chromadb-client 0.4.9.dev0__tar.gz → 0.4.11.dev0__tar.gz

chromadb-client-0.4.11.dev0/.github/workflows/chroma-js-release.yml

@@ -0,0 +1,42 @@
+name: Chroma Release JS Client
+
+on:
+  push:
+    tags:
+      - 'js_release_*.*.*'  # Match tags in the form js_release_X.Y.Z
+      - 'js_release_alpha_*.*.*'  # Match tags in the form js_release_alpha_X.Y.Z
+
+jobs:
+  build-and-release:
+    runs-on: ubuntu-latest
+    permissions: write-all
+    steps:
+    - name: Check if tag matches the pattern
+      run: |
+        if [[ "${{ github.ref }}" =~ ^refs/tags/js_release_alpha_[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
+          echo "Tag matches the pattern js_release_alpha_X.Y.Z"
+          echo "NPM_SCRIPT=release_alpha" >> "$GITHUB_ENV"
+        elif [[ "${{ github.ref }}" =~ ^refs/tags/js_release_[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
+          echo "Tag matches the pattern js_release_X.Y.Z"
+          echo "NPM_SCRIPT=release" >> "$GITHUB_ENV"
+        else
+          echo "Tag does not match the release tag pattern, exiting workflow"
+          exit 1
+        fi
+    - name: Checkout
+      uses: actions/checkout@v3
+      with:
+        fetch-depth: 0
+    - name: Set up JS
+      uses: actions/setup-node@v3
+      with:
+        node-version: '16.x'
+        registry-url: 'https://registry.npmjs.org'
+    - name: Install Client Dev Dependencies
+      run: npm install
+      working-directory: ./clients/js/
+    - name: npm Test & Publish
+      run: npm run db:run && PORT=8001 npm run $NPM_SCRIPT
+      working-directory: ./clients/js/
+      env:
+        NODE_AUTH_TOKEN: ${{ secrets.NPM_TOKEN }}

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/.github/workflows/chroma-release-python-client.yml

@@ -3,14 +3,27 @@ name: Chroma Release Python Client
 on:
   push:
     tags:
-      - '*'
+      - '[0-9]+.[0-9]+.[0-9]+'  # Match tags in the form X.Y.Z
     branches:
       - main
       - hammad/thin_client
 
 jobs:
+  check_tag:
+      runs-on: ubuntu-latest
+      outputs:
+        tag_matches: ${{ steps.check-tag.outputs.tag_matches }}
+      steps:
+        - name: Check Tag
+          id: check-tag
+          run: |
+            if [[ ${{ github.event.ref }} =~ ^refs/tags/[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
+                echo "tag_matches=true" >> $GITHUB_OUTPUT
+            fi
   build-and-release:
     runs-on: ubuntu-latest
+    needs: check_tag
+    if: needs.check_tag.outputs.tag_matches == 'true'
     permissions: write-all
     steps:
     - name: Checkout

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/.github/workflows/chroma-release.yml

@@ -8,24 +8,38 @@ on:
       - main
 
 env:
-  REGISTRY: ghcr.io
-  IMAGE_NAME: "ghcr.io/chroma-core/chroma"
+  GHCR_IMAGE_NAME: "ghcr.io/chroma-core/chroma"
+  DOCKERHUB_IMAGE_NAME: "chromadb/chroma"
+  PLATFORMS: linux/amd64,linux/arm64 #linux/riscv64, linux/arm/v7
 
 jobs:
+  check_tag:
+    runs-on: ubuntu-latest
+    outputs:
+      tag_matches: ${{ steps.check-tag.outputs.tag_matches }}
+    steps:
+      - name: Check Tag
+        id: check-tag
+        run: |
+          if [[ ${{ github.event.ref }} =~ ^refs/tags/[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
+              echo "tag_matches=true" >> $GITHUB_OUTPUT
+          fi
   build-and-release:
     runs-on: ubuntu-latest
+    needs: check_tag
     permissions: write-all
-#      id-token: write
-#      contents: read
-#      deployments: write
-#      packages: write
-#      pull-requests: read
-#      statuses: write
     steps:
     - name: Checkout
       uses: actions/checkout@v3
       with:
         fetch-depth: 0
+    # https://github.com/docker/setup-qemu-action - for multiplatform builds
+    - name: Set up QEMU
+      uses: docker/setup-qemu-action@v2
+    # https://github.com/docker/setup-buildx-action - for multiplatform builds
+    - name: Set up Docker Buildx
+      id: buildx
+      uses: docker/setup-buildx-action@v2
     - name: Set up Python
       uses: actions/setup-python@v4
       with:
@@ -36,34 +50,38 @@ jobs:
       run: python -m build
     - name: Test Client Package
       run: bin/test-package.sh dist/*.tar.gz
-    - name: Log in to the Container registry
+    - name: Log in to the Github Container registry
       uses: docker/login-action@v2.1.0
       with:
-        registry: ${{ env.REGISTRY }}
+        registry: ghcr.io
         username: ${{ github.actor }}
         password: ${{ secrets.GITHUB_TOKEN }}
+    - name: Login to DockerHub
+      uses: docker/login-action@v2.1.0
+      with:
+        username: ${{ secrets.DOCKERHUB_USERNAME }}
+        password: ${{ secrets.DOCKERHUB_TOKEN }}
     - name: Install setuptools_scm
       run: python -m pip install setuptools_scm
-    - name: Get Docker Tag
-      id: tag
-      run: echo "tag_name=$IMAGE_NAME:$(bin/version)" >> $GITHUB_OUTPUT
+    - name: Get Release Version
+      id: version
+      run: echo "version=$(python -m setuptools_scm)" >> $GITHUB_OUTPUT
     - name: Build and push prerelease Docker image
-      if: "!startsWith(github.ref, 'refs/tags/')"
+      if: "needs.check_tag.outputs.tag_matches != 'true'"
       uses: docker/build-push-action@v3.2.0
       with:
         context: .
+        platforms: ${{ env.PLATFORMS }}
         push: true
-        tags: ${{ steps.tag.outputs.tag_name}}
+        tags: "${{ env.GHCR_IMAGE_NAME }}:${{ steps.version.outputs.version }},${{ env.DOCKERHUB_IMAGE_NAME }}:${{ steps.version.outputs.version }}"
     - name: Build and push release Docker image
-      if: "startsWith(github.ref, 'refs/tags/')"
+      if: "needs.check_tag.outputs.tag_matches == 'true'"
       uses: docker/build-push-action@v3.2.0
       with:
         context: .
+        platforms: ${{ env.PLATFORMS }}
         push: true
-        tags: "${{ steps.tag.outputs.tag_name }},${{ env.IMAGE_NAME }}:latest"
-    - name: Get Release Version
-      id: version
-      run: echo "version=$(python -m setuptools_scm)" >> $GITHUB_OUTPUT
+        tags: "${{ env.GHCR_IMAGE_NAME }}:${{ steps.version.outputs.version }},${{ env.DOCKERHUB_IMAGE_NAME }}:${{ steps.version.outputs.version }},${{ env.GHCR_IMAGE_NAME }}:latest,${{ env.DOCKERHUB_IMAGE_NAME }}:latest"
     - name: Get current date
       id: builddate
       run: echo "builddate=$(date +'%Y-%m-%dT%H:%M')" >> $GITHUB_OUTPUT
@@ -73,7 +91,7 @@ jobs:
         password: ${{ secrets.TEST_PYPI_API_TOKEN }}
         repository_url: https://test.pypi.org/legacy/
     - name: Publish to PyPI
-      if: startsWith(github.ref, 'refs/tags')
+      if: "needs.check_tag.outputs.tag_matches == 'true'"
       uses: pypa/gh-action-pypi-publish@release/v1
       with:
         password: ${{ secrets.PYPI_API_TOKEN }}
@@ -84,31 +102,32 @@ jobs:
         aws-region: us-east-1
     - name: Generate CloudFormation template
       id: generate-cf
-      if: "startsWith(github.ref, 'refs/tags/')"
+      if: "needs.check_tag.outputs.tag_matches == 'true'"
       run: "pip install boto3 && python bin/generate_cloudformation.py"
     - name: Release Tagged Version
       uses: ncipollo/release-action@v1.11.1
-      if: "startsWith(github.ref, 'refs/tags/')"
+      if: "needs.check_tag.outputs.tag_matches == 'true'"
       with:
         body: |
           Version: `${{steps.version.outputs.version}}`
           Git ref: `${{github.ref}}`
           Build Date: `${{steps.builddate.outputs.builddate}}`
           PIP Package: `chroma-${{steps.version.outputs.version}}.tar.gz`
-          Docker Image: `${{steps.tag.outputs.tag_name}}`
+          Github Container Registry Image: `${{ env.GHCR_IMAGE_NAME }}:${{ steps.version.outputs.version }}`
+          DockerHub Image: `${{ env.DOCKERHUB_IMAGE_NAME }}:${{ steps.version.outputs.version }}`
         artifacts: "dist/chroma-${{steps.version.outputs.version}}.tar.gz"
         prerelease: true
         generateReleaseNotes: true
     - name: Update Tag
       uses: richardsimko/update-tag@v1.0.5
-      if: "!startsWith(github.ref, 'refs/tags/')"
+      if: "needs.check_tag.outputs.tag_matches != 'true'"
       with:
         tag_name: latest
       env:
         GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
     - name: Release Latest
       uses: ncipollo/release-action@v1.11.1
-      if: "!startsWith(github.ref, 'refs/tags/')"
+      if: "needs.check_tag.outputs.tag_matches != 'true'"
       with:
         tag: "latest"
         name: "Latest"
@@ -117,7 +136,8 @@ jobs:
           Git ref: `${{github.ref}}`
           Build Date: `${{steps.builddate.outputs.builddate}}`
           PIP Package: `chroma-${{steps.version.outputs.version}}.tar.gz`
-          Docker Image: `${{steps.tag.outputs.tag_name}}`
+          Github Container Registry Image: `${{ env.GHCR_IMAGE_NAME }}:${{ steps.version.outputs.version }}`
+          DockerHub Image: `${{ env.DOCKERHUB_IMAGE_NAME }}:${{ steps.version.outputs.version }}`
         artifacts: "dist/chroma-${{steps.version.outputs.version}}.tar.gz"
         allowUpdates: true
         prerelease: true

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/.gitignore

@@ -11,6 +11,10 @@
 
 index_data
 
+# Default configuration for persist_directory in chromadb/config.py
+# Currently it's located in "./chroma/"
+chroma/
+
 venv
 .env
 .chroma

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: chromadb-client
-Version: 0.4.9.dev0
+Version: 0.4.11.dev0
 Summary: Chroma Client.
 Author-email: Jeff Huber <jeff@trychroma.com>, Anton Troynikov <anton@trychroma.com>
 Project-URL: Homepage, https://github.com/chroma-core/chroma
@@ -11,6 +11,12 @@ Classifier: Operating System :: OS Independent
 Requires-Python: >=3.7
 Description-Content-Type: text/markdown
 License-File: LICENSE
+Requires-Dist: requests>=2.28
+Requires-Dist: pydantic<2.0,>=1.9
+Requires-Dist: numpy>=1.21.6
+Requires-Dist: posthog>=2.4.0
+Requires-Dist: typing_extensions>=4.5.0
+Requires-Dist: overrides>=7.3.1
 
 <p align="center">
   <a href="https://trychroma.com"><img src="https://user-images.githubusercontent.com/891664/227103090-6624bf7d-9524-4e05-9d2c-c28d5d451481.png" alt="Chroma logo"></a>

chromadb-client-0.4.11.dev0/bin/integration-test

@@ -0,0 +1,46 @@
+#!/usr/bin/env bash
+
+set -e
+
+export CHROMA_PORT=8000
+
+function cleanup {
+  docker compose -f docker-compose.test.yml down --rmi local --volumes
+  rm server.htpasswd .chroma_env
+}
+
+function setup_basic_auth {
+  # Generate htpasswd file
+  docker run --rm --entrypoint htpasswd httpd:2 -Bbn admin admin > server.htpasswd
+  # Create .chroma_env file
+  cat <<EOF > .chroma_env
+CHROMA_SERVER_AUTH_CREDENTIALS_FILE="/chroma/server.htpasswd"
+CHROMA_SERVER_AUTH_CREDENTIALS_PROVIDER='chromadb.auth.providers.HtpasswdFileServerAuthCredentialsProvider'
+CHROMA_SERVER_AUTH_PROVIDER='chromadb.auth.basic.BasicAuthServerProvider'
+EOF
+}
+
+trap cleanup EXIT
+
+docker compose -f docker-compose.test.yml up --build -d
+
+export CHROMA_INTEGRATION_TEST_ONLY=1
+export CHROMA_API_IMPL=chromadb.api.fastapi.FastAPI
+export CHROMA_SERVER_HOST=localhost
+export CHROMA_SERVER_HTTP_PORT=8000
+
+echo testing: python -m pytest "$@"
+python -m pytest "$@"
+
+cd clients/js
+yarn
+yarn test:run
+docker compose down
+cd ../..
+echo "Testing auth"
+setup_basic_auth #this is specific to the auth type, later on we'll have other auth types
+cd clients/js
+# Start docker compose - this should be auth agnostic
+docker compose --env-file ../../.chroma_env -f ../../docker-compose.test-auth.yml up --build -d
+yarn test:run-auth
+cd ../..

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/chromadb/__init__.py

@@ -44,7 +44,7 @@ logger = logging.getLogger(__name__)
 
 __settings = Settings()
 
-__version__ = "0.4.8"
+__version__ = "0.4.10"
 
 # Workaround to deal with Colab's old sqlite3 version
 try:

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/chromadb/api/fastapi.py

@@ -1,4 +1,5 @@
 import json
+import logging
 from typing import Optional, cast
 from typing import Sequence
 from uuid import UUID
@@ -30,27 +31,73 @@ from chromadb.auth.providers import RequestsClientAuthProtocolAdapter
 from chromadb.auth.registry import resolve_provider
 from chromadb.config import Settings, System
 from chromadb.telemetry import Telemetry
+from urllib.parse import urlparse, urlunparse, quote
+
+logger = logging.getLogger(__name__)
 
 
 class FastAPI(API):
     _settings: Settings
 
+    @staticmethod
+    def _validate_host(host: str) -> None:
+        parsed = urlparse(host)
+        if "/" in host and parsed.scheme not in {"http", "https"}:
+            raise ValueError(
+                "Invalid URL. " f"Unrecognized protocol - {parsed.scheme}."
+            )
+        if "/" in host and (not host.startswith("http")):
+            raise ValueError(
+                "Invalid URL. "
+                "Seems that you are trying to pass URL as a host but without specifying the protocol. "
+                "Please add http:// or https:// to the host."
+            )
+
+    @staticmethod
+    def resolve_url(
+        chroma_server_host: str,
+        chroma_server_ssl_enabled: Optional[bool] = False,
+        default_api_path: Optional[str] = "",
+        chroma_server_http_port: Optional[int] = 8000,
+    ) -> str:
+        _skip_port = False
+        _chroma_server_host = chroma_server_host
+        FastAPI._validate_host(_chroma_server_host)
+        if _chroma_server_host.startswith("http"):
+            logger.debug("Skipping port as the user is passing a full URL")
+            _skip_port = True
+        parsed = urlparse(_chroma_server_host)
+
+        scheme = "https" if chroma_server_ssl_enabled else parsed.scheme or "http"
+        net_loc = parsed.netloc or parsed.hostname or chroma_server_host
+        port = (
+            ":" + str(parsed.port or chroma_server_http_port) if not _skip_port else ""
+        )
+        path = parsed.path or default_api_path
+
+        if not path or path == net_loc:
+            path = default_api_path if default_api_path else ""
+        if not path.endswith(default_api_path or ""):
+            path = path + default_api_path if default_api_path else ""
+        full_url = urlunparse(
+            (scheme, f"{net_loc}{port}", quote(path.replace("//", "/")), "", "", "")
+        )
+
+        return full_url
+
     def __init__(self, system: System):
         super().__init__(system)
-        url_prefix = "https" if system.settings.chroma_server_ssl_enabled else "http"
         system.settings.require("chroma_server_host")
         system.settings.require("chroma_server_http_port")
 
         self._telemetry_client = self.require(Telemetry)
         self._settings = system.settings
 
-        port_suffix = (
-            f":{system.settings.chroma_server_http_port}"
-            if system.settings.chroma_server_http_port
-            else ""
-        )
-        self._api_url = (
-            f"{url_prefix}://{system.settings.chroma_server_host}{port_suffix}/api/v1"
+        self._api_url = FastAPI.resolve_url(
+            chroma_server_host=str(system.settings.chroma_server_host),
+            chroma_server_http_port=int(str(system.settings.chroma_server_http_port)),
+            chroma_server_ssl_enabled=system.settings.chroma_server_ssl_enabled,
+            default_api_path=system.settings.chroma_server_api_default_path,
         )
 
         self._header = system.settings.chroma_server_headers

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/chromadb/api/types.py

@@ -207,6 +207,8 @@ def validate_where(where: Where) -> Where:
         if (
             key != "$and"
             and key != "$or"
+            and key != "$in"
+            and key != "$nin"
             and not isinstance(value, (str, int, float, dict))
         ):
             raise ValueError(
@@ -238,15 +240,37 @@ def validate_where(where: Where) -> Where:
                         raise ValueError(
                             f"Expected operand value to be an int or a float for operator {operator}, got {operand}"
                         )
-
-                if operator not in ["$gt", "$gte", "$lt", "$lte", "$ne", "$eq"]:
+                if operator in ["$in", "$nin"]:
+                    if not isinstance(operand, list):
+                        raise ValueError(
+                            f"Expected operand value to be an list for operator {operator}, got {operand}"
+                        )
+                if operator not in [
+                    "$gt",
+                    "$gte",
+                    "$lt",
+                    "$lte",
+                    "$ne",
+                    "$eq",
+                    "$in",
+                    "$nin",
+                ]:
                     raise ValueError(
-                        f"Expected where operator to be one of $gt, $gte, $lt, $lte, $ne, $eq, got {operator}"
+                        f"Expected where operator to be one of $gt, $gte, $lt, $lte, $ne, $eq, $in, $nin, "
+                        f"got {operator}"
                     )
 
-                if not isinstance(operand, (str, int, float)):
+                if not isinstance(operand, (str, int, float, list)):
+                    raise ValueError(
+                        f"Expected where operand value to be a str, int, float, or list of those type, got {operand}"
+                    )
+                if isinstance(operand, list) and (
+                    len(operand) == 0
+                    or not all(isinstance(x, type(operand[0])) for x in operand)
+                ):
                     raise ValueError(
-                        f"Expected where operand value to be a str, int, or float, got {operand}"
+                        f"Expected where operand value to be a non-empty list, and all values to obe of the same type "
+                        f"got {operand}"
                     )
     return where
 

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/chromadb/auth/token/__init__.py

@@ -68,7 +68,10 @@ class TokenAuthClientAuthResponse(ClientAuthResponse):
 
 def check_token(token: str) -> None:
     token_str = str(token)
-    if not all(c in string.ascii_letters + string.digits for c in token_str):
+    if not all(
+        c in string.digits + string.ascii_letters + string.punctuation
+        for c in token_str
+    ):
         raise ValueError("Invalid token. Must contain only ASCII letters and digits.")
 
 

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/chromadb/config.py

@@ -88,6 +88,7 @@ class Settings(BaseSettings):  # type: ignore
     chroma_server_headers: Optional[Dict[str, str]] = None
     chroma_server_http_port: Optional[str] = None
     chroma_server_ssl_enabled: Optional[bool] = False
+    chroma_server_api_default_path: Optional[str] = "/api/v1"
     chroma_server_grpc_port: Optional[str] = None
     chroma_server_cors_allow_origins: List[str] = []  # eg ["http://localhost:3000"]
 

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/chromadb/segment/__init__.py

@@ -39,6 +39,11 @@ class SegmentImplementation(Component):
         segment. Validation errors will be reported to the user."""
         return None
 
+    @abstractmethod
+    def delete(self) -> None:
+        """Delete the segment and all its data"""
+        ...
+
 
 S = TypeVar("S", bound=SegmentImplementation)
 

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/chromadb/segment/impl/manager/local.py

@@ -112,6 +112,9 @@ class LocalSegmentManager(SegmentManager):
         segments = self._sysdb.get_segments(collection=collection_id)
         for segment in segments:
             if segment["id"] in self._instances:
+                if segment["type"] == SegmentType.HNSW_LOCAL_PERSISTED.value:
+                    instance = self.get_segment(collection_id, VectorReader)
+                    instance.delete()
                 del self._instances[segment["id"]]
             if collection_id in self._segment_cache:
                 if segment["scope"] in self._segment_cache[collection_id]:

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/chromadb/segment/impl/metadata/sqlite.py

@@ -1,8 +1,8 @@
-from typing import Optional, Sequence, Any, Tuple, cast, Generator, Union, Dict
+from typing import Optional, Sequence, Any, Tuple, cast, Generator, Union, Dict, List
 from chromadb.segment import MetadataReader
 from chromadb.ingest import Consumer
 from chromadb.config import System
-from chromadb.types import Segment
+from chromadb.types import Segment, InclusionExclusionOperator
 from chromadb.db.impl.sqlite import SqliteDB
 from overrides import override
 from chromadb.db.base import (
@@ -105,7 +105,6 @@ class SqliteMetadataSegment(MetadataReader):
         offset: Optional[int] = None,
     ) -> Sequence[MetadataEmbeddingRecord]:
         """Query for embedding metadata."""
-
         embeddings_t, metadata_t, fulltext_t = Tables(
             "embeddings", "embedding_metadata", "embedding_fulltext_search"
         )
@@ -135,7 +134,6 @@ class SqliteMetadataSegment(MetadataReader):
 
         if where:
             q = q.where(self._where_map_criterion(q, where, embeddings_t, metadata_t))
-
         if where_document:
             q = q.where(
                 self._where_doc_criterion(q, where_document, embeddings_t, fulltext_t)
@@ -146,7 +144,6 @@ class SqliteMetadataSegment(MetadataReader):
 
         limit = limit or 2**63 - 1
         offset = offset or 0
-
         with self._db.tx() as cur:
             return list(islice(self._records(cur, q), offset, offset + limit))
 
@@ -405,7 +402,6 @@ class SqliteMetadataSegment(MetadataReader):
         self, q: QueryBuilder, where: Where, embeddings_t: Table, metadata_t: Table
     ) -> Criterion:
         clause: list[Criterion] = []
-
         for k, v in where.items():
             if k == "$and":
                 criteria = [
@@ -466,6 +462,10 @@ class SqliteMetadataSegment(MetadataReader):
                 raise ValueError(f"Unknown where_doc operator {k}")
         raise ValueError("Empty where_doc")
 
+    @override
+    def delete(self) -> None:
+        raise NotImplementedError()
+
 
 def _encode_seq_id(seq_id: SeqId) -> bytes:
     """Encode a SeqID into a byte array"""
@@ -488,24 +488,31 @@ def _decode_seq_id(seq_id_bytes: bytes) -> SeqId:
 
 
 def _where_clause(
-    expr: Union[LiteralValue, Dict[WhereOperator, LiteralValue]],
+    expr: Union[
+        LiteralValue,
+        Dict[WhereOperator, LiteralValue],
+        Dict[InclusionExclusionOperator, List[LiteralValue]],
+    ],
     table: Table,
 ) -> Criterion:
     """Given a field name, an expression, and a table, construct a Pypika Criterion"""
 
     # Literal value case
     if isinstance(expr, (str, int, float, bool)):
-        return _where_clause({"$eq": expr}, table)
+        return _where_clause({cast(WhereOperator, "$eq"): expr}, table)
 
     # Operator dict case
     operator, value = next(iter(expr.items()))
     return _value_criterion(value, operator, table)
 
 
-def _value_criterion(value: LiteralValue, op: WhereOperator, table: Table) -> Criterion:
+def _value_criterion(
+    value: Union[LiteralValue, List[LiteralValue]],
+    op: Union[WhereOperator, InclusionExclusionOperator],
+    table: Table,
+) -> Criterion:
     """Return a criterion to compare a value with the appropriate columns given its type
     and the operation type."""
-
     if isinstance(value, str):
         cols = [table.string_value]
     # isinstance(True, int) evaluates to True, so we need to check for bools separately
@@ -515,6 +522,43 @@ def _value_criterion(value: LiteralValue, op: WhereOperator, table: Table) -> Cr
         cols = [table.int_value]
     elif isinstance(value, float) and op in ("$eq", "$ne"):
         cols = [table.float_value]
+    elif isinstance(value, list) and op in ("$in", "$nin"):
+        _v = value
+        if len(_v) == 0:
+            raise ValueError(f"Empty list for {op} operator")
+        if isinstance(value[0], str):
+            col_exprs = [
+                table.string_value.isin(ParameterValue(_v))
+                if op == "$in"
+                else table.str_value.notin(ParameterValue(_v))
+            ]
+        elif isinstance(value[0], bool):
+            col_exprs = [
+                table.bool_value.isin(ParameterValue(_v))
+                if op == "$in"
+                else table.bool_value.notin(ParameterValue(_v))
+            ]
+        elif isinstance(value[0], int):
+            col_exprs = [
+                table.int_value.isin(ParameterValue(_v))
+                if op == "$in"
+                else table.int_value.notin(ParameterValue(_v))
+            ]
+        elif isinstance(value[0], float):
+            col_exprs = [
+                table.float_value.isin(ParameterValue(_v))
+                if op == "$in"
+                else table.float_value.notin(ParameterValue(_v))
+            ]
+    elif isinstance(value, list) and op in ("$in", "$nin"):
+        col_exprs = [
+            table.int_value.isin(ParameterValue(value))
+            if op == "$in"
+            else table.int_value.notin(ParameterValue(value)),
+            table.float_value.isin(ParameterValue(value))
+            if op == "$in"
+            else table.float_value.notin(ParameterValue(value)),
+        ]
     else:
         cols = [table.int_value, table.float_value]
 

{chromadb-client-0.4.9.dev0 → chromadb-client-0.4.11.dev0}/chromadb/segment/impl/vector/local_hnsw.py

@@ -305,3 +305,7 @@ class LocalHnswSegment(VectorReader):
                     batch.apply(record, label is not None)
 
             self._apply_batch(batch)
+
+    @override
+    def delete(self) -> None:
+        raise NotImplementedError()