M17: ground knowledge and integration evidence

backend/app/services/knowledge/ragcore.py

@@ -17,9 +17,10 @@ _GROUNDED_ANSWERABILITY = {"answerable", "partially_answerable"}
 _LEAD_ANSWER_TEMPLATE = {
     "en-GB": 'Per "{title}": {excerpt}',
     "nl-BE": 'Volgens "{title}": {excerpt}',
-    "fr-BE": 'Selon « {title} » : {excerpt}',
+    "fr-BE": "Selon « {title} » : {excerpt}",
 }
 _DEFAULT_LANGUAGE = "en-GB"
+_MAX_SOURCE_CARDS = 3
 
 _DOMAIN_CONCEPTS: dict[str, tuple[str, ...]] = {
     "damage": ("damage", "damaged", "schade", "beschadigd", "dommage", "endommagé"),
@@ -48,14 +49,26 @@ def _question_concepts(question: str) -> set[str]:
 
 
 def _deduplicate_sources(sources: list[SourceCard]) -> list[SourceCard]:
-    seen: set[tuple[str, str]] = set()
+    """Collapse duplicate chunks and re-uploaded document versions.
+
+    RAGcore document/version UUIDs change across uploads, so they are not useful
+    deduplication keys. Human-visible citation identity is the normalized title,
+    section and excerpt.
+    """
+    seen: set[tuple[str, str, str]] = set()
     unique: list[SourceCard] = []
     for source in sources:
-        key = (source.document_id, source.section)
+        key = (
+            source.title.strip().casefold(),
+            source.section.strip().casefold(),
+            " ".join(source.excerpt.split()).casefold(),
+        )
         if key in seen:
             continue
         seen.add(key)
         unique.append(source)
+        if len(unique) == _MAX_SOURCE_CARDS:
+            break
     return unique
 
 
@@ -193,6 +206,7 @@ class RAGcoreKnowledgeProvider:
                 )
                 for citation in citations.values()
             ]
+            sources = _deduplicate_sources(sources)
             answerability = body.get("answerability", "not_answerable")
             is_grounded = answerability in _GROUNDED_ANSWERABILITY and sources
             evidence_state: EvidenceState = "grounded" if is_grounded else "insufficient"