nlp2sql 0.2.0rc6__tar.gz → 0.2.0rc7__tar.gz

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/CLAUDE.md

@@ -51,6 +51,8 @@ The codebase follows Clean Architecture (Hexagonal/Ports & Adapters):
 ```
 src/nlp2sql/
 ├── core/           # Business entities (pure Python, no external dependencies)
+│   ├── entities.py            # Query, SQLQuery, DatabaseType
+│   └── database_prompts.py    # SQL dialect hints for AI providers
 ├── ports/          # Interfaces/abstractions (contracts)
 │   ├── ai_provider.py          # AIProviderPort - interface for AI providers
 │   ├── embedding_provider.py   # EmbeddingProviderPort - interface for embeddings
@@ -69,7 +71,8 @@ src/nlp2sql/
 ├── schema/         # Schema management
 │   ├── schema_manager.py       # Coordinates filtering and strategies
 │   ├── schema_analyzer.py      # Scores schema relevance
-│   └── schema_embedding_manager.py  # FAISS vector embeddings
+│   ├── schema_embedding_manager.py  # FAISS vector embeddings
+│   └── example_store.py        # ExampleStore - FAISS-indexed few-shot examples
 ├── config/         # Configuration (Pydantic Settings)
 ├── exceptions/     # Custom exceptions hierarchy
 └── cli.py          # Click-based CLI

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: nlp2sql
-Version: 0.2.0rc6
+Version: 0.2.0rc7
 Summary: Enterprise-ready Natural Language to SQL converter with multi-provider support. Built for production scale (1000+ tables) with Clean Architecture.
 Project-URL: Homepage, https://github.com/luiscarbonel1991/nlp2sql
 Project-URL: Bug Tracker, https://github.com/luiscarbonel1991/nlp2sql/issues
@@ -180,7 +180,33 @@ async def main():
     result2 = await service.generate_sql("Show recent orders")
 ```
 
-### 4. Large Database with Schema Filtering
+### 4. Few-Shot Examples (Improved Accuracy)
+
+```python
+from nlp2sql import ExampleStore, create_embedding_provider, create_and_initialize_service
+
+# Create example store with domain-specific examples
+embedding_provider = create_embedding_provider("openai", api_key=os.getenv("OPENAI_API_KEY"))
+example_store = ExampleStore(embedding_provider=embedding_provider)
+await example_store.add_examples([
+    {
+        "question": "What was the total revenue last month?",
+        "sql": "SELECT SUM(gross_revenue) FROM sales WHERE date >= DATE_TRUNC('month', DATEADD(month, -1, CURRENT_DATE))",
+        "database_type": "redshift",
+        "metadata": {"category": "aggregation"}
+    }
+])
+
+service = await create_and_initialize_service(
+    database_url="redshift://user:pass@host:5439/db",
+    ai_provider="openai",
+    api_key=os.getenv("OPENAI_API_KEY"),
+    example_store=example_store
+)
+result = await service.generate_sql("Show me total sales this quarter")
+```
+
+### 5. Large Database with Schema Filtering
 
 ```python
 from nlp2sql import create_and_initialize_service

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/README.md

@@ -101,7 +101,33 @@ async def main():
     result2 = await service.generate_sql("Show recent orders")
 ```
 
-### 4. Large Database with Schema Filtering
+### 4. Few-Shot Examples (Improved Accuracy)
+
+```python
+from nlp2sql import ExampleStore, create_embedding_provider, create_and_initialize_service
+
+# Create example store with domain-specific examples
+embedding_provider = create_embedding_provider("openai", api_key=os.getenv("OPENAI_API_KEY"))
+example_store = ExampleStore(embedding_provider=embedding_provider)
+await example_store.add_examples([
+    {
+        "question": "What was the total revenue last month?",
+        "sql": "SELECT SUM(gross_revenue) FROM sales WHERE date >= DATE_TRUNC('month', DATEADD(month, -1, CURRENT_DATE))",
+        "database_type": "redshift",
+        "metadata": {"category": "aggregation"}
+    }
+])
+
+service = await create_and_initialize_service(
+    database_url="redshift://user:pass@host:5439/db",
+    ai_provider="openai",
+    api_key=os.getenv("OPENAI_API_KEY"),
+    example_store=example_store
+)
+result = await service.generate_sql("Show me total sales this quarter")
+```
+
+### 5. Large Database with Schema Filtering
 
 ```python
 from nlp2sql import create_and_initialize_service

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/docs/API.md

@@ -80,6 +80,7 @@ print(result['validation'])   # Validation results
 | `api_key` | `str` | Yes | Provider API key |
 | `embedding_provider_type` | `str` | No | Embedding provider: `local` (default), `openai` |
 | `schema_filters` | `dict` | No | Schema filtering options |
+| `example_store` | `ExampleStore` | No | Pre-loaded few-shot examples for better SQL generation |
 
 **Returns:** `dict` with keys `sql`, `confidence`, `explanation`, `validation`
 
@@ -111,6 +112,7 @@ result2 = await service.generate_sql("Show recent orders")
 | `schema_filters` | `dict` | No | Schema filtering options |
 | `embedding_provider_type` | `str` | No | Embedding provider type (see note below) |
 | `database_type` | `DatabaseType` | No | Database type (auto-detected) |
+| `example_store` | `ExampleStore` | No | Pre-loaded few-shot examples for better SQL generation |
 
 **Note on `embedding_provider_type`:** This controls how schema relevance filtering works. With `"local"` or `"openai"`, the system builds a FAISS vector index for semantic search (recommended for databases with 50+ tables). With `None` (default), the system attempts to load local embeddings silently and falls back to text-only matching if `sentence-transformers` is not installed. Text-only matching uses exact name, normalized singular/plural, and token overlap -- effective for small schemas but less accurate for large ones.
 
@@ -141,6 +143,42 @@ result = await service.generate_sql(
 )
 ```
 
+### Few-Shot Examples
+
+Improve SQL accuracy by providing domain-specific examples via `ExampleStore`. Examples are indexed with FAISS for semantic retrieval -- the most relevant examples are automatically included in the LLM prompt.
+
+```python
+from nlp2sql import ExampleStore, create_embedding_provider, create_and_initialize_service
+
+# Create example store
+embedding_provider = create_embedding_provider("openai", api_key="your-key")
+example_store = ExampleStore(embedding_provider=embedding_provider)
+
+# Add domain-specific examples
+await example_store.add_examples([
+    {
+        "question": "What was the total revenue last month?",
+        "sql": "SELECT SUM(gross_revenue) FROM sales WHERE date >= DATE_TRUNC('month', DATEADD(month, -1, CURRENT_DATE))",
+        "database_type": "redshift",
+        "metadata": {"category": "aggregation"}
+    },
+    {
+        "question": "Show orders by country",
+        "sql": "SELECT country, COUNT(*) AS total_orders FROM orders GROUP BY country ORDER BY total_orders DESC",
+        "database_type": "redshift",
+        "metadata": {"category": "aggregation"}
+    }
+])
+
+# Pass to service
+service = await create_and_initialize_service(
+    database_url="redshift://user:pass@host:5439/db",
+    ai_provider="openai",
+    api_key="your-key",
+    example_store=example_store
+)
+```
+
 ### Embedding Providers
 
 ```python

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/docs/ARCHITECTURE.md

@@ -229,7 +229,8 @@ The system has multiple caching layers:
 ```
 src/nlp2sql/
 ├── core/           # Business entities (pure Python, no external dependencies)
-│   └── entities.py         # Query, SQLQuery, DatabaseType
+│   ├── entities.py         # Query, SQLQuery, DatabaseType
+│   └── database_prompts.py # SQL dialect hints for AI providers (centralized)
 ├── ports/          # Interfaces/abstractions
 │   ├── ai_provider.py      # AIProviderPort, QueryContext, QueryResponse
 │   ├── schema_repository.py # SchemaRepositoryPort, TableInfo
@@ -250,7 +251,8 @@ src/nlp2sql/
 ├── schema/         # Schema management
 │   ├── manager.py           # SchemaManager (coordinates strategies)
 │   ├── analyzer.py          # SchemaAnalyzer (scoring, compression)
-│   └── embedding_manager.py # SchemaEmbeddingManager (FAISS + TF-IDF)
+│   ├── embedding_manager.py # SchemaEmbeddingManager (FAISS + TF-IDF)
+│   └── example_store.py     # ExampleStore (FAISS-indexed few-shot examples)
 ├── config/         # Configuration
 │   └── settings.py          # Pydantic Settings (env vars)
 ├── exceptions/     # Custom exception hierarchy

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "nlp2sql"
-version = "0.2.0rc6"
+version = "0.2.0rc7"
 description = "Enterprise-ready Natural Language to SQL converter with multi-provider support. Built for production scale (1000+ tables) with Clean Architecture."
 readme = "README.md"
 license = {text = "MIT"}

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/src/nlp2sql/__init__.py

@@ -11,9 +11,10 @@ from .exceptions import *
 from .factories import RepositoryFactory
 from .ports.embedding_provider import EmbeddingProviderPort
 from .ports.schema_repository import SchemaRepositoryPort
+from .schema.example_store import ExampleStore
 from .services.query_service import QueryGenerationService
 
-__version__ = "0.2.0rc6"
+__version__ = "0.2.0rc7"
 __author__ = "Luis Carbonel"
 __email__ = "devhighlevel@gmail.com"
 
@@ -38,6 +39,8 @@ __all__ = [
     "SQLQuery",
     # Embedding Provider
     "EmbeddingProviderPort",
+    # Example Store
+    "ExampleStore",
     # Configuration
     "settings",
     # Exceptions
@@ -130,9 +133,7 @@ async def create_repository(
         repo = await create_repository("postgresql://user:pass@localhost/db")
         tables = await repo.get_tables()
     """
-    return await RepositoryFactory.create_and_initialize(
-        database_url, schema_name, database_type
-    )
+    return await RepositoryFactory.create_and_initialize(database_url, schema_name, database_type)
 
 
 def create_query_service(
@@ -144,6 +145,7 @@ def create_query_service(
     embedding_provider: Optional[EmbeddingProviderPort] = None,
     embedding_provider_type: Optional[str] = None,
     schema_name: str = "public",
+    example_store: Optional[ExampleStore] = None,
 ) -> QueryGenerationService:
     """
     Create a configured query service instance.
@@ -224,6 +226,7 @@ def create_query_service(
         schema_filters=schema_filters,
         embedding_provider=embedding_provider,
         schema_name=schema_name,
+        example_store=example_store,
     )
 
     return service
@@ -238,6 +241,7 @@ async def create_and_initialize_service(
     embedding_provider: Optional[EmbeddingProviderPort] = None,
     embedding_provider_type: Optional[str] = None,
     schema_name: str = "public",
+    example_store: Optional[ExampleStore] = None,
 ) -> QueryGenerationService:
     """
     Create and initialize a query service with automatic schema loading.
@@ -274,6 +278,7 @@ async def create_and_initialize_service(
         embedding_provider,
         embedding_provider_type,
         schema_name,
+        example_store,
     )
     await service.initialize(database_type)
     return service
@@ -289,6 +294,7 @@ async def generate_sql_from_db(
     embedding_provider: Optional[EmbeddingProviderPort] = None,
     embedding_provider_type: Optional[str] = None,
     schema_name: str = "public",
+    example_store: Optional[ExampleStore] = None,
     **kwargs,
 ) -> Dict[str, Any]:
     """
@@ -346,5 +352,6 @@ async def generate_sql_from_db(
         embedding_provider,
         embedding_provider_type,
         schema_name,
+        example_store,
     )
     return await service.generate_sql(question=question, database_type=database_type, **kwargs)

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/src/nlp2sql/adapters/anthropic_adapter.py

@@ -129,7 +129,11 @@ class AnthropicAdapter(AIProviderPort):
         # Add instructions
         prompt_parts.append("Instructions:")
         prompt_parts.append(f"1. Generate a {context.database_type} SQL query that answers the question")
-        prompt_parts.append("2. Use only the tables and columns from the provided schema")
+        prompt_parts.append(
+            "2. Use EXACT column names as listed in the schema above. "
+            "NEVER abbreviate, shorten, or infer column names. "
+            "If a column is not explicitly listed in the schema, DO NOT use it."
+        )
         prompt_parts.append("3. Follow SQL best practices and optimize for performance")
         prompt_parts.append("4. Include appropriate JOINs, WHERE clauses, and ORDER BY if needed")
         prompt_parts.append("5. Return the response in JSON format with 'sql', 'explanation', and 'confidence' fields")
@@ -151,15 +155,9 @@ Your response must be a valid JSON object with exactly these fields:
 
 Ensure the JSON is properly formatted with no syntax errors. Escape any quotes inside strings properly."""
 
-        database_specific = {
-            "postgres": "You specialize in PostgreSQL syntax and features. Use PostgreSQL-specific functions and syntax when appropriate.",
-            "mysql": "You specialize in MySQL syntax and features. Use MySQL-specific functions and syntax when appropriate.",
-            "sqlite": "You specialize in SQLite syntax and features. Use SQLite-specific functions and syntax when appropriate.",
-            "mssql": "You specialize in SQL Server syntax and features. Use T-SQL syntax when appropriate.",
-            "oracle": "You specialize in Oracle SQL syntax and features. Use Oracle-specific functions and syntax when appropriate.",
-        }
+        from ..core.database_prompts import get_database_hint
 
-        specific_prompt = database_specific.get(database_type.lower(), "")
+        specific_prompt = get_database_hint(database_type)
 
         return f"{base_prompt} {specific_prompt}"
 

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/src/nlp2sql/adapters/gemini_adapter.py

@@ -142,7 +142,11 @@ class GeminiAdapter(AIProviderPort):
         # Add instructions
         prompt_parts.append("Instructions:")
         prompt_parts.append(f"1. Generate a {context.database_type} SQL query that answers the question")
-        prompt_parts.append("2. Use only the tables and columns from the provided schema")
+        prompt_parts.append(
+            "2. Use EXACT column names as listed in the schema above. "
+            "NEVER abbreviate, shorten, or infer column names. "
+            "If a column is not explicitly listed in the schema, DO NOT use it."
+        )
         prompt_parts.append("3. Follow SQL best practices and optimize for performance")
         prompt_parts.append("4. Include appropriate JOINs, WHERE clauses, and ORDER BY if needed")
         prompt_parts.append("5. Return the response in JSON format with 'sql', 'explanation', and 'confidence' fields")
@@ -164,15 +168,9 @@ Your response must be a valid JSON object with exactly these fields:
 
 Ensure the JSON is properly formatted with no syntax errors. Escape any quotes inside strings properly."""
 
-        database_specific = {
-            "postgres": "You specialize in PostgreSQL syntax and features. Use PostgreSQL-specific functions and syntax when appropriate.",
-            "mysql": "You specialize in MySQL syntax and features. Use MySQL-specific functions and syntax when appropriate.",
-            "sqlite": "You specialize in SQLite syntax and features. Use SQLite-specific functions and syntax when appropriate.",
-            "mssql": "You specialize in SQL Server syntax and features. Use T-SQL syntax when appropriate.",
-            "oracle": "You specialize in Oracle SQL syntax and features. Use Oracle-specific functions and syntax when appropriate.",
-        }
+        from ..core.database_prompts import get_database_hint
 
-        specific_prompt = database_specific.get(database_type.lower(), "")
+        specific_prompt = get_database_hint(database_type)
 
         return f"{base_prompt} {specific_prompt}"
 

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/src/nlp2sql/adapters/openai_adapter.py

@@ -163,7 +163,11 @@ class OpenAIAdapter(AIProviderPort):
         # Add instructions
         prompt_parts.append("Instructions:")
         prompt_parts.append(f"1. Generate a {context.database_type} SQL query that answers the question")
-        prompt_parts.append("2. Use only the tables and columns from the provided schema")
+        prompt_parts.append(
+            "2. Use EXACT column names as listed in the schema above. "
+            "NEVER abbreviate, shorten, or infer column names. "
+            "If a column is not explicitly listed in the schema, DO NOT use it."
+        )
         prompt_parts.append("3. Follow SQL best practices and optimize for performance")
         prompt_parts.append("4. Include appropriate JOINs, WHERE clauses, and ORDER BY if needed")
         prompt_parts.append("5. Return the response in JSON format with 'sql', 'explanation', and 'confidence' fields")
@@ -204,15 +208,9 @@ Your response must be a valid JSON object with exactly these fields:
 
 Ensure the JSON is properly formatted with no syntax errors. Escape any quotes inside strings properly."""
 
-        database_specific = {
-            "postgres": "You specialize in PostgreSQL syntax and features. Use PostgreSQL-specific functions and syntax when appropriate.",
-            "mysql": "You specialize in MySQL syntax and features. Use MySQL-specific functions and syntax when appropriate.",
-            "sqlite": "You specialize in SQLite syntax and features. Use SQLite-specific functions and syntax when appropriate.",
-            "mssql": "You specialize in SQL Server syntax and features. Use T-SQL syntax when appropriate.",
-            "oracle": "You specialize in Oracle SQL syntax and features. Use Oracle-specific functions and syntax when appropriate.",
-        }
+        from ..core.database_prompts import get_database_hint
 
-        specific_prompt = database_specific.get(database_type.lower(), "")
+        specific_prompt = get_database_hint(database_type)
 
         return f"{base_prompt} {specific_prompt}"
 

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/src/nlp2sql/adapters/postgres_repository.py

@@ -406,16 +406,18 @@ class PostgreSQLRepository(SchemaRepositoryPort):
                 col_name = row[5]
                 if col_name and col_name not in tables_dict[table_name]["_seen_columns"]:
                     tables_dict[table_name]["_seen_columns"].add(col_name)
-                    tables_dict[table_name]["columns"].append({
-                        "name": col_name,
-                        "type": row[6],
-                        "nullable": row[7] == "YES",
-                        "default": row[8],
-                        "max_length": row[9],
-                        "precision": row[10],
-                        "scale": row[11],
-                        "description": row[12],
-                    })
+                    tables_dict[table_name]["columns"].append(
+                        {
+                            "name": col_name,
+                            "type": row[6],
+                            "nullable": row[7] == "YES",
+                            "default": row[8],
+                            "max_length": row[9],
+                            "precision": row[10],
+                            "scale": row[11],
+                            "description": row[12],
+                        }
+                    )
 
                 # Add foreign key (if present and not seen)
                 fk_column = row[15]
@@ -424,12 +426,14 @@ class PostgreSQLRepository(SchemaRepositoryPort):
                     fk_key = f"{fk_constraint}:{fk_column}"
                     if fk_key not in tables_dict[table_name]["_seen_fks"]:
                         tables_dict[table_name]["_seen_fks"].add(fk_key)
-                        tables_dict[table_name]["foreign_keys"].append({
-                            "column": fk_column,
-                            "ref_table": row[16],
-                            "ref_column": row[17],
-                            "constraint_name": fk_constraint,
-                        })
+                        tables_dict[table_name]["foreign_keys"].append(
+                            {
+                                "column": fk_column,
+                                "ref_table": row[16],
+                                "ref_column": row[17],
+                                "constraint_name": fk_constraint,
+                            }
+                        )
 
             # Convert to TableInfo objects
             tables = []
@@ -438,18 +442,20 @@ class PostgreSQLRepository(SchemaRepositoryPort):
                 del table_data["_seen_columns"]
                 del table_data["_seen_fks"]
 
-                tables.append(TableInfo(
-                    name=table_data["name"],
-                    schema=table_data["schema"],
-                    columns=table_data["columns"],
-                    primary_keys=table_data["primary_keys"],
-                    foreign_keys=table_data["foreign_keys"],
-                    indexes=table_data["indexes"],
-                    row_count=table_data["row_count"],
-                    size_bytes=table_data["size_bytes"],
-                    description=table_data["description"],
-                    last_updated=datetime.now(),
-                ))
+                tables.append(
+                    TableInfo(
+                        name=table_data["name"],
+                        schema=table_data["schema"],
+                        columns=table_data["columns"],
+                        primary_keys=table_data["primary_keys"],
+                        foreign_keys=table_data["foreign_keys"],
+                        indexes=table_data["indexes"],
+                        row_count=table_data["row_count"],
+                        size_bytes=table_data["size_bytes"],
+                        description=table_data["description"],
+                        last_updated=datetime.now(),
+                    )
+                )
 
             logger.info("Tables processed from bulk query", count=len(tables))
             return tables

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/src/nlp2sql/adapters/redshift_adapter.py

@@ -12,6 +12,7 @@ from typing import Any, Dict, List, Optional, Tuple
 
 import psycopg2
 import structlog
+from psycopg2 import sql as psycopg2_sql
 from psycopg2.extras import RealDictCursor
 
 from ..exceptions import SchemaException, SecurityException
@@ -333,9 +334,7 @@ class RedshiftRepository(SchemaRepositoryPort):
 
         try:
             start_time = time.time()
-            rows = await asyncio.to_thread(
-                self._execute_query, bulk_query, (schema,)
-            )
+            rows = await asyncio.to_thread(self._execute_query, bulk_query, (schema,))
 
             elapsed_ms = round((time.time() - start_time) * 1000, 2)
             logger.info("Bulk query completed", rows=len(rows), elapsed_ms=elapsed_ms)
@@ -374,34 +373,38 @@ class RedshiftRepository(SchemaRepositoryPort):
                 col_name = row.get("column_name")
                 if col_name and col_name not in tables_dict[table_name]["_seen_columns"]:
                     tables_dict[table_name]["_seen_columns"].add(col_name)
-                    tables_dict[table_name]["columns"].append({
-                        "name": col_name,
-                        "type": row.get("data_type", ""),
-                        "nullable": row.get("is_nullable") == "YES",
-                        "default": row.get("column_default"),
-                        "max_length": row.get("character_maximum_length"),
-                        "precision": row.get("numeric_precision"),
-                        "scale": row.get("numeric_scale"),
-                        "description": "",
-                    })
+                    tables_dict[table_name]["columns"].append(
+                        {
+                            "name": col_name,
+                            "type": row.get("data_type", ""),
+                            "nullable": row.get("is_nullable") == "YES",
+                            "default": row.get("column_default"),
+                            "max_length": row.get("character_maximum_length"),
+                            "precision": row.get("numeric_precision"),
+                            "scale": row.get("numeric_scale"),
+                            "description": "",
+                        }
+                    )
 
             # Convert to TableInfo objects
             tables = []
             for table_data in tables_dict.values():
                 del table_data["_seen_columns"]
 
-                tables.append(TableInfo(
-                    name=table_data["name"],
-                    schema=table_data["schema"],
-                    columns=table_data["columns"],
-                    primary_keys=table_data["primary_keys"],
-                    foreign_keys=table_data["foreign_keys"],
-                    indexes=table_data["indexes"],
-                    row_count=0,
-                    size_bytes=0,
-                    description=table_data["description"],
-                    last_updated=datetime.now(),
-                ))
+                tables.append(
+                    TableInfo(
+                        name=table_data["name"],
+                        schema=table_data["schema"],
+                        columns=table_data["columns"],
+                        primary_keys=table_data["primary_keys"],
+                        foreign_keys=table_data["foreign_keys"],
+                        indexes=table_data["indexes"],
+                        row_count=0,
+                        size_bytes=0,
+                        description=table_data["description"],
+                        last_updated=datetime.now(),
+                    )
+                )
 
             logger.info("Tables processed from bulk query", count=len(tables))
             return tables
@@ -646,8 +649,11 @@ class RedshiftRepository(SchemaRepositoryPort):
             start_time = time.time()
             conn = self._get_connection()
             try:
-                # Set statement timeout (in milliseconds for Redshift)
                 cursor = conn.cursor()
+                # Set search_path so unqualified table names resolve to our schema
+                cursor.execute(
+                    psycopg2_sql.SQL("SET search_path TO {}, public").format(psycopg2_sql.Identifier(self.schema_name))
+                )
                 cursor.execute(f"SET statement_timeout TO {timeout_seconds * 1000}")
                 cursor.close()
 

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/src/nlp2sql/cli.py

@@ -913,6 +913,7 @@ def cache_clear(clear_all: bool, embeddings: bool, queries: bool, tables: bool):
             if Path(path).exists():
                 if Path(path).is_dir():
                     import shutil
+
                     shutil.rmtree(path)
                 else:
                     Path(path).unlink()
@@ -925,6 +926,7 @@ def cache_clear(clear_all: bool, embeddings: bool, queries: bool, tables: bool):
             if Path(path).exists():
                 if Path(path).is_dir():
                     import shutil
+
                     shutil.rmtree(path)
                 else:
                     Path(path).unlink()

{nlp2sql-0.2.0rc6 → nlp2sql-0.2.0rc7}/src/nlp2sql/config/settings.py

@@ -29,7 +29,7 @@ class Settings(BaseSettings):
 
     # General settings
     app_name: str = "nlp2sql"
-    version: str = "0.2.0rc6"
+    version: str = "0.2.0rc7"
     debug: bool = Field(default=False, validation_alias="NLP2SQL_DEBUG")
     log_level: LogLevel = Field(default=LogLevel.INFO, validation_alias="NLP2SQL_LOG_LEVEL")
 

nlp2sql-0.2.0rc7/src/nlp2sql/core/database_prompts.py

@@ -0,0 +1,48 @@
+"""SQL dialect hints for AI providers.
+
+Each entry provides database-specific guidance that helps AI providers
+generate syntactically correct SQL for the target database. Centralized
+here to avoid duplication across adapters.
+"""
+
+DATABASE_SQL_HINTS: dict[str, str] = {
+    "postgres": (
+        "You specialize in PostgreSQL syntax and features. "
+        "Use PostgreSQL-specific functions and syntax when appropriate."
+    ),
+    "mysql": ("You specialize in MySQL syntax and features. Use MySQL-specific functions and syntax when appropriate."),
+    "sqlite": (
+        "You specialize in SQLite syntax and features. Use SQLite-specific functions and syntax when appropriate."
+    ),
+    "mssql": ("You specialize in SQL Server syntax and features. Use T-SQL syntax when appropriate."),
+    "oracle": (
+        "You specialize in Oracle SQL syntax and features. Use Oracle-specific functions and syntax when appropriate."
+    ),
+    "redshift": (
+        "You specialize in Amazon Redshift SQL syntax for data warehouse analytics. "
+        "CRITICAL RULES: "
+        "1) Use DATEADD(datepart, n, date) instead of INTERVAL arithmetic "
+        "(e.g., DATEADD(month, -1, CURRENT_DATE) not CURRENT_DATE - INTERVAL '1 month'). "
+        "2) Use DATE_TRUNC('unit', date) for truncation - NEVER use TRUNC(date, 'format'). "
+        "3) Use LISTAGG(col, delimiter) WITHIN GROUP (ORDER BY ...) instead of STRING_AGG. "
+        "4) CONCAT() only takes 2 arguments - use || operator to chain more. "
+        "5) Use GETDATE() or CURRENT_DATE instead of NOW(). "
+        "6) DISTINCT ON is not supported - use ROW_NUMBER() OVER (PARTITION BY ... ORDER BY ...) = 1 pattern. "
+        "7) generate_series() is not supported - use CTEs or date tables for sequences. "
+        "8) No ARRAY types or ARRAY_AGG - use SUPER type for semi-structured data. "
+        "9) Use APPROXIMATE COUNT(DISTINCT col) for large cardinality counts. "
+        "10) Avoid SELECT * - only select needed columns for columnar storage efficiency."
+    ),
+}
+
+
+def get_database_hint(database_type: str) -> str:
+    """Get SQL dialect hint for a database type.
+
+    Args:
+        database_type: Database type string (e.g., 'postgres', 'redshift').
+
+    Returns:
+        Dialect-specific prompt hint, or empty string if unknown.
+    """
+    return DATABASE_SQL_HINTS.get(database_type.lower(), "")