
+     */
+    private double allowedLanguageRatio(String text) {
+        long total = text.codePoints().filter(cp -> 
!Character.isWhitespace(cp)).count();
+        if (total == 0) {
+            return 1.0;
+        }
+        long allowed = text.codePoints()
+                .filter(cp -> !Character.isWhitespace(cp))
+                .filter(cp -> {
+                    String ch = new String(Character.toChars(cp));
+                    return allowedLanguages.stream().anyMatch(lang -> 
lang.isPresent(ch));
+                })
+                .count();
+        return (double) allowed / total;
+    }
+
     /**
      * @return the set of allowed languages
      */
@@ -229,6 +258,13 @@ public class LanguageGuardrail implements InputGuardrail {
         return new HashSet<>(blockedLanguages);
     }
 
+    /**
+     * @return the minimum required ratio of allowed-language characters (0.0 
disables the check)
+     */
+    public double getMinLanguageRatio() {
+        return minLanguageRatio;
+    }
+
     /**
      * Builder for creating LanguageGuardrail instances.
      */
diff --git 
a/components/camel-ai/camel-langchain4j-agent-api/src/test/java/org/apache/camel/component/langchain4j/agent/api/guardrails/LanguageGuardrailTest.java
 
b/components/camel-ai/camel-langchain4j-agent-api/src/test/java/org/apache/camel/component/langchain4j/agent/api/guardrails/LanguageGuardrailTest.java
index 3c2f65737e9e..565fdd0a8198 100644
--- 
a/components/camel-ai/camel-langchain4j-agent-api/src/test/java/org/apache/camel/component/langchain4j/agent/api/guardrails/LanguageGuardrailTest.java
+++ 
b/components/camel-ai/camel-langchain4j-agent-api/src/test/java/org/apache/camel/component/langchain4j/agent/api/guardrails/LanguageGuardrailTest.java
@@ -153,4 +153,17 @@ class LanguageGuardrailTest {
 
         
assertTrue(guardrail.getBlockedLanguages().contains(LanguageGuardrail.Language.CYRILLIC));
     }
+
+    @Test
+    void testMinLanguageRatioIsEnforced() {
+        LanguageGuardrail guardrail = LanguageGuardrail.builder()
+                .allowedLanguages(LanguageGuardrail.Language.LATIN_SCRIPT)
+                .minLanguageRatio(0.8)
+                .build();
+
+        // "Hello 你好世界": 5 Latin characters out of 9 non-whitespace ~= 0.56, 
below the 0.8 threshold
+        assertFalse(guardrail.validate(UserMessage.from("Hello 
你好世界")).isSuccess());
+        // All-Latin text meets the ratio
+        assertTrue(guardrail.validate(UserMessage.from("Hello 
world")).isSuccess());
+    }
 }