cocoindex 0.2.0__tar.gz → 0.2.2__tar.gz

{cocoindex-0.2.0 → cocoindex-0.2.2}/Cargo.lock

@@ -1283,7 +1283,7 @@ dependencies = [
 
 [[package]]
 name = "cocoindex"
-version = "0.2.0"
+version = "0.2.2"
 dependencies = [
  "anyhow",
  "async-openai",

{cocoindex-0.2.0 → cocoindex-0.2.2}/Cargo.toml

@@ -2,7 +2,7 @@
 name = "cocoindex"
 # Version used for local development is always higher than others to take precedence.
 # Will be overridden for specific release versions.
-version = "0.2.0"
+version = "0.2.2"
 edition = "2024"
 rust-version = "1.88"
 readme = "README.md"

{cocoindex-0.2.0 → cocoindex-0.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cocoindex
-Version: 0.2.0
+Version: 0.2.2
 Requires-Dist: click>=8.1.8
 Requires-Dist: rich>=14.0.0
 Requires-Dist: python-dotenv>=1.1.0
@@ -84,7 +84,7 @@ CocoIndex makes it effortless to transform data with AI, and keep source data an
 </br>
 
 <p align="center">
-<img width="4187" height="1883" alt="CocoIndex Features" src="https://github.com/user-attachments/assets/6147673f-0daf-4313-a0c7-7c4205ba2e31" />
+<img alt="CocoIndex Features" src="https://cocoindex.io/images/venn2.svg" />
 </p>
 
 </br>

{cocoindex-0.2.0 → cocoindex-0.2.2}/README.md

@@ -55,7 +55,7 @@ CocoIndex makes it effortless to transform data with AI, and keep source data an
 </br>
 
 <p align="center">
-<img width="4187" height="1883" alt="CocoIndex Features" src="https://github.com/user-attachments/assets/6147673f-0daf-4313-a0c7-7c4205ba2e31" />
+<img alt="CocoIndex Features" src="https://cocoindex.io/images/venn2.svg" />
 </p>
 
 </br>

{cocoindex-0.2.0 → cocoindex-0.2.2}/docs/docs/core/settings.mdx

@@ -96,11 +96,11 @@ If not set, all flows are in a default unnamed namespace.
 
     :::
 
-*   `max_connections` (type: `int`, default: `64`): The maximum number of connections to keep in the pool.
+*   `max_connections` (type: `int`, default: `25`): The maximum number of connections to keep in the pool.
 
     *Environment variable* for `Settings.database.max_connections`: `COCOINDEX_DATABASE_MAX_CONNECTIONS`
 
-*   `min_connections` (type: `int`, default: `16`): The minimum number of connections to keep in the pool.
+*   `min_connections` (type: `int`, default: `5`): The minimum number of connections to keep in the pool.
 
     *Environment variable* for `Settings.database.min_connections`: `COCOINDEX_DATABASE_MIN_CONNECTIONS`
 
@@ -134,7 +134,7 @@ This is the list of environment variables, each of which has a corresponding fie
 | `COCOINDEX_DATABASE_URL` | `database.url` | Yes |
 | `COCOINDEX_DATABASE_USER` | `database.user` | No |
 | `COCOINDEX_DATABASE_PASSWORD` | `database.password` | No |
-| `COCOINDEX_DATABASE_MAX_CONNECTIONS` | `database.max_connections` | No (default: `64`) |
-| `COCOINDEX_DATABASE_MIN_CONNECTIONS` | `database.min_connections` | No (default: `16`) |
+| `COCOINDEX_DATABASE_MAX_CONNECTIONS` | `database.max_connections` | No (default: `25`) |
+| `COCOINDEX_DATABASE_MIN_CONNECTIONS` | `database.min_connections` | No (default: `5`) |
 | `COCOINDEX_SOURCE_MAX_INFLIGHT_ROWS` | `global_execution_options.source_max_inflight_rows` | No (default: `1024`) |
 | `COCOINDEX_SOURCE_MAX_INFLIGHT_BYTES` | `global_execution_options.source_max_inflight_bytes` | No |

{cocoindex-0.2.0 → cocoindex-0.2.2}/docs/docs/examples/examples/academic_papers_index.md

@@ -12,7 +12,9 @@ tags: [vector-index, metadata]
 
 import { GitHubButton, YouTubeButton, DocumentationButton } from '../../../src/components/GitHubButton';
 
-<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/paper_metadata"/>
+<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/paper_metadata" margin="0 0 24px 0" />
+
+![Academic Papers Index](/img/examples/academic_papers_index/cover.png)
 
 ## What we will achieve
 
@@ -45,7 +47,7 @@ to answer questions like "Give me all the papers by Jeff Dean."
   CocoIndex uses PostgreSQL internally for incremental processing.
 - [Configure your OpenAI API key](https://cocoindex.io/docs/ai/llm#openai). Alternatively, we have native support for Gemini, Ollama, LiteLLM. You can choose your favorite LLM provider and work completely on-premises.
 
-  <DocumentationButton href="https://cocoindex.io/docs/ai/llm" text="LLM" margin="0 0 16px 0" />
+  <DocumentationButton url="https://cocoindex.io/docs/ai/llm" text="LLM" margin="0 0 16px 0" />
 
 ## Import the Papers
 
@@ -61,7 +63,7 @@ def paper_metadata_flow(
 ```
 
 `flow_builder.add_source` will create a table with sub fields (`filename`, `content`).
-<DocumentationButton href="https://cocoindex.io/docs/ops/sources" text="Sources" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/ops/sources" text="Sources" margin="0 0 16px 0" />
 
 ## Extract and collect metadata
 
@@ -321,7 +323,7 @@ metadata_embeddings.export(
 
 In this example we use PGVector as embedding store. With CocoIndex, you can do one line switch on other supported Vector databases.
 
-<DocumentationButton href="https://cocoindex.io/docs/ops/targets#entry-oriented-targets" text="Entry Oriented Targets" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/ops/targets#entry-oriented-targets" text="Entry Oriented Targets" margin="0 0 16px 0" />
 
 ## Query the index
 

{cocoindex-0.2.0 → cocoindex-0.2.2}/docs/docs/examples/examples/codebase_index.md

@@ -12,23 +12,23 @@ tags: [vector-index, codebase]
 
 import { GitHubButton, YouTubeButton, DocumentationButton } from '../../../src/components/GitHubButton';
 
-<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/code_embedding"/>
-<YouTubeButton url="https://youtu.be/G3WstvhHO24?si=ndYfM0XRs03_hVPR" />
+<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/code_embedding" margin="0 0 24px 0" />
+<YouTubeButton url="https://youtu.be/G3WstvhHO24?si=ndYfM0XRs03_hVPR" margin="0 0 24px 0" />
+
+![Codebase Index](/img/examples/codebase_index/cover.png)
 
 ## Overview
 In this tutorial, we will build codebase index. [CocoIndex](https://github.com/cocoindex-io/cocoindex) provides built-in support for codebase chunking, with native Tree-sitter support. It works with large codebases, and can be updated in near real-time with incremental processing - only reprocess what's changed. 
 
 ## Use Cases
-A wide range of applications can be built with an effective codebase index that is always up-to-date. Some examples include:
-
-![Use case illustration](/img/examples/codebase_index/usecase.png)
+A wide range of applications can be built with an effective codebase index that is always up-to-date.
 
 - Semantic code context for AI coding agents like Claude, Codex, Gemini CLI.
 - MCP for code editors such as Cursor, Windsurf, and VSCode.
 - Context-aware code search applications—semantic code search, natural language code retrieval.
 - Context for code review agents—AI code review, automated code analysis, code quality checks, pull request summarization.
 - Automated code refactoring, large-scale code migration.
-- Enhance SRE workflows: enable rapid root cause analysis, incident response, and change impact assessment by indexing infrastructure-as-code, deployment scripts, and config files for semantic search and lineage tracking.
+- SRE workflows: enable rapid root cause analysis, incident response, and change impact assessment by indexing infrastructure-as-code, deployment scripts, and config files for semantic search and lineage tracking.
 - Automatically generate design documentation from code—keep design docs up-to-date.
 
 ## Flow Overview
@@ -67,7 +67,7 @@ def code_embedding_flow(flow_builder: cocoindex.FlowBuilder, data_scope: cocoind
 - Exclude files and directories starting `.`,  `target` in the root and `node_modules` under any directory.
 
 `flow_builder.add_source` will create a table with sub fields (`filename`, `content`). 
-<DocumentationButton href="https://cocoindex.io/docs/ops/sources" text="Sources" />
+<DocumentationButton url="https://cocoindex.io/docs/ops/sources" text="Sources" />
 
 
 ## Process each file and collect the information
@@ -84,7 +84,7 @@ def extract_extension(filename: str) -> str:
     return os.path.splitext(filename)[1]
 ```
 
-<DocumentationButton href="https://cocoindex.io/docs/custom_ops/custom_functions" text="Custom Function" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/custom_ops/custom_functions" text="Custom Function" margin="0 0 16px 0" />
 
 ### Split the file into chunks
 We use the `SplitRecursively` function to split the file into chunks.  `SplitRecursively` is CocoIndex building block, with native integration with Tree-sitter. You need to pass in the language to the `language` parameter if you are processing code.
@@ -97,7 +97,7 @@ with data_scope["files"].row() as file:
           cocoindex.functions.SplitRecursively(),
           language=file["extension"], chunk_size=1000, chunk_overlap=300) 
 ```
-<DocumentationButton href="https://cocoindex.io/docs/ops/functions#splitrecursively" text="SplitRecursively" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/ops/functions#splitrecursively" text="SplitRecursively" margin="0 0 16px 0" />
 
 ![SplitRecursively](/img/examples/codebase_index/chunk.png)
 
@@ -112,13 +112,13 @@ def code_to_embedding(text: cocoindex.DataSlice[str]) -> cocoindex.DataSlice[lis
             model="sentence-transformers/all-MiniLM-L6-v2"))
 ```
 
-<DocumentationButton href="https://cocoindex.io/docs/ops/functions#sentencetransformerembed" text="SentenceTransformerEmbed" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/ops/functions#sentencetransformerembed" text="SentenceTransformerEmbed" margin="0 0 16px 0" />
 
 :::tip
 `@cocoindex.transform_flow()` is needed to share the transformation across indexing and query. When building a vector index and querying against it, the embedding computation must remain consistent between indexing and querying.
 :::
 
-<DocumentationButton href="https://cocoindex.io/docs/query#transform-flow" text="Transform Flow" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/query#transform-flow" text="Transform Flow" margin="0 0 16px 0" />
 
 Then for each chunk, we will embed it using the `code_to_embedding` function, and collect the embeddings to the `code_embeddings` collector.
 
@@ -222,4 +222,11 @@ To spin up, it is super easy.
 ```
 cocoindex server main.py -ci
 ```
-Follow the url from the terminal - `https://cocoindex.io/cocoinsight` to access the CocoInsight.
+Follow the url from the terminal - `https://cocoindex.io/cocoinsight` to access the CocoInsight.
+
+
+## Supported Languages
+
+SplitRecursively has native support for all major programming languages.
+
+<DocumentationButton url="https://cocoindex.io/docs/ops/functions#supported-languages" text="Supported Languages" margin="0 0 16px 0" />

{cocoindex-0.2.0 → cocoindex-0.2.2}/docs/docs/examples/examples/custom_targets.md

@@ -11,7 +11,9 @@ tags: [custom-building-blocks]
 ---
 import { GitHubButton, YouTubeButton, DocumentationButton } from '../../../src/components/GitHubButton';
 
-<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/custom_output_files"/>
+<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/custom_output_files" margin="0 0 24px 0" />
+
+![Custom Targets](/img/examples/custom_targets/cover.png)
 
 ## Overview
 
@@ -33,7 +35,7 @@ flow_builder: cocoindex.FlowBuilder, data_scope: cocoindex.DataScope
 	)
 ```
 This ingestion creates a table with `filename` and `content` fields. 
-<DocumentationButton href="https://cocoindex.io/docs/ops/sources" text="Sources" />
+<DocumentationButton url="https://cocoindex.io/docs/ops/sources" text="Sources" />
 
 ## Process each file and collect
 
@@ -45,7 +47,7 @@ def markdown_to_html(text: str) -> str:
     return _markdown_it.render(text)
 ```
 
-<DocumentationButton href="https://cocoindex.io/docs/custom_ops/custom_functions" text="Custom Function" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/custom_ops/custom_functions" text="Custom Function" margin="0 0 16px 0" />
 
 Define data collector and transform each document to html.
 
@@ -62,7 +64,7 @@ with data_scope["documents"].row() as doc:
 
 ### Define the target spec
 
-<DocumentationButton href="https://cocoindex.io/docs/custom_ops/custom_targets#target-spec" text="Target Spec" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/custom_ops/custom_targets#target-spec" text="Target Spec" margin="0 0 16px 0" />
 
 The target spec contains a directory for output files:
 
@@ -74,7 +76,7 @@ class LocalFileTarget(cocoindex.op.TargetSpec):
 
 ### Implement the connector
 
-<DocumentationButton href="https://cocoindex.io/docs/custom_ops/custom_targets#target-connector" text="Target Connector" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/custom_ops/custom_targets#target-connector" text="Target Connector" margin="0 0 16px 0" />
 
 `get_persistent_key()` defines the persistent key,
 which uniquely identifies the target for change tracking and incremental updates. Here, we simply use the target directory as the key (e.g., `./data/output`).

{cocoindex-0.2.0 → cocoindex-0.2.2}/docs/docs/examples/examples/docs_to_knowledge_graph.md

@@ -12,8 +12,10 @@ tags: [knowledge-graph, structured-data-extraction]
 
 import { GitHubButton, YouTubeButton, DocumentationButton } from '../../../src/components/GitHubButton';
 
-<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/docs_to_knowledge_graph"/>
-<YouTubeButton url="https://youtu.be/2KVkpUGRtnk?si=MRalDweWrid-IFje" />
+<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/docs_to_knowledge_graph" margin="0 0 24px 0" />
+<YouTubeButton url="https://youtu.be/2KVkpUGRtnk?si=MRalDweWrid-IFje" margin="0 0 24px 0" />
+
+![Knowledge Graph for Docs](/img/examples/docs_to_knowledge_graph/cover.png)
 
 ## Overview
 [CocoIndex](https://github.com/cocoindex-io/cocoindex) makes it easy to build and maintain knowledge graphs with continuous source updates. In this tutorial, we will use LLM to extract relationships between the concepts in each document, and generate two kinds of relationships:
@@ -37,11 +39,11 @@ and then build a knowledge graph.
 *   [Install Neo4j](https://cocoindex.io/docs/ops/targets#neo4j-dev-instance), a graph database.
 *   [Configure your OpenAI API key](https://cocoindex.io/docs/ai/llm#openai).  Alternatively, we have native support for Gemini, Ollama, LiteLLM. You can choose your favorite LLM provider and work completely on-premises.
 
-    <DocumentationButton href="https://cocoindex.io/docs/ai/llm" text="LLM" margin="0 0 16px 0" />
+    <DocumentationButton url="https://cocoindex.io/docs/ai/llm" text="LLM" margin="0 0 16px 0" />
 
 
 ## Documentation
-<DocumentationButton href="https://cocoindex.io/docs/ops/targets#property-graph-targets" text="Property Graph Targets" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/ops/targets#property-graph-targets" text="Property Graph Targets" margin="0 0 16px 0" />
 
 
 ## Data flow to build knowledge graph
@@ -61,7 +63,7 @@ def docs_to_kg_flow(flow_builder: cocoindex.FlowBuilder, data_scope: cocoindex.D
 Here `flow_builder.add_source` creates a [KTable](https://cocoindex.io/docs/core/data_types#KTable).
 `filename` is the key of the KTable.
 
-<DocumentationButton href="https://cocoindex.io/docs/ops/sources" text="Sources" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/ops/sources" text="Sources" margin="0 0 16px 0" />
 
 
 ### Add data collectors
@@ -366,7 +368,7 @@ MATCH p=()-->() RETURN p
 ## Kuzu
 Cocoindex natively supports Kuzu - a high performant, embedded open source graph database.
 
-<DocumentationButton href="https://cocoindex.io/docs/ops/targets#kuzu" text="Kuzu" margin="0 0 16px 0" /> 
+<DocumentationButton url="https://cocoindex.io/docs/ops/targets#kuzu" text="Kuzu" margin="0 0 16px 0" /> 
 
 The GraphDB interface in CocoIndex is standardized, you just need to **switch the configuration** without any additional code changes. CocoIndex supports exporting to Kuzu through its API server. You can bring up a Kuzu API server locally by running:
 

cocoindex-0.2.2/docs/docs/examples/examples/image_search.md

@@ -0,0 +1,270 @@
+---
+title: Image Search App with ColPali and FastAPI
+description: Build image search index with ColPali and FastAPI
+sidebar_class_name: hidden
+slug: /examples/image_search
+canonicalUrl: '/examples/image_search'
+sidebar_custom_props:
+  image: /img/examples/image_search/cover.png
+  tags: [vector-index, multi-modal]
+tags: [vector-index, multi-modal]
+---
+
+import { GitHubButton, YouTubeButton, DocumentationButton } from '../../../src/components/GitHubButton';
+
+<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/image_search" margin="0 0 24px 0" />
+
+![Image Search](/img/examples/image_search/cover.png)
+
+## Overview
+CocoIndex supports native integration with ColPali - with just a few lines of code, you embed and index images with ColPali’s late-interaction architecture. We also build a light weight image search application with FastAPI.
+
+
+## ColPali 
+
+**ColPali (Contextual Late-interaction over Patches)** is a powerful model for multimodal retrieval.
+
+It fundamentally rethinks how documents—especially visually complex or image-rich ones—are represented and searched. Instead of reducing each image or page to a single dense vector (as in traditional bi-encoders), ColPali breaks an image into many smaller patches, preserving local spatial and semantic structure. Each patch receives its own embedding, which together form a multi-vector representation of the complete document.
+
+![ColPali Architecture](/img/examples/image_search/multi_modal_architecture.png)
+
+
+## Flow Overview
+![Flow](/img/examples/image_search/flow.png)
+
+1. Ingest image files from the local filesystem
+2. Use **ColPali** to embed each image into patch-level multi-vectors
+3. Optionally extract image captions using an LLM
+4. Export the embeddings (and optional captions) to a Qdrant collection
+
+## Setup
+- [Install Postgres](https://cocoindex.io/docs/getting_started/installation#-install-postgres) if you don't have one.
+
+- Make sure Qdrant is running
+  ```
+  docker run -d -p 6334:6334 -p 6333:6333 qdrant/qdrant
+  ```
+
+
+## Add Source
+
+We start by defining a flow to read `.jpg`, `.jpeg`, and `.png` files from a local directory using `LocalFile`.
+
+```python
+@cocoindex.flow_def(name="ImageObjectEmbeddingColpali")
+def image_object_embedding_flow(flow_builder, data_scope):
+    data_scope["images"] = flow_builder.add_source(
+        cocoindex.sources.LocalFile(
+            path="img",
+            included_patterns=["*.jpg", "*.jpeg", "*.png"],
+            binary=True
+        ),
+        refresh_interval=datetime.timedelta(minutes=1),
+    )
+```
+
+The `add_source` function sets up a table with fields like `filename` and `content`. Images are automatically re-scanned every minute.
+
+<DocumentationButton url="https://cocoindex.io/docs/ops/sources#localfile" text="LocalFile" />
+
+
+## Process Each Image and Collect the Embedding
+
+We use CocoIndex's built-in `ColPaliEmbedImage` function, which returns a **multi-vector representation** for each image. Each patch receives its own vector, preserving spatial and semantic information.
+
+<DocumentationButton url="https://cocoindex.io/docs/ops/functions#colpaliembedimage" text="ColPaliEmbedImage" margin="0 0 16px 0" />
+
+```python
+img_embeddings = data_scope.add_collector()
+with data_scope["images"].row() as img:
+    img["embedding"] = img["content"].transform(cocoindex.functions.ColPaliEmbedImage(model="vidore/colpali-v1.2"))
+    collect_fields = {
+        "id": cocoindex.GeneratedField.UUID,
+        "filename": img["filename"],
+        "embedding": img["embedding"],
+    }
+    img_embeddings.collect(**collect_fields)
+```
+
+This transformation turns the raw image bytes into a list of vectors — one per patch — that can later be used for **late interaction search**. And then we collect the embeddings.
+
+![ColPali Embedding](/img/examples/image_search/embedding.png)
+
+## Export the Embeddings
+
+```python
+img_embeddings.export(
+    "img_embeddings",
+    cocoindex.targets.Qdrant(collection_name="ImageSearchColpali"),
+    primary_key_fields=["id"],
+)
+```
+
+This creates a vector collection in Qdrant that supports **multi-vector fields** — required for ColPali-style late interaction search.
+
+
+## Enable Real-Time Indexing
+
+To keep the image index up to date automatically, we wrap the flow in a `FlowLiveUpdater`:
+
+```python
+@asynccontextmanager
+async def lifespan(app: FastAPI):
+    load_dotenv()
+    cocoindex.init()
+    image_object_embedding_flow.setup(report_to_stdout=True)
+    app.state.live_updater = cocoindex.FlowLiveUpdater(image_object_embedding_flow)
+    app.state.live_updater.start()
+    yield
+```
+
+This keeps your vector index fresh as new images arrive.
+
+## Fast API Application
+
+We build a simple FastAPI application to query the index.
+
+```python
+app = FastAPI(lifespan=lifespan)
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+# Serve images from the 'img' directory at /img
+app.mount("/img", StaticFiles(directory="img"), name="img")
+```
+
+## Search API & Query the index
+
+We use `ColPaliEmbedQuery` to embed the query text into a multi-vector format.
+
+<DocumentationButton url="https://cocoindex.io/docs/ops/functions#colpaliembedquery" text="ColPaliEmbedQuery" margin="0 0 16px 0" />
+
+```python
+@cocoindex.transform_flow()
+def text_to_colpali_embedding(
+    text: cocoindex.DataSlice[str],
+) -> cocoindex.DataSlice[list[list[float]]]:
+    return text.transform(
+        cocoindex.functions.ColPaliEmbedQuery(model=COLPALI_MODEL_NAME)
+    )
+```
+Then we build a search API to query the index.
+
+```python
+# --- Search API ---
+@app.get("/search")
+def search(
+    q: str = Query(..., description="Search query"),
+    limit: int = Query(5, description="Number of results"),
+) -> Any:
+    # Get the multi-vector embedding for the query
+    query_embedding = text_to_colpali_embedding.eval(q)
+    print(
+        f"🔍 Query multi-vector shape: {len(query_embedding)} tokens x {len(query_embedding[0]) if query_embedding else 0} dims"
+    )
+
+    # Search in Qdrant with multi-vector MaxSim scoring using query_points API
+    search_results = app.state.qdrant_client.query_points(
+        collection_name=QDRANT_COLLECTION,
+        query=query_embedding,  # Multi-vector format: list[list[float]]
+        using="embedding",  # Specify the vector field name
+        limit=limit,
+        with_payload=True,
+    )
+
+    print(f"📈 Found {len(search_results.points)} results with MaxSim scoring")
+
+    return {
+        "results": [
+            {
+                "filename": result.payload["filename"],
+                "score": result.score,
+                "caption": result.payload.get("caption"),
+            }
+            for result in search_results.points
+        ]
+    }
+```
+
+## Run the application
+
+- Install dependencies:
+  ```
+  pip install -e .
+  pip install 'cocoindex[colpali]'  # Adds ColPali support
+  ```
+
+- Configure model (optional):
+  ```sh
+  # All ColVision models supported by colpali-engine are available
+  # See https://github.com/illuin-tech/colpali#list-of-colvision-models for the complete list
+
+  # ColPali models (colpali-*) - PaliGemma-based, best for general document retrieval
+  export COLPALI_MODEL="vidore/colpali-v1.2"  # Default model
+  export COLPALI_MODEL="vidore/colpali-v1.3"  # Latest version
+
+  # ColQwen2 models (colqwen-*) - Qwen2-VL-based, excellent for multilingual text (29+ languages) and general vision
+  export COLPALI_MODEL="vidore/colqwen2-v1.0"
+  export COLPALI_MODEL="vidore/colqwen2.5-v0.2"  # Latest Qwen2.5 model
+
+  # ColSmol models (colsmol-*) - Lightweight, good for resource-constrained environments
+  export COLPALI_MODEL="vidore/colSmol-256M"
+
+  # Any other ColVision models from https://github.com/illuin-tech/colpali are supported
+  ```
+
+- Run ColPali Backend:
+  ```
+  uvicorn colpali_main:app --reload --host 0.0.0.0 --port 8000
+  ```
+    :::warning
+    Note that recent Nvidia GPUs (such as the RTX 5090) are not supported by the stable PyTorch version up to 2.7.1.
+    :::
+
+    If you get this error:
+
+    ```
+    The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_61 sm_70 sm_75 sm_80 sm_86 sm_90 compute_37.
+    ```
+
+    You can install the nightly pytorch build here: https://pytorch.org/get-started/locally/
+
+    ```sh
+    pip install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu129
+    ```
+- Run Frontend:
+  ```
+  cd frontend
+  npm install
+  npm run dev
+  ```
+
+ Go to `http://localhost:5173` to search. The frontend works with both backends identically.
+
+ ![Result](/img/examples/image_search/result.png)
+
+## CLIP Model & Comparison with ColPali
+We've also had a similar application built with CLIP model.
+
+<DocumentationButton url="https://cocoindex.io/blogs/live-image-search" text="Image Search App with CLIP" margin="0 0 16px 0" />
+
+In general,
+- CLIP: Faster, good for general image-text matching
+- ColPali: More accurate for document images and text-heavy content, supports multi-vector late interaction for better precision
+
+## Connect to Any Data Source
+
+One of CocoIndex’s core strengths is its ability to connect to your existing data sources and automatically keep your index fresh. Beyond local files, CocoIndex natively supports source connectors including:
+
+- Google Drive
+- Amazon S3 / SQS
+- Azure Blob Storage
+
+<DocumentationButton url="https://cocoindex.io/docs/ops/sources" text="Sources" margin="0 0 16px 0" />
+
+Once connected, CocoIndex continuously watches for changes — new uploads, updates, or deletions — and applies them to your index in real time.

{cocoindex-0.2.0 → cocoindex-0.2.2}/docs/docs/examples/examples/manual_extraction.md

@@ -12,7 +12,9 @@ tags: [structured-data-extraction, data-mapping]
 
 import { GitHubButton, YouTubeButton, DocumentationButton } from '../../../src/components/GitHubButton';
 
-<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/manuals_llm_extraction"/>
+<GitHubButton url="https://github.com/cocoindex-io/cocoindex/tree/main/examples/manuals_llm_extraction" margin="0 0 24px 0" />
+
+![Manual Extraction](/img/examples/manual_extraction/cover.png)
 
 ## Overview
 This example shows how to extract structured data from Python Manuals using Ollama.
@@ -36,11 +38,11 @@ This example shows how to extract structured data from Python Manuals using Olla
     ollama pull llama3.2
     ```
 
-    <DocumentationButton href="https://cocoindex.io/docs/ai/llm#ollama" text="Ollama" margin="0 0 16px 0" />
+    <DocumentationButton url="https://cocoindex.io/docs/ai/llm#ollama" text="Ollama" margin="0 0 16px 0" />
 
     Alternatively, CocoIndex have native support for Gemini, Ollama, LiteLLM. You can choose your favorite LLM provider and work completely on-premises.
 
-    <DocumentationButton href="https://cocoindex.io/docs/ai/llm" text="LLM" margin="0 0 16px 0" />
+    <DocumentationButton url="https://cocoindex.io/docs/ai/llm" text="LLM" margin="0 0 16px 0" />
 
 ## Add Source
 Let's add Python docs as a source.
@@ -64,7 +66,7 @@ def manual_extraction_flow(
 - `filename` (key, type: `str`): the filename of the file, e.g. `dir1/file1.md`
 - `content` (type: `str` if `binary` is `False`, otherwise `bytes`): the content of the file
 
-<DocumentationButton href="https://cocoindex.io/docs/ops/sources" text="LocalFile" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/ops/sources" text="LocalFile" margin="0 0 16px 0" />
 
 ## Parse Markdown
 
@@ -97,7 +99,7 @@ class PdfToMarkdownExecutor:
 ```
 You may wonder why we want to define a spec + executor (instead of using a standalone function) here. The main reason is there're some heavy preparation work (initialize the parser) needs to be done before being ready to process real data.
 
-<DocumentationButton href="https://cocoindex.io/docs/custom_ops/custom_functions" text="Custom Function" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/custom_ops/custom_functions" text="Custom Function" margin="0 0 16px 0" />
 
 Plug in the function to the flow.
 
@@ -160,7 +162,7 @@ with data_scope["documents"].row() as doc:
             instruction="Please extract Python module information from the manual."))
 ```
 
-<DocumentationButton href="https://cocoindex.io/docs/core/functions#extractbyllm" text="ExtractByLlm" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/core/functions#extractbyllm" text="ExtractByLlm" margin="0 0 16px 0" />
 
 ![ExtractByLlm](/img/examples/manual_extraction/extraction.png)
 
@@ -194,7 +196,7 @@ with data_scope["documents"].row() as doc:
     doc["module_summary"] = doc["module_info"].transform(summarize_module)
 ```
 
-<DocumentationButton href="https://cocoindex.io/docs/custom_ops/custom_functions" text="Custom Function" margin="0 0 16px 0" />
+<DocumentationButton url="https://cocoindex.io/docs/custom_ops/custom_functions" text="Custom Function" margin="0 0 16px 0" />
 
 ![Summarize Module](/img/examples/manual_extraction/summary.png)