feat: enhance Turkish language support for Islamic terminology

config.py

@@ -1,9 +1,7 @@
 from functools import lru_cache
-
 from pydantic import Field, field_validator
 from pydantic_settings import BaseSettings, SettingsConfigDict
 
-
 class Settings(BaseSettings):
     model_config = SettingsConfigDict(
         env_file=".env",
@@ -16,7 +14,7 @@ class Settings(BaseSettings):
 
     gemini_api_key: str = Field(default="test-key")
 
-    model_name: str = "gemini-1.5-flush"
+    model_name: str = "gemini-1.5-flash"
 
     temperature: float = Field(default=0.7, ge=0, le=2)
     top_p: float = Field(default=0.8, ge=0, le=1)
@@ -49,6 +47,12 @@ class Settings(BaseSettings):
 
     port: int = Field(default=8000, ge=1)
 
+    turkish_support_enabled: bool = Field(default=True)
+    turkish_terminology_db_path: str = Field(default="data/turkish_islamic_terms.json")
+    turkish_ottoman_recognition_enabled: bool = Field(default=True)
+    turkish_arabic_mapping_enabled: bool = Field(default=True)
+    turkish_transliteration_enabled: bool = Field(default=True)
+
     # Respectful Disagreement Enforcement settings
     enforce_respectful_disagreement: bool = True
     disrespectful_language_patterns: list[str] = Field(
@@ -72,9 +76,8 @@ class Settings(BaseSettings):
 
     @field_validator("cors_origins", mode="before")
     @classmethod
-    def parse_cors_origins(cls, value):
-        if isinstance(value, str):
-            return [item.strip() for item in value.split(",") if item.strip()]
+    def parse_cors_origins(cls,value):
+        if isinstance(value,str): return [item.strip() for item in value.split(",") if item.strip()]
         return value
 
     @field_validator("disrespectful_language_patterns", mode="before")
@@ -84,7 +87,5 @@ def parse_disrespectful_language_patterns(cls, value):
             return [item.strip() for item in value.split(",") if item.strip()]
         return value
 
-
 @lru_cache
-def get_settings() -> Settings:
-    return Settings()
+def get_settings()->Settings: return Settings()

query_optimizer.py

@@ -40,6 +40,94 @@
 
 router = APIRouter(prefix="/db-optimizer", tags=["db-optimizer"])
 
+# ---------------------------------------------------------------------------
+# Turkish language support
+# ---------------------------------------------------------------------------
+
+def turkish_lower(text: str) -> str:
+    """Turkish-aware lowercasing that respects the dotted/dotless I distinction."""
+    return text.replace("I", "ı").replace("İ", "i").lower()
+
+
+_TURKISH_SUFFIXES = (
+    "lar", "ler", "dan", "den", "tan", "ten",
+    "yla", "yle", "sin", "sın", "sun", "sün",
+    "dir", "dır", "dur", "dür", "tir", "tır", "tur", "tür",
+    "ken", "ince", "erek", "arak",
+)
+
+TURKISH_ISLAMIC_TERMS: dict[str, tuple[str, ...]] = {
+    "namaz": ("salat", "salât"),
+    "oruç": ("savm", "sawm"),
+    "zekat": ("zakat", "zekât"),
+    "hac": ("hajj", "hacc"),
+    "peygamber": ("nebi", "resul", "rasûl"),
+    "melek": ("malaik", "melekût"),
+    "kuran": ("qur'an", "kur'an", "quran"),
+    "hadis": ("hadith", "hadîs"),
+    "sünnet": ("sunna", "sünnet"),
+    "kelam": ("kalam", "kelâm"),
+    "tasavvuf": ("sufism", "tasawwuf"),
+    "tevhid": ("tawhid", "tevhit"),
+    "ibadet": ("ibada", "ibâdet"),
+    "sevap": ("thawab", "sevâb"),
+    "günah": ("dhanb", "gunah", "günâh"),
+    "helal": ("halal", "helâl"),
+    "haram": ("haram", "harâm"),
+    "dua": ("du'a", "duâ"),
+    "iman": ("iman", "îman"),
+    "islam": ("islam", "islâm"),
+}
+
+_OTTOMAN_AFFIXES = ("ullah", "ül", "il", "el", "al", "ibn", "bin", "zade", "efendi")
+
+
+def turkish_tokenize(text: str) -> list[str]:
+    """Split Turkish text into lowercase tokens, respecting Turkish characters."""
+    return re.findall(r"[a-z0-9çğıöşüâîû']+", turkish_lower(text))
+
+
+def strip_turkish_suffixes(word: str) -> str:
+    """Remove common agglutinative suffixes for a conservative Turkish stem."""
+    stem = word
+    changed = True
+    while changed and len(stem) > 3:
+        changed = False
+        for suffix in _TURKISH_SUFFIXES:
+            if stem.endswith(suffix) and len(stem) - len(suffix) >= 2:
+                stem = stem[: -len(suffix)]
+                changed = True
+                break
+    return stem
+
+
+def normalize_turkish_islamic_term(term: str) -> tuple[str, ...] | str:
+    """Return known Arabic/Ottoman equivalents for a Turkish Islamic term."""
+    normalized = turkish_lower(term).replace("'", "").replace(" ", "")
+    return TURKISH_ISLAMIC_TERMS.get(normalized, turkish_lower(term))
+
+
+def expand_turkish_islamic_query(text: str) -> str:
+    """Expand Turkish Islamic terms with Arabic/Ottoman equivalents for recall."""
+    tokens = turkish_tokenize(text)
+    expanded: list[str] = []
+    for token in tokens:
+        root = strip_turkish_suffixes(token)
+        equivalents = normalize_turkish_islamic_term(root)
+        if isinstance(equivalents, tuple):
+            expanded.append(token)
+            expanded.extend(equivalents)
+        else:
+            expanded.append(token)
+    return " ".join(expanded)
+
+
+def detect_ottoman_turkish_influence(text: str) -> list[str]:
+    """Return Ottoman affixes found in a religious text fragment."""
+    lowered = turkish_lower(text)
+    return [affix for affix in _OTTOMAN_AFFIXES if affix in lowered]
+
+
 # Cross-Surah Reference System
 
 class ReferenceType(str, Enum):
@@ -105,6 +193,9 @@ def references_for(self, surah: int, ayah: int) -> list[CrossReferenceRecord]:
 
 def _tokenize(text: str) -> set[str]:
     return {token for token in re.findall(r"[a-z0-9']+", text.lower()) if
+            token not in _STOPWORDS}
+
+
 # Latency at or above this (milliseconds) counts a query as "slow". Mirrors the
 # issue's "zero queries exceeding 500ms" success criterion.
 DEFAULT_SLOW_QUERY_MS = 500.0
@@ -875,4 +966,4 @@ def _classify_relationship(a: AyahText, b: AyahText, jaccard: float, topic_overl
         return "example"
     if jaccard >= 0.15:
         return "elaboration"
-    return "complement"
+    return "complement"

swahili_processor.py

@@ -51,6 +51,15 @@ def analyze_swahili(text: str) -> SwahiliAnalysisResult:
     )
 
 
+def analyze_turkish(text: str) -> dict:
+    """Analyze Turkish text.
+
+    Placeholder for Turkish language support.  The actual implementation
+    will be added in a dedicated Turkish processor module.
+    """
+    raise NotImplementedError("Turkish language support is under development.")
+
+
 __all__ = [
     "CodeSwitchResult",
     "CodeSwitchType",
@@ -64,6 +73,7 @@ def analyze_swahili(text: str) -> SwahiliAnalysisResult:
     "SwahiliPromptEnhancement",
     "SwahiliToken",
     "analyze_swahili",
+    "analyze_turkish",
     "code_switch_processor",
     "cultural_context_engine",
     "dialect_classifier",