Skip to content

Postgres

PostgresKVStore #

Bases: BaseKVStore

Postgres Key-Value store.

Parameters:

Name Type Description Default
connection_string str

psycopg2 connection string

 required
async_connection_string str

asyncpg connection string

 required
table_name str

table name

 required
schema_name Optional[str]

schema name

 'public'
perform_setup Optional[bool]

perform table setup

 True
debug Optional[bool]

debug mode

 False
use_jsonb Optional[bool]

use JSONB data type for storage

 False
Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
class PostgresKVStore(BaseKVStore):
    """Postgres Key-Value store.

    Args:
        connection_string (str): psycopg2 connection string
        async_connection_string (str): asyncpg connection string
        table_name (str): table name
        schema_name (Optional[str]): schema name
        perform_setup (Optional[bool]): perform table setup
        debug (Optional[bool]): debug mode
        use_jsonb (Optional[bool]): use JSONB data type for storage
    """

    connection_string: str
    async_connection_string: str
    table_name: str
    schema_name: str
    perform_setup: bool
    debug: bool
    use_jsonb: bool

    def __init__(
        self,
        connection_string: str,
        async_connection_string: str,
        table_name: str,
        schema_name: str = "public",
        perform_setup: bool = True,
        debug: bool = False,
        use_jsonb: bool = False,
    ) -> None:
        try:
            import asyncpg  # noqa
            import psycopg2  # noqa
            import sqlalchemy
            import sqlalchemy.ext.asyncio  # noqa
        except ImportError:
            raise ImportError(
                "`sqlalchemy[asyncio]`, `psycopg2-binary` and `asyncpg` "
                "packages should be pre installed"
            )

        table_name = table_name.lower()
        schema_name = schema_name.lower()
        self.connection_string = connection_string
        self.async_connection_string = async_connection_string
        self.table_name = table_name
        self.schema_name = schema_name
        self.perform_setup = perform_setup
        self.debug = debug
        self.use_jsonb = use_jsonb
        self._is_initialized = False

        from sqlalchemy.orm import declarative_base

        # sqlalchemy model
        self._base = declarative_base()
        self._table_class = get_data_model(
            self._base,
            table_name,
            schema_name,
            use_jsonb=use_jsonb,
        )

    @classmethod
    def from_params(
        cls,
        host: Optional[str] = None,
        port: Optional[str] = None,
        database: Optional[str] = None,
        user: Optional[str] = None,
        password: Optional[str] = None,
        table_name: str = "kvstore",
        schema_name: str = "public",
        connection_string: Optional[str] = None,
        async_connection_string: Optional[str] = None,
        perform_setup: bool = True,
        debug: bool = False,
        use_jsonb: bool = False,
    ) -> "PostgresKVStore":
        """Return connection string from database parameters."""
        conn_str = (
            connection_string
            or f"postgresql+psycopg2://{user}:{password}@{host}:{port}/{database}"
        )
        async_conn_str = async_connection_string or (
            f"postgresql+asyncpg://{user}:{password}@{host}:{port}/{database}"
        )
        return cls(
            connection_string=conn_str,
            async_connection_string=async_conn_str,
            table_name=table_name,
            schema_name=schema_name,
            perform_setup=perform_setup,
            debug=debug,
            use_jsonb=use_jsonb,
        )

    @classmethod
    def from_uri(
        cls,
        uri: str,
        table_name: str = "kvstore",
        schema_name: str = "public",
        perform_setup: bool = True,
        debug: bool = False,
        use_jsonb: bool = False,
    ) -> "PostgresKVStore":
        """Return connection string from database parameters."""
        params = params_from_uri(uri)
        return cls.from_params(
            **params,
            table_name=table_name,
            schema_name=schema_name,
            perform_setup=perform_setup,
            debug=debug,
            use_jsonb=use_jsonb,
        )

    def _connect(self) -> Any:
        from sqlalchemy import create_engine
        from sqlalchemy.ext.asyncio import AsyncSession, create_async_engine
        from sqlalchemy.orm import sessionmaker

        self._engine = create_engine(self.connection_string, echo=self.debug)
        self._session = sessionmaker(self._engine)

        self._async_engine = create_async_engine(self.async_connection_string)
        self._async_session = sessionmaker(self._async_engine, class_=AsyncSession)

    def _create_schema_if_not_exists(self) -> None:
        with self._session() as session, session.begin():
            from sqlalchemy import text

            # Check if the specified schema exists with "CREATE" statement
            check_schema_statement = text(
                f"SELECT schema_name FROM information_schema.schemata WHERE schema_name = '{self.schema_name}'"
            )
            result = session.execute(check_schema_statement).fetchone()

            # If the schema does not exist, then create it
            if not result:
                create_schema_statement = text(
                    f"CREATE SCHEMA IF NOT EXISTS {self.schema_name}"
                )
                session.execute(create_schema_statement)

            session.commit()

    def _create_tables_if_not_exists(self) -> None:
        with self._session() as session, session.begin():
            self._base.metadata.create_all(session.connection())

    def _initialize(self) -> None:
        if not self._is_initialized:
            self._connect()
            if self.perform_setup:
                self._create_schema_if_not_exists()
                self._create_tables_if_not_exists()
            self._is_initialized = True

    def put(
        self,
        key: str,
        val: dict,
        collection: str = DEFAULT_COLLECTION,
    ) -> None:
        """Put a key-value pair into the store.

        Args:
            key (str): key
            val (dict): value
            collection (str): collection name

        """
        self.put_all([(key, val)], collection=collection)

    async def aput(
        self,
        key: str,
        val: dict,
        collection: str = DEFAULT_COLLECTION,
    ) -> None:
        """Put a key-value pair into the store.

        Args:
            key (str): key
            val (dict): value
            collection (str): collection name

        """
        await self.aput_all([(key, val)], collection=collection)

    def put_all(
        self,
        kv_pairs: List[Tuple[str, dict]],
        collection: str = DEFAULT_COLLECTION,
        batch_size: int = DEFAULT_BATCH_SIZE,
    ) -> None:
        from sqlalchemy import text

        self._initialize()
        with self._session() as session:
            for i in range(0, len(kv_pairs), batch_size):
                batch = kv_pairs[i : i + batch_size]

                # Prepare the VALUES part of the SQL statement
                values_clause = ", ".join(
                    f"(:key_{i}, :namespace_{i}, :value_{i})"
                    for i, _ in enumerate(batch)
                )

                # Prepare the raw SQL for bulk upsert
                # Note: This SQL is PostgreSQL-specific. Adjust for other databases.
                stmt = text(
                    f"""
                INSERT INTO {self.schema_name}.{self._table_class.__tablename__} (key, namespace, value)
                VALUES {values_clause}
                ON CONFLICT (key, namespace)
                DO UPDATE SET
                value = EXCLUDED.value;
                """
                )

                # Flatten the list of tuples for execute parameters
                params = {}
                for i, (key, value) in enumerate(batch):
                    params[f"key_{i}"] = key
                    params[f"namespace_{i}"] = collection
                    params[f"value_{i}"] = json.dumps(value)

                # Execute the bulk upsert
                session.execute(stmt, params)
                session.commit()

    async def aput_all(
        self,
        kv_pairs: List[Tuple[str, dict]],
        collection: str = DEFAULT_COLLECTION,
        batch_size: int = DEFAULT_BATCH_SIZE,
    ) -> None:
        from sqlalchemy import text

        self._initialize()
        async with self._async_session() as session:
            for i in range(0, len(kv_pairs), batch_size):
                batch = kv_pairs[i : i + batch_size]

                # Prepare the VALUES part of the SQL statement
                values_clause = ", ".join(
                    f"(:key_{i}, :namespace_{i}, :value_{i})"
                    for i, _ in enumerate(batch)
                )

                # Prepare the raw SQL for bulk upsert
                # Note: This SQL is PostgreSQL-specific. Adjust for other databases.
                stmt = text(
                    f"""
                INSERT INTO {self.schema_name}.{self._table_class.__tablename__} (key, namespace, value)
                VALUES {values_clause}
                ON CONFLICT (key, namespace)
                DO UPDATE SET
                value = EXCLUDED.value;
                """
                )

                # Flatten the list of tuples for execute parameters
                params = {}
                for i, (key, value) in enumerate(batch):
                    params[f"key_{i}"] = key
                    params[f"namespace_{i}"] = collection
                    params[f"value_{i}"] = json.dumps(value)

                # Execute the bulk upsert
                await session.execute(stmt, params)
                await session.commit()

    def get(self, key: str, collection: str = DEFAULT_COLLECTION) -> Optional[dict]:
        """Get a value from the store.

        Args:
            key (str): key
            collection (str): collection name

        """
        from sqlalchemy import select

        self._initialize()
        with self._session() as session:
            result = session.execute(
                select(self._table_class)
                .filter_by(key=key)
                .filter_by(namespace=collection)
            )
            result = result.scalars().first()
            if result:
                return result.value
        return None

    async def aget(
        self, key: str, collection: str = DEFAULT_COLLECTION
    ) -> Optional[dict]:
        """Get a value from the store.

        Args:
            key (str): key
            collection (str): collection name

        """
        from sqlalchemy import select

        self._initialize()
        async with self._async_session() as session:
            result = await session.execute(
                select(self._table_class)
                .filter_by(key=key)
                .filter_by(namespace=collection)
            )
            result = result.scalars().first()
            if result:
                return result.value
        return None

    def get_all(self, collection: str = DEFAULT_COLLECTION) -> Dict[str, dict]:
        """Get all values from the store.

        Args:
            collection (str): collection name

        """
        from sqlalchemy import select

        self._initialize()
        with self._session() as session:
            results = session.execute(
                select(self._table_class).filter_by(namespace=collection)
            )
            results = results.scalars().all()
        return {result.key: result.value for result in results} if results else {}

    async def aget_all(self, collection: str = DEFAULT_COLLECTION) -> Dict[str, dict]:
        """Get all values from the store.

        Args:
            collection (str): collection name

        """
        from sqlalchemy import select

        self._initialize()
        async with self._async_session() as session:
            results = await session.execute(
                select(self._table_class).filter_by(namespace=collection)
            )
            results = results.scalars().all()
        return {result.key: result.value for result in results} if results else {}

    def delete(self, key: str, collection: str = DEFAULT_COLLECTION) -> bool:
        """Delete a value from the store.

        Args:
            key (str): key
            collection (str): collection name

        """
        from sqlalchemy import delete

        self._initialize()
        with self._session() as session:
            result = session.execute(
                delete(self._table_class)
                .filter_by(namespace=collection)
                .filter_by(key=key)
            )
            session.commit()
        return result.rowcount > 0

    async def adelete(self, key: str, collection: str = DEFAULT_COLLECTION) -> bool:
        """Delete a value from the store.

        Args:
            key (str): key
            collection (str): collection name

        """
        from sqlalchemy import delete

        self._initialize()
        async with self._async_session() as session:
            async with session.begin():
                result = await session.execute(
                    delete(self._table_class)
                    .filter_by(namespace=collection)
                    .filter_by(key=key)
                )
        return result.rowcount > 0

from_params classmethod #

from_params(host: Optional[str] = None, port: Optional[str] = None, database: Optional[str] = None, user: Optional[str] = None, password: Optional[str] = None, table_name: str = 'kvstore', schema_name: str = 'public', connection_string: Optional[str] = None, async_connection_string: Optional[str] = None, perform_setup: bool = True, debug: bool = False, use_jsonb: bool = False) -> PostgresKVStore

Return connection string from database parameters.

Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
@classmethod
def from_params(
    cls,
    host: Optional[str] = None,
    port: Optional[str] = None,
    database: Optional[str] = None,
    user: Optional[str] = None,
    password: Optional[str] = None,
    table_name: str = "kvstore",
    schema_name: str = "public",
    connection_string: Optional[str] = None,
    async_connection_string: Optional[str] = None,
    perform_setup: bool = True,
    debug: bool = False,
    use_jsonb: bool = False,
) -> "PostgresKVStore":
    """Return connection string from database parameters."""
    conn_str = (
        connection_string
        or f"postgresql+psycopg2://{user}:{password}@{host}:{port}/{database}"
    )
    async_conn_str = async_connection_string or (
        f"postgresql+asyncpg://{user}:{password}@{host}:{port}/{database}"
    )
    return cls(
        connection_string=conn_str,
        async_connection_string=async_conn_str,
        table_name=table_name,
        schema_name=schema_name,
        perform_setup=perform_setup,
        debug=debug,
        use_jsonb=use_jsonb,
    )

from_uri classmethod #

from_uri(uri: str, table_name: str = 'kvstore', schema_name: str = 'public', perform_setup: bool = True, debug: bool = False, use_jsonb: bool = False) -> PostgresKVStore

Return connection string from database parameters.

Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
@classmethod
def from_uri(
    cls,
    uri: str,
    table_name: str = "kvstore",
    schema_name: str = "public",
    perform_setup: bool = True,
    debug: bool = False,
    use_jsonb: bool = False,
) -> "PostgresKVStore":
    """Return connection string from database parameters."""
    params = params_from_uri(uri)
    return cls.from_params(
        **params,
        table_name=table_name,
        schema_name=schema_name,
        perform_setup=perform_setup,
        debug=debug,
        use_jsonb=use_jsonb,
    )

put #

put(key: str, val: dict, collection: str = DEFAULT_COLLECTION) -> None

Put a key-value pair into the store.

Parameters:

Name Type Description Default
key str

key

 required
val dict

value

 required
collection str

collection name

 DEFAULT_COLLECTION
Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
def put(
    self,
    key: str,
    val: dict,
    collection: str = DEFAULT_COLLECTION,
) -> None:
    """Put a key-value pair into the store.

    Args:
        key (str): key
        val (dict): value
        collection (str): collection name

    """
    self.put_all([(key, val)], collection=collection)

aput async #

aput(key: str, val: dict, collection: str = DEFAULT_COLLECTION) -> None

Put a key-value pair into the store.

Parameters:

Name Type Description Default
key str

key

 required
val dict

value

 required
collection str

collection name

 DEFAULT_COLLECTION
Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
async def aput(
    self,
    key: str,
    val: dict,
    collection: str = DEFAULT_COLLECTION,
) -> None:
    """Put a key-value pair into the store.

    Args:
        key (str): key
        val (dict): value
        collection (str): collection name

    """
    await self.aput_all([(key, val)], collection=collection)

get #

get(key: str, collection: str = DEFAULT_COLLECTION) -> Optional[dict]

Get a value from the store.

Parameters:

Name Type Description Default
key str

key

 required
collection str

collection name

 DEFAULT_COLLECTION
Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
def get(self, key: str, collection: str = DEFAULT_COLLECTION) -> Optional[dict]:
    """Get a value from the store.

    Args:
        key (str): key
        collection (str): collection name

    """
    from sqlalchemy import select

    self._initialize()
    with self._session() as session:
        result = session.execute(
            select(self._table_class)
            .filter_by(key=key)
            .filter_by(namespace=collection)
        )
        result = result.scalars().first()
        if result:
            return result.value
    return None

aget async #

aget(key: str, collection: str = DEFAULT_COLLECTION) -> Optional[dict]

Get a value from the store.

Parameters:

Name Type Description Default
key str

key

 required
collection str

collection name

 DEFAULT_COLLECTION
Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
async def aget(
    self, key: str, collection: str = DEFAULT_COLLECTION
) -> Optional[dict]:
    """Get a value from the store.

    Args:
        key (str): key
        collection (str): collection name

    """
    from sqlalchemy import select

    self._initialize()
    async with self._async_session() as session:
        result = await session.execute(
            select(self._table_class)
            .filter_by(key=key)
            .filter_by(namespace=collection)
        )
        result = result.scalars().first()
        if result:
            return result.value
    return None

get_all #

get_all(collection: str = DEFAULT_COLLECTION) -> Dict[str, dict]

Get all values from the store.

Parameters:

Name Type Description Default
collection str

collection name

 DEFAULT_COLLECTION
Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
def get_all(self, collection: str = DEFAULT_COLLECTION) -> Dict[str, dict]:
    """Get all values from the store.

    Args:
        collection (str): collection name

    """
    from sqlalchemy import select

    self._initialize()
    with self._session() as session:
        results = session.execute(
            select(self._table_class).filter_by(namespace=collection)
        )
        results = results.scalars().all()
    return {result.key: result.value for result in results} if results else {}

aget_all async #

aget_all(collection: str = DEFAULT_COLLECTION) -> Dict[str, dict]

Get all values from the store.

Parameters:

Name Type Description Default
collection str

collection name

 DEFAULT_COLLECTION
Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
async def aget_all(self, collection: str = DEFAULT_COLLECTION) -> Dict[str, dict]:
    """Get all values from the store.

    Args:
        collection (str): collection name

    """
    from sqlalchemy import select

    self._initialize()
    async with self._async_session() as session:
        results = await session.execute(
            select(self._table_class).filter_by(namespace=collection)
        )
        results = results.scalars().all()
    return {result.key: result.value for result in results} if results else {}

delete #

delete(key: str, collection: str = DEFAULT_COLLECTION) -> bool

Delete a value from the store.

Parameters:

Name Type Description Default
key str

key

 required
collection str

collection name

 DEFAULT_COLLECTION
Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
def delete(self, key: str, collection: str = DEFAULT_COLLECTION) -> bool:
    """Delete a value from the store.

    Args:
        key (str): key
        collection (str): collection name

    """
    from sqlalchemy import delete

    self._initialize()
    with self._session() as session:
        result = session.execute(
            delete(self._table_class)
            .filter_by(namespace=collection)
            .filter_by(key=key)
        )
        session.commit()
    return result.rowcount > 0

adelete async #

adelete(key: str, collection: str = DEFAULT_COLLECTION) -> bool

Delete a value from the store.

Parameters:

Name Type Description Default
key str

key

 required
collection str

collection name

 DEFAULT_COLLECTION
Source code in llama-index-integrations/storage/kvstore/llama-index-storage-kvstore-postgres/llama_index/storage/kvstore/postgres/base.py 
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
async def adelete(self, key: str, collection: str = DEFAULT_COLLECTION) -> bool:
    """Delete a value from the store.

    Args:
        key (str): key
        collection (str): collection name

    """
    from sqlalchemy import delete

    self._initialize()
    async with self._async_session() as session:
        async with session.begin():
            result = await session.execute(
                delete(self._table_class)
                .filter_by(namespace=collection)
                .filter_by(key=key)
            )
    return result.rowcount > 0