@fllf/agent-sdk 0.1.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +240 -0
- package/dist/agent/Agent.d.ts +34 -0
- package/dist/agent/Agent.d.ts.map +1 -0
- package/dist/agent/Agent.js +168 -0
- package/dist/agent/Agent.js.map +1 -0
- package/dist/agent/createAgent.d.ts +4 -0
- package/dist/agent/createAgent.d.ts.map +1 -0
- package/dist/agent/createAgent.js +8 -0
- package/dist/agent/createAgent.js.map +1 -0
- package/dist/agent/index.d.ts +4 -0
- package/dist/agent/index.d.ts.map +1 -0
- package/dist/agent/index.js +8 -0
- package/dist/agent/index.js.map +1 -0
- package/dist/agent/types.d.ts +23 -0
- package/dist/agent/types.d.ts.map +1 -0
- package/dist/agent/types.js +3 -0
- package/dist/agent/types.js.map +1 -0
- package/dist/config/config.d.ts +35 -0
- package/dist/config/config.d.ts.map +1 -0
- package/dist/config/config.js +123 -0
- package/dist/config/config.js.map +1 -0
- package/dist/config/index.d.ts +3 -0
- package/dist/config/index.d.ts.map +1 -0
- package/dist/config/index.js +6 -0
- package/dist/config/index.js.map +1 -0
- package/dist/executors/base.d.ts +44 -0
- package/dist/executors/base.d.ts.map +1 -0
- package/dist/executors/base.js +3 -0
- package/dist/executors/base.js.map +1 -0
- package/dist/executors/index.d.ts +7 -0
- package/dist/executors/index.d.ts.map +1 -0
- package/dist/executors/index.js +10 -0
- package/dist/executors/index.js.map +1 -0
- package/dist/executors/rag-executor.d.ts +16 -0
- package/dist/executors/rag-executor.d.ts.map +1 -0
- package/dist/executors/rag-executor.js +120 -0
- package/dist/executors/rag-executor.js.map +1 -0
- package/dist/executors/simple-chat-executor.d.ts +5 -0
- package/dist/executors/simple-chat-executor.d.ts.map +1 -0
- package/dist/executors/simple-chat-executor.js +77 -0
- package/dist/executors/simple-chat-executor.js.map +1 -0
- package/dist/executors/tool-calling-executor.d.ts +10 -0
- package/dist/executors/tool-calling-executor.d.ts.map +1 -0
- package/dist/executors/tool-calling-executor.js +151 -0
- package/dist/executors/tool-calling-executor.js.map +1 -0
- package/dist/history/base.d.ts +16 -0
- package/dist/history/base.d.ts.map +1 -0
- package/dist/history/base.js +10 -0
- package/dist/history/base.js.map +1 -0
- package/dist/history/in-memory.d.ts +26 -0
- package/dist/history/in-memory.d.ts.map +1 -0
- package/dist/history/in-memory.js +88 -0
- package/dist/history/in-memory.js.map +1 -0
- package/dist/history/index.d.ts +5 -0
- package/dist/history/index.d.ts.map +1 -0
- package/dist/history/index.js +10 -0
- package/dist/history/index.js.map +1 -0
- package/dist/index.d.ts +10 -0
- package/dist/index.d.ts.map +1 -0
- package/dist/index.js +26 -0
- package/dist/index.js.map +1 -0
- package/dist/llm/LLM.d.ts +23 -0
- package/dist/llm/LLM.d.ts.map +1 -0
- package/dist/llm/LLM.js +404 -0
- package/dist/llm/LLM.js.map +1 -0
- package/dist/llm/errors.d.ts +43 -0
- package/dist/llm/errors.d.ts.map +1 -0
- package/dist/llm/errors.js +128 -0
- package/dist/llm/errors.js.map +1 -0
- package/dist/llm/factory.d.ts +9 -0
- package/dist/llm/factory.d.ts.map +1 -0
- package/dist/llm/factory.js +28 -0
- package/dist/llm/factory.js.map +1 -0
- package/dist/llm/index.d.ts +6 -0
- package/dist/llm/index.d.ts.map +1 -0
- package/dist/llm/index.js +18 -0
- package/dist/llm/index.js.map +1 -0
- package/dist/llm/providers/base.d.ts +3 -0
- package/dist/llm/providers/base.d.ts.map +1 -0
- package/dist/llm/providers/base.js +3 -0
- package/dist/llm/providers/base.js.map +1 -0
- package/dist/llm/providers/local.d.ts +6 -0
- package/dist/llm/providers/local.d.ts.map +1 -0
- package/dist/llm/providers/local.js +19 -0
- package/dist/llm/providers/local.js.map +1 -0
- package/dist/llm/providers/openai-compatible.d.ts +13 -0
- package/dist/llm/providers/openai-compatible.d.ts.map +1 -0
- package/dist/llm/providers/openai-compatible.js +176 -0
- package/dist/llm/providers/openai-compatible.js.map +1 -0
- package/dist/llm/providers/openai.d.ts +6 -0
- package/dist/llm/providers/openai.d.ts.map +1 -0
- package/dist/llm/providers/openai.js +23 -0
- package/dist/llm/providers/openai.js.map +1 -0
- package/dist/llm/types.d.ts +116 -0
- package/dist/llm/types.d.ts.map +1 -0
- package/dist/llm/types.js +3 -0
- package/dist/llm/types.js.map +1 -0
- package/dist/messages/index.d.ts +3 -0
- package/dist/messages/index.d.ts.map +1 -0
- package/dist/messages/index.js +6 -0
- package/dist/messages/index.js.map +1 -0
- package/dist/messages/message.d.ts +25 -0
- package/dist/messages/message.d.ts.map +1 -0
- package/dist/messages/message.js +110 -0
- package/dist/messages/message.js.map +1 -0
- package/dist/messages/types.d.ts +20 -0
- package/dist/messages/types.d.ts.map +1 -0
- package/dist/messages/types.js +3 -0
- package/dist/messages/types.js.map +1 -0
- package/dist/observability/console-observer.d.ts +10 -0
- package/dist/observability/console-observer.d.ts.map +1 -0
- package/dist/observability/console-observer.js +28 -0
- package/dist/observability/console-observer.js.map +1 -0
- package/dist/observability/index.d.ts +5 -0
- package/dist/observability/index.d.ts.map +1 -0
- package/dist/observability/index.js +11 -0
- package/dist/observability/index.js.map +1 -0
- package/dist/observability/observer.d.ts +8 -0
- package/dist/observability/observer.d.ts.map +1 -0
- package/dist/observability/observer.js +46 -0
- package/dist/observability/observer.js.map +1 -0
- package/dist/observability/types.d.ts +34 -0
- package/dist/observability/types.d.ts.map +1 -0
- package/dist/observability/types.js +3 -0
- package/dist/observability/types.js.map +1 -0
- package/dist/rag/chunking/auto-chunker.d.ts +25 -0
- package/dist/rag/chunking/auto-chunker.d.ts.map +1 -0
- package/dist/rag/chunking/auto-chunker.js +101 -0
- package/dist/rag/chunking/auto-chunker.js.map +1 -0
- package/dist/rag/chunking/chunker.d.ts +44 -0
- package/dist/rag/chunking/chunker.d.ts.map +1 -0
- package/dist/rag/chunking/chunker.js +162 -0
- package/dist/rag/chunking/chunker.js.map +1 -0
- package/dist/rag/chunking/index.d.ts +6 -0
- package/dist/rag/chunking/index.d.ts.map +1 -0
- package/dist/rag/chunking/index.js +22 -0
- package/dist/rag/chunking/index.js.map +1 -0
- package/dist/rag/chunking/markdown-chunker.d.ts +13 -0
- package/dist/rag/chunking/markdown-chunker.d.ts.map +1 -0
- package/dist/rag/chunking/markdown-chunker.js +144 -0
- package/dist/rag/chunking/markdown-chunker.js.map +1 -0
- package/dist/rag/chunking/parent-child.d.ts +16 -0
- package/dist/rag/chunking/parent-child.d.ts.map +1 -0
- package/dist/rag/chunking/parent-child.js +76 -0
- package/dist/rag/chunking/parent-child.js.map +1 -0
- package/dist/rag/chunking/recursive-chunker.d.ts +11 -0
- package/dist/rag/chunking/recursive-chunker.d.ts.map +1 -0
- package/dist/rag/chunking/recursive-chunker.js +27 -0
- package/dist/rag/chunking/recursive-chunker.js.map +1 -0
- package/dist/rag/embeddings/embedder.d.ts +24 -0
- package/dist/rag/embeddings/embedder.d.ts.map +1 -0
- package/dist/rag/embeddings/embedder.js +73 -0
- package/dist/rag/embeddings/embedder.js.map +1 -0
- package/dist/rag/embeddings/embedding-cache.d.ts +40 -0
- package/dist/rag/embeddings/embedding-cache.d.ts.map +1 -0
- package/dist/rag/embeddings/embedding-cache.js +118 -0
- package/dist/rag/embeddings/embedding-cache.js.map +1 -0
- package/dist/rag/embeddings/fake-embedder.d.ts +20 -0
- package/dist/rag/embeddings/fake-embedder.d.ts.map +1 -0
- package/dist/rag/embeddings/fake-embedder.js +55 -0
- package/dist/rag/embeddings/fake-embedder.js.map +1 -0
- package/dist/rag/embeddings/index.d.ts +5 -0
- package/dist/rag/embeddings/index.d.ts.map +1 -0
- package/dist/rag/embeddings/index.js +21 -0
- package/dist/rag/embeddings/index.js.map +1 -0
- package/dist/rag/embeddings/openai-compatible-embedder.d.ts +46 -0
- package/dist/rag/embeddings/openai-compatible-embedder.d.ts.map +1 -0
- package/dist/rag/embeddings/openai-compatible-embedder.js +145 -0
- package/dist/rag/embeddings/openai-compatible-embedder.js.map +1 -0
- package/dist/rag/generation/context-builder.d.ts +22 -0
- package/dist/rag/generation/context-builder.d.ts.map +1 -0
- package/dist/rag/generation/context-builder.js +166 -0
- package/dist/rag/generation/context-builder.js.map +1 -0
- package/dist/rag/generation/generator.d.ts +25 -0
- package/dist/rag/generation/generator.d.ts.map +1 -0
- package/dist/rag/generation/generator.js +185 -0
- package/dist/rag/generation/generator.js.map +1 -0
- package/dist/rag/generation/index.d.ts +4 -0
- package/dist/rag/generation/index.d.ts.map +1 -0
- package/dist/rag/generation/index.js +20 -0
- package/dist/rag/generation/index.js.map +1 -0
- package/dist/rag/generation/verifier.d.ts +13 -0
- package/dist/rag/generation/verifier.d.ts.map +1 -0
- package/dist/rag/generation/verifier.js +43 -0
- package/dist/rag/generation/verifier.js.map +1 -0
- package/dist/rag/index.d.ts +9 -0
- package/dist/rag/index.d.ts.map +1 -0
- package/dist/rag/index.js +25 -0
- package/dist/rag/index.js.map +1 -0
- package/dist/rag/ingestion/index.d.ts +4 -0
- package/dist/rag/ingestion/index.d.ts.map +1 -0
- package/dist/rag/ingestion/index.js +20 -0
- package/dist/rag/ingestion/index.js.map +1 -0
- package/dist/rag/ingestion/loaders.d.ts +69 -0
- package/dist/rag/ingestion/loaders.d.ts.map +1 -0
- package/dist/rag/ingestion/loaders.js +653 -0
- package/dist/rag/ingestion/loaders.js.map +1 -0
- package/dist/rag/ingestion/metadata.d.ts +31 -0
- package/dist/rag/ingestion/metadata.d.ts.map +1 -0
- package/dist/rag/ingestion/metadata.js +81 -0
- package/dist/rag/ingestion/metadata.js.map +1 -0
- package/dist/rag/ingestion/normalizer.d.ts +20 -0
- package/dist/rag/ingestion/normalizer.d.ts.map +1 -0
- package/dist/rag/ingestion/normalizer.js +161 -0
- package/dist/rag/ingestion/normalizer.js.map +1 -0
- package/dist/rag/pipeline.d.ts +62 -0
- package/dist/rag/pipeline.d.ts.map +1 -0
- package/dist/rag/pipeline.js +75 -0
- package/dist/rag/pipeline.js.map +1 -0
- package/dist/rag/retrieval/dense-retriever.d.ts +17 -0
- package/dist/rag/retrieval/dense-retriever.d.ts.map +1 -0
- package/dist/rag/retrieval/dense-retriever.js +21 -0
- package/dist/rag/retrieval/dense-retriever.js.map +1 -0
- package/dist/rag/retrieval/fusion.d.ts +12 -0
- package/dist/rag/retrieval/fusion.d.ts.map +1 -0
- package/dist/rag/retrieval/fusion.js +54 -0
- package/dist/rag/retrieval/fusion.js.map +1 -0
- package/dist/rag/retrieval/http-rerank-model-client.d.ts +41 -0
- package/dist/rag/retrieval/http-rerank-model-client.d.ts.map +1 -0
- package/dist/rag/retrieval/http-rerank-model-client.js +130 -0
- package/dist/rag/retrieval/http-rerank-model-client.js.map +1 -0
- package/dist/rag/retrieval/hybrid-retriever.d.ts +22 -0
- package/dist/rag/retrieval/hybrid-retriever.d.ts.map +1 -0
- package/dist/rag/retrieval/hybrid-retriever.js +49 -0
- package/dist/rag/retrieval/hybrid-retriever.js.map +1 -0
- package/dist/rag/retrieval/index.d.ts +8 -0
- package/dist/rag/retrieval/index.d.ts.map +1 -0
- package/dist/rag/retrieval/index.js +24 -0
- package/dist/rag/retrieval/index.js.map +1 -0
- package/dist/rag/retrieval/reranker.d.ts +30 -0
- package/dist/rag/retrieval/reranker.d.ts.map +1 -0
- package/dist/rag/retrieval/reranker.js +85 -0
- package/dist/rag/retrieval/reranker.js.map +1 -0
- package/dist/rag/retrieval/sparse-retriever.d.ts +14 -0
- package/dist/rag/retrieval/sparse-retriever.d.ts.map +1 -0
- package/dist/rag/retrieval/sparse-retriever.js +18 -0
- package/dist/rag/retrieval/sparse-retriever.js.map +1 -0
- package/dist/rag/retrieval/types.d.ts +11 -0
- package/dist/rag/retrieval/types.d.ts.map +1 -0
- package/dist/rag/retrieval/types.js +10 -0
- package/dist/rag/retrieval/types.js.map +1 -0
- package/dist/rag/retrieval/utils.d.ts +8 -0
- package/dist/rag/retrieval/utils.d.ts.map +1 -0
- package/dist/rag/retrieval/utils.js +114 -0
- package/dist/rag/retrieval/utils.js.map +1 -0
- package/dist/rag/stores/in-memory-document-store.d.ts +14 -0
- package/dist/rag/stores/in-memory-document-store.d.ts.map +1 -0
- package/dist/rag/stores/in-memory-document-store.js +64 -0
- package/dist/rag/stores/in-memory-document-store.js.map +1 -0
- package/dist/rag/stores/in-memory-keyword-store.d.ts +10 -0
- package/dist/rag/stores/in-memory-keyword-store.d.ts.map +1 -0
- package/dist/rag/stores/in-memory-keyword-store.js +99 -0
- package/dist/rag/stores/in-memory-keyword-store.js.map +1 -0
- package/dist/rag/stores/in-memory-vector-store.d.ts +14 -0
- package/dist/rag/stores/in-memory-vector-store.d.ts.map +1 -0
- package/dist/rag/stores/in-memory-vector-store.js +61 -0
- package/dist/rag/stores/in-memory-vector-store.js.map +1 -0
- package/dist/rag/stores/index.d.ts +6 -0
- package/dist/rag/stores/index.d.ts.map +1 -0
- package/dist/rag/stores/index.js +22 -0
- package/dist/rag/stores/index.js.map +1 -0
- package/dist/rag/stores/postgres/index.d.ts +6 -0
- package/dist/rag/stores/postgres/index.d.ts.map +1 -0
- package/dist/rag/stores/postgres/index.js +22 -0
- package/dist/rag/stores/postgres/index.js.map +1 -0
- package/dist/rag/stores/postgres/pg-vector-store.d.ts +16 -0
- package/dist/rag/stores/postgres/pg-vector-store.d.ts.map +1 -0
- package/dist/rag/stores/postgres/pg-vector-store.js +77 -0
- package/dist/rag/stores/postgres/pg-vector-store.js.map +1 -0
- package/dist/rag/stores/postgres/postgres-document-store.d.ts +15 -0
- package/dist/rag/stores/postgres/postgres-document-store.d.ts.map +1 -0
- package/dist/rag/stores/postgres/postgres-document-store.js +174 -0
- package/dist/rag/stores/postgres/postgres-document-store.js.map +1 -0
- package/dist/rag/stores/postgres/postgres-keyword-store.d.ts +15 -0
- package/dist/rag/stores/postgres/postgres-keyword-store.d.ts.map +1 -0
- package/dist/rag/stores/postgres/postgres-keyword-store.js +163 -0
- package/dist/rag/stores/postgres/postgres-keyword-store.js.map +1 -0
- package/dist/rag/stores/postgres/schema.d.ts +8 -0
- package/dist/rag/stores/postgres/schema.d.ts.map +1 -0
- package/dist/rag/stores/postgres/schema.js +91 -0
- package/dist/rag/stores/postgres/schema.js.map +1 -0
- package/dist/rag/stores/postgres/sql.d.ts +23 -0
- package/dist/rag/stores/postgres/sql.d.ts.map +1 -0
- package/dist/rag/stores/postgres/sql.js +177 -0
- package/dist/rag/stores/postgres/sql.js.map +1 -0
- package/dist/rag/stores/postgres/types.d.ts +17 -0
- package/dist/rag/stores/postgres/types.d.ts.map +1 -0
- package/dist/rag/stores/postgres/types.js +3 -0
- package/dist/rag/stores/postgres/types.js.map +1 -0
- package/dist/rag/stores/types.d.ts +36 -0
- package/dist/rag/stores/types.d.ts.map +1 -0
- package/dist/rag/stores/types.js +3 -0
- package/dist/rag/stores/types.js.map +1 -0
- package/dist/rag/stores/utils.d.ts +11 -0
- package/dist/rag/stores/utils.d.ts.map +1 -0
- package/dist/rag/stores/utils.js +200 -0
- package/dist/rag/stores/utils.js.map +1 -0
- package/dist/rag/types.d.ts +174 -0
- package/dist/rag/types.d.ts.map +1 -0
- package/dist/rag/types.js +3 -0
- package/dist/rag/types.js.map +1 -0
- package/dist/tools/base.d.ts +28 -0
- package/dist/tools/base.d.ts.map +1 -0
- package/dist/tools/base.js +68 -0
- package/dist/tools/base.js.map +1 -0
- package/dist/tools/builtin/advancedSearchTool.d.ts +24 -0
- package/dist/tools/builtin/advancedSearchTool.d.ts.map +1 -0
- package/dist/tools/builtin/advancedSearchTool.js +134 -0
- package/dist/tools/builtin/advancedSearchTool.js.map +1 -0
- package/dist/tools/builtin/ragSearchTool.d.ts +29 -0
- package/dist/tools/builtin/ragSearchTool.d.ts.map +1 -0
- package/dist/tools/builtin/ragSearchTool.js +91 -0
- package/dist/tools/builtin/ragSearchTool.js.map +1 -0
- package/dist/tools/executor.d.ts +10 -0
- package/dist/tools/executor.d.ts.map +1 -0
- package/dist/tools/executor.js +86 -0
- package/dist/tools/executor.js.map +1 -0
- package/dist/tools/index.d.ts +9 -0
- package/dist/tools/index.d.ts.map +1 -0
- package/dist/tools/index.js +16 -0
- package/dist/tools/index.js.map +1 -0
- package/dist/tools/registry.d.ts +16 -0
- package/dist/tools/registry.d.ts.map +1 -0
- package/dist/tools/registry.js +35 -0
- package/dist/tools/registry.js.map +1 -0
- package/package.json +43 -0
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../../src/messages/types.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,cAAc,CAAC;AAEjD,MAAM,MAAM,WAAW,GAAG,MAAM,GAAG,WAAW,GAAG,QAAQ,GAAG,MAAM,CAAC;AAEnE,MAAM,WAAW,eAAe;IAC5B;;;OAGG;IACH,UAAU,CAAC,EAAE,MAAM,CAAC;IAEpB;;;OAGG;IACH,SAAS,CAAC,EAAE,YAAY,EAAE,CAAC;IAE3B,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC;CAC1B;AAED,MAAM,WAAW,cAAc;IAC3B,SAAS,CAAC,EAAE,IAAI,CAAC;IACjB,QAAQ,CAAC,EAAE,eAAe,CAAC;CAC9B"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.js","sourceRoot":"","sources":["../../src/messages/types.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,10 @@
|
|
|
1
|
+
import type { Observer, RuntimeEvent } from './types';
|
|
2
|
+
export interface ConsoleObserverOptions {
|
|
3
|
+
includeMetadata?: boolean;
|
|
4
|
+
}
|
|
5
|
+
export declare class ConsoleObserver implements Observer {
|
|
6
|
+
private readonly options;
|
|
7
|
+
constructor(options?: ConsoleObserverOptions);
|
|
8
|
+
onEvent(event: RuntimeEvent): void;
|
|
9
|
+
}
|
|
10
|
+
//# sourceMappingURL=console-observer.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"console-observer.d.ts","sourceRoot":"","sources":["../../src/observability/console-observer.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,QAAQ,EAAE,YAAY,EAAE,MAAM,SAAS,CAAC;AAEtD,MAAM,WAAW,sBAAsB;IACnC,eAAe,CAAC,EAAE,OAAO,CAAC;CAC7B;AAED,qBAAa,eAAgB,YAAW,QAAQ;IAChC,OAAO,CAAC,QAAQ,CAAC,OAAO;gBAAP,OAAO,GAAE,sBAA2B;IAEjE,OAAO,CAAC,KAAK,EAAE,YAAY,GAAG,IAAI;CAmBrC"}
|
|
@@ -0,0 +1,28 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.ConsoleObserver = void 0;
|
|
4
|
+
class ConsoleObserver {
|
|
5
|
+
options;
|
|
6
|
+
constructor(options = {}) {
|
|
7
|
+
this.options = options;
|
|
8
|
+
}
|
|
9
|
+
onEvent(event) {
|
|
10
|
+
const parts = [
|
|
11
|
+
`[${event.type}]`,
|
|
12
|
+
`run=${event.runId}`,
|
|
13
|
+
event.agentName ? `agent=${event.agentName}` : undefined,
|
|
14
|
+
event.sessionId ? `session=${event.sessionId}` : undefined,
|
|
15
|
+
event.provider ? `provider=${event.provider}` : undefined,
|
|
16
|
+
event.model ? `model=${event.model}` : undefined,
|
|
17
|
+
event.toolName ? `tool=${event.toolName}` : undefined,
|
|
18
|
+
event.durationMs !== undefined ? `duration=${event.durationMs}ms` : undefined,
|
|
19
|
+
event.error ? `error="${event.error.message}"` : undefined,
|
|
20
|
+
].filter(Boolean);
|
|
21
|
+
console.log(parts.join(' '));
|
|
22
|
+
if (this.options.includeMetadata && event.metadata) {
|
|
23
|
+
console.log(event.metadata);
|
|
24
|
+
}
|
|
25
|
+
}
|
|
26
|
+
}
|
|
27
|
+
exports.ConsoleObserver = ConsoleObserver;
|
|
28
|
+
//# sourceMappingURL=console-observer.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"console-observer.js","sourceRoot":"","sources":["../../src/observability/console-observer.ts"],"names":[],"mappings":";;;AAMA,MAAa,eAAe;IACK;IAA7B,YAA6B,UAAkC,EAAE;QAApC,YAAO,GAAP,OAAO,CAA6B;IAAG,CAAC;IAErE,OAAO,CAAC,KAAmB;QACvB,MAAM,KAAK,GAAG;YACV,IAAI,KAAK,CAAC,IAAI,GAAG;YACjB,OAAO,KAAK,CAAC,KAAK,EAAE;YACpB,KAAK,CAAC,SAAS,CAAC,CAAC,CAAC,SAAS,KAAK,CAAC,SAAS,EAAE,CAAC,CAAC,CAAC,SAAS;YACxD,KAAK,CAAC,SAAS,CAAC,CAAC,CAAC,WAAW,KAAK,CAAC,SAAS,EAAE,CAAC,CAAC,CAAC,SAAS;YAC1D,KAAK,CAAC,QAAQ,CAAC,CAAC,CAAC,YAAY,KAAK,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC,SAAS;YACzD,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,SAAS,KAAK,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC,SAAS;YAChD,KAAK,CAAC,QAAQ,CAAC,CAAC,CAAC,QAAQ,KAAK,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC,SAAS;YACrD,KAAK,CAAC,UAAU,KAAK,SAAS,CAAC,CAAC,CAAC,YAAY,KAAK,CAAC,UAAU,IAAI,CAAC,CAAC,CAAC,SAAS;YAC7E,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,UAAU,KAAK,CAAC,KAAK,CAAC,OAAO,GAAG,CAAC,CAAC,CAAC,SAAS;SAC7D,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC;QAElB,OAAO,CAAC,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;QAE7B,IAAI,IAAI,CAAC,OAAO,CAAC,eAAe,IAAI,KAAK,CAAC,QAAQ,EAAE,CAAC;YACjD,OAAO,CAAC,GAAG,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC;QAChC,CAAC;IACL,CAAC;CACJ;AAtBD,0CAsBC"}
|
|
@@ -0,0 +1,5 @@
|
|
|
1
|
+
export { ConsoleObserver } from './console-observer';
|
|
2
|
+
export type { ConsoleObserverOptions } from './console-observer';
|
|
3
|
+
export { NoopObserver, createRunId, emitRuntimeEvent, toObservedError, } from './observer';
|
|
4
|
+
export type { ObservedError, Observer, RuntimeEvent, RuntimeEventType, TraceContext, } from './types';
|
|
5
|
+
//# sourceMappingURL=index.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/observability/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,eAAe,EAAE,MAAM,oBAAoB,CAAC;AACrD,YAAY,EAAE,sBAAsB,EAAE,MAAM,oBAAoB,CAAC;AACjE,OAAO,EACH,YAAY,EACZ,WAAW,EACX,gBAAgB,EAChB,eAAe,GAClB,MAAM,YAAY,CAAC;AACpB,YAAY,EACR,aAAa,EACb,QAAQ,EACR,YAAY,EACZ,gBAAgB,EAChB,YAAY,GACf,MAAM,SAAS,CAAC"}
|
|
@@ -0,0 +1,11 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.toObservedError = exports.emitRuntimeEvent = exports.createRunId = exports.NoopObserver = exports.ConsoleObserver = void 0;
|
|
4
|
+
var console_observer_1 = require("./console-observer");
|
|
5
|
+
Object.defineProperty(exports, "ConsoleObserver", { enumerable: true, get: function () { return console_observer_1.ConsoleObserver; } });
|
|
6
|
+
var observer_1 = require("./observer");
|
|
7
|
+
Object.defineProperty(exports, "NoopObserver", { enumerable: true, get: function () { return observer_1.NoopObserver; } });
|
|
8
|
+
Object.defineProperty(exports, "createRunId", { enumerable: true, get: function () { return observer_1.createRunId; } });
|
|
9
|
+
Object.defineProperty(exports, "emitRuntimeEvent", { enumerable: true, get: function () { return observer_1.emitRuntimeEvent; } });
|
|
10
|
+
Object.defineProperty(exports, "toObservedError", { enumerable: true, get: function () { return observer_1.toObservedError; } });
|
|
11
|
+
//# sourceMappingURL=index.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.js","sourceRoot":"","sources":["../../src/observability/index.ts"],"names":[],"mappings":";;;AAAA,uDAAqD;AAA5C,mHAAA,eAAe,OAAA;AAExB,uCAKoB;AAJhB,wGAAA,YAAY,OAAA;AACZ,uGAAA,WAAW,OAAA;AACX,4GAAA,gBAAgB,OAAA;AAChB,2GAAA,eAAe,OAAA"}
|
|
@@ -0,0 +1,8 @@
|
|
|
1
|
+
import type { ObservedError, Observer, RuntimeEvent } from './types';
|
|
2
|
+
export declare class NoopObserver implements Observer {
|
|
3
|
+
onEvent(_event: RuntimeEvent): void;
|
|
4
|
+
}
|
|
5
|
+
export declare function emitRuntimeEvent(observer: Observer | undefined, event: RuntimeEvent): Promise<void>;
|
|
6
|
+
export declare function createRunId(): string;
|
|
7
|
+
export declare function toObservedError(error: unknown): ObservedError;
|
|
8
|
+
//# sourceMappingURL=observer.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"observer.d.ts","sourceRoot":"","sources":["../../src/observability/observer.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,aAAa,EAAE,QAAQ,EAAE,YAAY,EAAE,MAAM,SAAS,CAAC;AAErE,qBAAa,YAAa,YAAW,QAAQ;IACzC,OAAO,CAAC,MAAM,EAAE,YAAY,GAAG,IAAI;CAGtC;AAED,wBAAsB,gBAAgB,CAClC,QAAQ,EAAE,QAAQ,GAAG,SAAS,EAC9B,KAAK,EAAE,YAAY,GACpB,OAAO,CAAC,IAAI,CAAC,CAUf;AAED,wBAAgB,WAAW,IAAI,MAAM,CAGpC;AAED,wBAAgB,eAAe,CAAC,KAAK,EAAE,OAAO,GAAG,aAAa,CAc7D"}
|
|
@@ -0,0 +1,46 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.NoopObserver = void 0;
|
|
4
|
+
exports.emitRuntimeEvent = emitRuntimeEvent;
|
|
5
|
+
exports.createRunId = createRunId;
|
|
6
|
+
exports.toObservedError = toObservedError;
|
|
7
|
+
class NoopObserver {
|
|
8
|
+
onEvent(_event) {
|
|
9
|
+
// 默认 observer 什么都不做,库代码不应该主动污染用户控制台。
|
|
10
|
+
}
|
|
11
|
+
}
|
|
12
|
+
exports.NoopObserver = NoopObserver;
|
|
13
|
+
async function emitRuntimeEvent(observer, event) {
|
|
14
|
+
if (!observer) {
|
|
15
|
+
return;
|
|
16
|
+
}
|
|
17
|
+
try {
|
|
18
|
+
await observer.onEvent(event);
|
|
19
|
+
}
|
|
20
|
+
catch {
|
|
21
|
+
// Observer 是旁路能力,日志/监控失败不应该影响 Agent 主流程。
|
|
22
|
+
}
|
|
23
|
+
}
|
|
24
|
+
function createRunId() {
|
|
25
|
+
const randomId = globalThis.crypto?.randomUUID?.();
|
|
26
|
+
return randomId ? `run_${randomId}` : `run_${Date.now()}_${Math.random().toString(36).slice(2)}`;
|
|
27
|
+
}
|
|
28
|
+
function toObservedError(error) {
|
|
29
|
+
if (error instanceof Error) {
|
|
30
|
+
return {
|
|
31
|
+
name: error.name,
|
|
32
|
+
message: error.message,
|
|
33
|
+
code: getErrorCode(error),
|
|
34
|
+
stack: error.stack,
|
|
35
|
+
};
|
|
36
|
+
}
|
|
37
|
+
return {
|
|
38
|
+
name: 'Error',
|
|
39
|
+
message: String(error),
|
|
40
|
+
};
|
|
41
|
+
}
|
|
42
|
+
function getErrorCode(error) {
|
|
43
|
+
const maybeCode = error.code;
|
|
44
|
+
return typeof maybeCode === 'string' ? maybeCode : undefined;
|
|
45
|
+
}
|
|
46
|
+
//# sourceMappingURL=observer.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"observer.js","sourceRoot":"","sources":["../../src/observability/observer.ts"],"names":[],"mappings":";;;AAQA,4CAaC;AAED,kCAGC;AAED,0CAcC;AAxCD,MAAa,YAAY;IACrB,OAAO,CAAC,MAAoB;QACxB,qCAAqC;IACzC,CAAC;CACJ;AAJD,oCAIC;AAEM,KAAK,UAAU,gBAAgB,CAClC,QAA8B,EAC9B,KAAmB;IAEnB,IAAI,CAAC,QAAQ,EAAE,CAAC;QACZ,OAAO;IACX,CAAC;IAED,IAAI,CAAC;QACD,MAAM,QAAQ,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;IAClC,CAAC;IAAC,MAAM,CAAC;QACL,yCAAyC;IAC7C,CAAC;AACL,CAAC;AAED,SAAgB,WAAW;IACvB,MAAM,QAAQ,GAAG,UAAU,CAAC,MAAM,EAAE,UAAU,EAAE,EAAE,CAAC;IACnD,OAAO,QAAQ,CAAC,CAAC,CAAC,OAAO,QAAQ,EAAE,CAAC,CAAC,CAAC,OAAO,IAAI,CAAC,GAAG,EAAE,IAAI,IAAI,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,CAAC;AACrG,CAAC;AAED,SAAgB,eAAe,CAAC,KAAc;IAC1C,IAAI,KAAK,YAAY,KAAK,EAAE,CAAC;QACzB,OAAO;YACH,IAAI,EAAE,KAAK,CAAC,IAAI;YAChB,OAAO,EAAE,KAAK,CAAC,OAAO;YACtB,IAAI,EAAE,YAAY,CAAC,KAAK,CAAC;YACzB,KAAK,EAAE,KAAK,CAAC,KAAK;SACrB,CAAC;IACN,CAAC;IAED,OAAO;QACH,IAAI,EAAE,OAAO;QACb,OAAO,EAAE,MAAM,CAAC,KAAK,CAAC;KACzB,CAAC;AACN,CAAC;AAED,SAAS,YAAY,CAAC,KAAY;IAC9B,MAAM,SAAS,GAAI,KAAoC,CAAC,IAAI,CAAC;IAC7D,OAAO,OAAO,SAAS,KAAK,QAAQ,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,SAAS,CAAC;AACjE,CAAC"}
|
|
@@ -0,0 +1,34 @@
|
|
|
1
|
+
export type RuntimeEventType = 'agent:start' | 'agent:end' | 'agent:error' | 'llm:start' | 'llm:end' | 'llm:error' | 'tool:start' | 'tool:end' | 'tool:error';
|
|
2
|
+
export interface ObservedError {
|
|
3
|
+
name: string;
|
|
4
|
+
message: string;
|
|
5
|
+
code?: string | undefined;
|
|
6
|
+
stack?: string | undefined;
|
|
7
|
+
}
|
|
8
|
+
export interface RuntimeEvent {
|
|
9
|
+
type: RuntimeEventType;
|
|
10
|
+
runId: string;
|
|
11
|
+
timestamp: Date;
|
|
12
|
+
durationMs?: number | undefined;
|
|
13
|
+
agentName?: string | undefined;
|
|
14
|
+
sessionId?: string | undefined;
|
|
15
|
+
provider?: string | undefined;
|
|
16
|
+
model?: string | undefined;
|
|
17
|
+
toolName?: string | undefined;
|
|
18
|
+
toolCallId?: string | undefined;
|
|
19
|
+
messageCount?: number | undefined;
|
|
20
|
+
toolCount?: number | undefined;
|
|
21
|
+
outputLength?: number | undefined;
|
|
22
|
+
error?: ObservedError | undefined;
|
|
23
|
+
metadata?: Record<string, unknown> | undefined;
|
|
24
|
+
}
|
|
25
|
+
export interface Observer {
|
|
26
|
+
onEvent(event: RuntimeEvent): void | Promise<void>;
|
|
27
|
+
}
|
|
28
|
+
export interface TraceContext {
|
|
29
|
+
runId: string;
|
|
30
|
+
observer?: Observer | undefined;
|
|
31
|
+
agentName?: string | undefined;
|
|
32
|
+
sessionId?: string | undefined;
|
|
33
|
+
}
|
|
34
|
+
//# sourceMappingURL=types.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../../src/observability/types.ts"],"names":[],"mappings":"AAAA,MAAM,MAAM,gBAAgB,GACtB,aAAa,GACb,WAAW,GACX,aAAa,GACb,WAAW,GACX,SAAS,GACT,WAAW,GACX,YAAY,GACZ,UAAU,GACV,YAAY,CAAC;AAEnB,MAAM,WAAW,aAAa;IAC1B,IAAI,EAAE,MAAM,CAAC;IACb,OAAO,EAAE,MAAM,CAAC;IAChB,IAAI,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC1B,KAAK,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;CAC9B;AAED,MAAM,WAAW,YAAY;IACzB,IAAI,EAAE,gBAAgB,CAAC;IACvB,KAAK,EAAE,MAAM,CAAC;IACd,SAAS,EAAE,IAAI,CAAC;IAChB,UAAU,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAChC,SAAS,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC/B,SAAS,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC/B,QAAQ,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC9B,KAAK,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC3B,QAAQ,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC9B,UAAU,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAChC,YAAY,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAClC,SAAS,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC/B,YAAY,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAClC,KAAK,CAAC,EAAE,aAAa,GAAG,SAAS,CAAC;IAClC,QAAQ,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAAG,SAAS,CAAC;CAClD;AAED,MAAM,WAAW,QAAQ;IACrB,OAAO,CAAC,KAAK,EAAE,YAAY,GAAG,IAAI,GAAG,OAAO,CAAC,IAAI,CAAC,CAAC;CACtD;AAED,MAAM,WAAW,YAAY;IACzB,KAAK,EAAE,MAAM,CAAC;IACd,QAAQ,CAAC,EAAE,QAAQ,GAAG,SAAS,CAAC;IAChC,SAAS,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC/B,SAAS,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;CAClC"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.js","sourceRoot":"","sources":["../../src/observability/types.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,25 @@
|
|
|
1
|
+
import type { Chunk, RawDocument } from '../types';
|
|
2
|
+
import type { Chunker } from './chunker';
|
|
3
|
+
export interface AutoChunkerOptions {
|
|
4
|
+
markdownChunker?: Chunker;
|
|
5
|
+
parentChildChunker?: Chunker;
|
|
6
|
+
recursiveChunker?: Chunker;
|
|
7
|
+
parentChildMinContentLength?: number;
|
|
8
|
+
}
|
|
9
|
+
/**
|
|
10
|
+
* AutoChunker 根据文档形态选择切块策略:
|
|
11
|
+
* - 有 Markdown 结构或明确来自 Markdown 转换链路:使用 MarkdownChunker,保留标题、表格、代码等结构。
|
|
12
|
+
* - 无结构但内容较长:使用 ParentChildChunker,用小块召回、大块补上下文。
|
|
13
|
+
* - 其余短文本:使用 RecursiveChunker,按通用文本边界切块。
|
|
14
|
+
*/
|
|
15
|
+
export declare class AutoChunker implements Chunker {
|
|
16
|
+
private readonly markdownChunker;
|
|
17
|
+
private readonly parentChildChunker;
|
|
18
|
+
private readonly recursiveChunker;
|
|
19
|
+
private readonly parentChildMinContentLength;
|
|
20
|
+
constructor(options?: AutoChunkerOptions);
|
|
21
|
+
chunk(document: RawDocument): Promise<Chunk[]>;
|
|
22
|
+
chunkMany(documents: RawDocument[]): Promise<Chunk[]>;
|
|
23
|
+
private resolveChunker;
|
|
24
|
+
}
|
|
25
|
+
//# sourceMappingURL=auto-chunker.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"auto-chunker.d.ts","sourceRoot":"","sources":["../../../src/rag/chunking/auto-chunker.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACR,KAAK,EAEL,WAAW,EACd,MAAM,UAAU,CAAC;AAIlB,OAAO,KAAK,EACR,OAAO,EACV,MAAM,WAAW,CAAC;AAWnB,MAAM,WAAW,kBAAkB;IAC/B,eAAe,CAAC,EAAE,OAAO,CAAC;IAC1B,kBAAkB,CAAC,EAAE,OAAO,CAAC;IAC7B,gBAAgB,CAAC,EAAE,OAAO,CAAC;IAC3B,2BAA2B,CAAC,EAAE,MAAM,CAAC;CACxC;AAcD;;;;;GAKG;AACH,qBAAa,WAAY,YAAW,OAAO;IACvC,OAAO,CAAC,QAAQ,CAAC,eAAe,CAAU;IAC1C,OAAO,CAAC,QAAQ,CAAC,kBAAkB,CAAU;IAC7C,OAAO,CAAC,QAAQ,CAAC,gBAAgB,CAAU;IAC3C,OAAO,CAAC,QAAQ,CAAC,2BAA2B,CAAS;gBAEzC,OAAO,GAAE,kBAAuB;IAUtC,KAAK,CAAC,QAAQ,EAAE,WAAW,GAAG,OAAO,CAAC,KAAK,EAAE,CAAC;IAI9C,SAAS,CAAC,SAAS,EAAE,WAAW,EAAE,GAAG,OAAO,CAAC,KAAK,EAAE,CAAC;IAI3D,OAAO,CAAC,cAAc;CAWzB"}
|
|
@@ -0,0 +1,101 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.AutoChunker = void 0;
|
|
4
|
+
const chunker_1 = require("./chunker");
|
|
5
|
+
const markdown_chunker_1 = require("./markdown-chunker");
|
|
6
|
+
const parent_child_1 = require("./parent-child");
|
|
7
|
+
const recursive_chunker_1 = require("./recursive-chunker");
|
|
8
|
+
const DEFAULT_PARENT_CHILD_MIN_CONTENT_LENGTH = 6000;
|
|
9
|
+
const MARKDOWN_MIME_TYPES = new Set([
|
|
10
|
+
'text/markdown',
|
|
11
|
+
'text/x-markdown',
|
|
12
|
+
]);
|
|
13
|
+
const MARKDOWN_EXTENSIONS = new Set([
|
|
14
|
+
'.md',
|
|
15
|
+
'.markdown',
|
|
16
|
+
'.mdown',
|
|
17
|
+
'.mkd',
|
|
18
|
+
]);
|
|
19
|
+
/**
|
|
20
|
+
* AutoChunker 根据文档形态选择切块策略:
|
|
21
|
+
* - 有 Markdown 结构或明确来自 Markdown 转换链路:使用 MarkdownChunker,保留标题、表格、代码等结构。
|
|
22
|
+
* - 无结构但内容较长:使用 ParentChildChunker,用小块召回、大块补上下文。
|
|
23
|
+
* - 其余短文本:使用 RecursiveChunker,按通用文本边界切块。
|
|
24
|
+
*/
|
|
25
|
+
class AutoChunker {
|
|
26
|
+
markdownChunker;
|
|
27
|
+
parentChildChunker;
|
|
28
|
+
recursiveChunker;
|
|
29
|
+
parentChildMinContentLength;
|
|
30
|
+
constructor(options = {}) {
|
|
31
|
+
this.markdownChunker = options.markdownChunker ?? new markdown_chunker_1.MarkdownChunker();
|
|
32
|
+
this.parentChildChunker = options.parentChildChunker ?? new parent_child_1.ParentChildChunker();
|
|
33
|
+
this.recursiveChunker = options.recursiveChunker ?? new recursive_chunker_1.RecursiveChunker();
|
|
34
|
+
this.parentChildMinContentLength = toPositiveInteger(options.parentChildMinContentLength, DEFAULT_PARENT_CHILD_MIN_CONTENT_LENGTH);
|
|
35
|
+
}
|
|
36
|
+
async chunk(document) {
|
|
37
|
+
return await this.resolveChunker(document).chunk(document);
|
|
38
|
+
}
|
|
39
|
+
async chunkMany(documents) {
|
|
40
|
+
return await (0, chunker_1.chunkManyDocuments)(this, documents);
|
|
41
|
+
}
|
|
42
|
+
resolveChunker(document) {
|
|
43
|
+
if (shouldUseMarkdownChunker(document)) {
|
|
44
|
+
return this.markdownChunker;
|
|
45
|
+
}
|
|
46
|
+
if (document.content.length >= this.parentChildMinContentLength) {
|
|
47
|
+
return this.parentChildChunker;
|
|
48
|
+
}
|
|
49
|
+
return this.recursiveChunker;
|
|
50
|
+
}
|
|
51
|
+
}
|
|
52
|
+
exports.AutoChunker = AutoChunker;
|
|
53
|
+
function shouldUseMarkdownChunker(document) {
|
|
54
|
+
if (hasMarkdownStructure(document)) {
|
|
55
|
+
return true;
|
|
56
|
+
}
|
|
57
|
+
return isMarkdownLikeMetadata(document.metadata);
|
|
58
|
+
}
|
|
59
|
+
function hasMarkdownStructure(document) {
|
|
60
|
+
return document.blocks?.some((block) => block.type !== 'paragraph') ?? false;
|
|
61
|
+
}
|
|
62
|
+
function isMarkdownLikeMetadata(metadata) {
|
|
63
|
+
const mimeType = normalizeMimeType(metadata.mimeType);
|
|
64
|
+
if (mimeType !== undefined && MARKDOWN_MIME_TYPES.has(mimeType)) {
|
|
65
|
+
return true;
|
|
66
|
+
}
|
|
67
|
+
const convertedToMimeType = normalizeMimeType(readString(metadata.extra, 'convertedToMimeType'));
|
|
68
|
+
if (convertedToMimeType !== undefined && MARKDOWN_MIME_TYPES.has(convertedToMimeType)) {
|
|
69
|
+
return true;
|
|
70
|
+
}
|
|
71
|
+
const extension = getSourceExtension(metadata.source);
|
|
72
|
+
return extension !== undefined && MARKDOWN_EXTENSIONS.has(extension);
|
|
73
|
+
}
|
|
74
|
+
function readString(record, key) {
|
|
75
|
+
const value = record?.[key];
|
|
76
|
+
return typeof value === 'string' ? value : undefined;
|
|
77
|
+
}
|
|
78
|
+
function getSourceExtension(source) {
|
|
79
|
+
const withoutQuery = source.split(/[?#]/, 1)[0] ?? source;
|
|
80
|
+
const normalizedPath = withoutQuery.replace(/\\/g, '/');
|
|
81
|
+
const filename = normalizedPath.split('/').filter(Boolean).at(-1);
|
|
82
|
+
if (filename === undefined) {
|
|
83
|
+
return undefined;
|
|
84
|
+
}
|
|
85
|
+
const dotIndex = filename.lastIndexOf('.');
|
|
86
|
+
if (dotIndex <= 0 || dotIndex === filename.length - 1) {
|
|
87
|
+
return undefined;
|
|
88
|
+
}
|
|
89
|
+
return filename.slice(dotIndex).toLowerCase();
|
|
90
|
+
}
|
|
91
|
+
function normalizeMimeType(mimeType) {
|
|
92
|
+
const normalized = (mimeType?.split(';', 1)[0] ?? '').trim().toLowerCase();
|
|
93
|
+
return normalized.length > 0 ? normalized : undefined;
|
|
94
|
+
}
|
|
95
|
+
function toPositiveInteger(value, fallback) {
|
|
96
|
+
if (value === undefined || !Number.isFinite(value) || value <= 0) {
|
|
97
|
+
return fallback;
|
|
98
|
+
}
|
|
99
|
+
return Math.floor(value);
|
|
100
|
+
}
|
|
101
|
+
//# sourceMappingURL=auto-chunker.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"auto-chunker.js","sourceRoot":"","sources":["../../../src/rag/chunking/auto-chunker.ts"],"names":[],"mappings":";;;AAKA,uCAEmB;AAInB,yDAE4B;AAC5B,iDAEwB;AACxB,2DAE6B;AAS7B,MAAM,uCAAuC,GAAG,IAAI,CAAC;AACrD,MAAM,mBAAmB,GAAG,IAAI,GAAG,CAAC;IAChC,eAAe;IACf,iBAAiB;CACpB,CAAC,CAAC;AACH,MAAM,mBAAmB,GAAG,IAAI,GAAG,CAAC;IAChC,KAAK;IACL,WAAW;IACX,QAAQ;IACR,MAAM;CACT,CAAC,CAAC;AAEH;;;;;GAKG;AACH,MAAa,WAAW;IACH,eAAe,CAAU;IACzB,kBAAkB,CAAU;IAC5B,gBAAgB,CAAU;IAC1B,2BAA2B,CAAS;IAErD,YAAY,UAA8B,EAAE;QACxC,IAAI,CAAC,eAAe,GAAG,OAAO,CAAC,eAAe,IAAI,IAAI,kCAAe,EAAE,CAAC;QACxE,IAAI,CAAC,kBAAkB,GAAG,OAAO,CAAC,kBAAkB,IAAI,IAAI,iCAAkB,EAAE,CAAC;QACjF,IAAI,CAAC,gBAAgB,GAAG,OAAO,CAAC,gBAAgB,IAAI,IAAI,oCAAgB,EAAE,CAAC;QAC3E,IAAI,CAAC,2BAA2B,GAAG,iBAAiB,CAChD,OAAO,CAAC,2BAA2B,EACnC,uCAAuC,CAC1C,CAAC;IACN,CAAC;IAED,KAAK,CAAC,KAAK,CAAC,QAAqB;QAC7B,OAAO,MAAM,IAAI,CAAC,cAAc,CAAC,QAAQ,CAAC,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC;IAC/D,CAAC;IAED,KAAK,CAAC,SAAS,CAAC,SAAwB;QACpC,OAAO,MAAM,IAAA,4BAAkB,EAAC,IAAI,EAAE,SAAS,CAAC,CAAC;IACrD,CAAC;IAEO,cAAc,CAAC,QAAqB;QACxC,IAAI,wBAAwB,CAAC,QAAQ,CAAC,EAAE,CAAC;YACrC,OAAO,IAAI,CAAC,eAAe,CAAC;QAChC,CAAC;QAED,IAAI,QAAQ,CAAC,OAAO,CAAC,MAAM,IAAI,IAAI,CAAC,2BAA2B,EAAE,CAAC;YAC9D,OAAO,IAAI,CAAC,kBAAkB,CAAC;QACnC,CAAC;QAED,OAAO,IAAI,CAAC,gBAAgB,CAAC;IACjC,CAAC;CACJ;AAnCD,kCAmCC;AAED,SAAS,wBAAwB,CAAC,QAAqB;IACnD,IAAI,oBAAoB,CAAC,QAAQ,CAAC,EAAE,CAAC;QACjC,OAAO,IAAI,CAAC;IAChB,CAAC;IAED,OAAO,sBAAsB,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;AACrD,CAAC;AAED,SAAS,oBAAoB,CAAC,QAAqB;IAC/C,OAAO,QAAQ,CAAC,MAAM,EAAE,IAAI,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,IAAI,KAAK,WAAW,CAAC,IAAI,KAAK,CAAC;AACjF,CAAC;AAED,SAAS,sBAAsB,CAAC,QAA0B;IACtD,MAAM,QAAQ,GAAG,iBAAiB,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACtD,IAAI,QAAQ,KAAK,SAAS,IAAI,mBAAmB,CAAC,GAAG,CAAC,QAAQ,CAAC,EAAE,CAAC;QAC9D,OAAO,IAAI,CAAC;IAChB,CAAC;IAED,MAAM,mBAAmB,GAAG,iBAAiB,CAAC,UAAU,CAAC,QAAQ,CAAC,KAAK,EAAE,qBAAqB,CAAC,CAAC,CAAC;IACjG,IAAI,mBAAmB,KAAK,SAAS,IAAI,mBAAmB,CAAC,GAAG,CAAC,mBAAmB,CAAC,EAAE,CAAC;QACpF,OAAO,IAAI,CAAC;IAChB,CAAC;IAED,MAAM,SAAS,GAAG,kBAAkB,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;IAEtD,OAAO,SAAS,KAAK,SAAS,IAAI,mBAAmB,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC;AACzE,CAAC;AAED,SAAS,UAAU,CAAC,MAA2C,EAAE,GAAW;IACxE,MAAM,KAAK,GAAG,MAAM,EAAE,CAAC,GAAG,CAAC,CAAC;IAE5B,OAAO,OAAO,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,SAAS,CAAC;AACzD,CAAC;AAED,SAAS,kBAAkB,CAAC,MAAc;IACtC,MAAM,YAAY,GAAG,MAAM,CAAC,KAAK,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,MAAM,CAAC;IAC1D,MAAM,cAAc,GAAG,YAAY,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;IACxD,MAAM,QAAQ,GAAG,cAAc,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;IAClE,IAAI,QAAQ,KAAK,SAAS,EAAE,CAAC;QACzB,OAAO,SAAS,CAAC;IACrB,CAAC;IAED,MAAM,QAAQ,GAAG,QAAQ,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;IAC3C,IAAI,QAAQ,IAAI,CAAC,IAAI,QAAQ,KAAK,QAAQ,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACpD,OAAO,SAAS,CAAC;IACrB,CAAC;IAED,OAAO,QAAQ,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;AAClD,CAAC;AAED,SAAS,iBAAiB,CAAC,QAA4B;IACnD,MAAM,UAAU,GAAG,CAAC,QAAQ,EAAE,KAAK,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;IAE3E,OAAO,UAAU,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,CAAC,SAAS,CAAC;AAC1D,CAAC;AAED,SAAS,iBAAiB,CAAC,KAAyB,EAAE,QAAgB;IAClE,IAAI,KAAK,KAAK,SAAS,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,KAAK,IAAI,CAAC,EAAE,CAAC;QAC/D,OAAO,QAAQ,CAAC;IACpB,CAAC;IAED,OAAO,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;AAC7B,CAAC"}
|
|
@@ -0,0 +1,44 @@
|
|
|
1
|
+
import type { Chunk, ChunkKind, RawDocument } from '../types';
|
|
2
|
+
export interface Chunker {
|
|
3
|
+
chunk(document: RawDocument): Promise<Chunk[]>;
|
|
4
|
+
chunkMany(documents: RawDocument[]): Promise<Chunk[]>;
|
|
5
|
+
}
|
|
6
|
+
export interface TextSplitOptions {
|
|
7
|
+
/**
|
|
8
|
+
* 初始实现用字符长度近似 token 预算,避免引入 tokenizer 依赖。
|
|
9
|
+
* 后续接入具体模型 tokenizer 时,可以在 chunker 内部替换长度计算。
|
|
10
|
+
*/
|
|
11
|
+
maxChunkLength?: number;
|
|
12
|
+
minChunkLength?: number;
|
|
13
|
+
overlapLength?: number;
|
|
14
|
+
}
|
|
15
|
+
export interface TextSegment {
|
|
16
|
+
content: string;
|
|
17
|
+
startOffset: number;
|
|
18
|
+
endOffset: number;
|
|
19
|
+
}
|
|
20
|
+
export interface CreateChunkInput {
|
|
21
|
+
document: RawDocument;
|
|
22
|
+
content: string;
|
|
23
|
+
kind: ChunkKind;
|
|
24
|
+
chunkIndex: number;
|
|
25
|
+
parentId?: string;
|
|
26
|
+
headingPath?: string[];
|
|
27
|
+
startOffset?: number;
|
|
28
|
+
endOffset?: number;
|
|
29
|
+
extra?: Record<string, unknown>;
|
|
30
|
+
}
|
|
31
|
+
export declare const DEFAULT_MAX_CHUNK_LENGTH = 1200;
|
|
32
|
+
export declare const DEFAULT_MIN_CHUNK_LENGTH = 240;
|
|
33
|
+
export declare const DEFAULT_OVERLAP_LENGTH = 120;
|
|
34
|
+
export declare function chunkManyDocuments(chunker: Chunker, documents: RawDocument[]): Promise<Chunk[]>;
|
|
35
|
+
export declare function splitTextSegments(text: string, options?: TextSplitOptions): TextSegment[];
|
|
36
|
+
export declare function createChunk(input: CreateChunkInput): Chunk;
|
|
37
|
+
export interface ChunkIdInput {
|
|
38
|
+
documentId: string;
|
|
39
|
+
chunkIndex: number;
|
|
40
|
+
content: string;
|
|
41
|
+
parentId?: string;
|
|
42
|
+
}
|
|
43
|
+
export declare function createChunkId(input: ChunkIdInput): string;
|
|
44
|
+
//# sourceMappingURL=chunker.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"chunker.d.ts","sourceRoot":"","sources":["../../../src/rag/chunking/chunker.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EACR,KAAK,EACL,SAAS,EAET,WAAW,EACd,MAAM,UAAU,CAAC;AAElB,MAAM,WAAW,OAAO;IACpB,KAAK,CAAC,QAAQ,EAAE,WAAW,GAAG,OAAO,CAAC,KAAK,EAAE,CAAC,CAAC;IAC/C,SAAS,CAAC,SAAS,EAAE,WAAW,EAAE,GAAG,OAAO,CAAC,KAAK,EAAE,CAAC,CAAC;CACzD;AAED,MAAM,WAAW,gBAAgB;IAC7B;;;OAGG;IACH,cAAc,CAAC,EAAE,MAAM,CAAC;IACxB,cAAc,CAAC,EAAE,MAAM,CAAC;IACxB,aAAa,CAAC,EAAE,MAAM,CAAC;CAC1B;AAED,MAAM,WAAW,WAAW;IACxB,OAAO,EAAE,MAAM,CAAC;IAChB,WAAW,EAAE,MAAM,CAAC;IACpB,SAAS,EAAE,MAAM,CAAC;CACrB;AAED,MAAM,WAAW,gBAAgB;IAC7B,QAAQ,EAAE,WAAW,CAAC;IACtB,OAAO,EAAE,MAAM,CAAC;IAChB,IAAI,EAAE,SAAS,CAAC;IAChB,UAAU,EAAE,MAAM,CAAC;IACnB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,WAAW,CAAC,EAAE,MAAM,EAAE,CAAC;IACvB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,KAAK,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CACnC;AAED,eAAO,MAAM,wBAAwB,OAAO,CAAC;AAC7C,eAAO,MAAM,wBAAwB,MAAM,CAAC;AAC5C,eAAO,MAAM,sBAAsB,MAAM,CAAC;AAe1C,wBAAsB,kBAAkB,CACpC,OAAO,EAAE,OAAO,EAChB,SAAS,EAAE,WAAW,EAAE,GACzB,OAAO,CAAC,KAAK,EAAE,CAAC,CAIlB;AAED,wBAAgB,iBAAiB,CAC7B,IAAI,EAAE,MAAM,EACZ,OAAO,GAAE,gBAAqB,GAC/B,WAAW,EAAE,CA4Cf;AAED,wBAAgB,WAAW,CAAC,KAAK,EAAE,gBAAgB,GAAG,KAAK,CA2C1D;AAED,MAAM,WAAW,YAAY;IACzB,UAAU,EAAE,MAAM,CAAC;IACnB,UAAU,EAAE,MAAM,CAAC;IACnB,OAAO,EAAE,MAAM,CAAC;IAChB,QAAQ,CAAC,EAAE,MAAM,CAAC;CACrB;AAED,wBAAgB,aAAa,CAAC,KAAK,EAAE,YAAY,GAAG,MAAM,CASzD"}
|
|
@@ -0,0 +1,162 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.DEFAULT_OVERLAP_LENGTH = exports.DEFAULT_MIN_CHUNK_LENGTH = exports.DEFAULT_MAX_CHUNK_LENGTH = void 0;
|
|
4
|
+
exports.chunkManyDocuments = chunkManyDocuments;
|
|
5
|
+
exports.splitTextSegments = splitTextSegments;
|
|
6
|
+
exports.createChunk = createChunk;
|
|
7
|
+
exports.createChunkId = createChunkId;
|
|
8
|
+
const node_crypto_1 = require("node:crypto");
|
|
9
|
+
exports.DEFAULT_MAX_CHUNK_LENGTH = 1200;
|
|
10
|
+
exports.DEFAULT_MIN_CHUNK_LENGTH = 240;
|
|
11
|
+
exports.DEFAULT_OVERLAP_LENGTH = 120;
|
|
12
|
+
const DEFAULT_SEPARATORS = [
|
|
13
|
+
'\n\n',
|
|
14
|
+
'\n',
|
|
15
|
+
'。 ',
|
|
16
|
+
'。', // 中文文档常见句末符号,不能只依赖空格。
|
|
17
|
+
'. ',
|
|
18
|
+
'! ',
|
|
19
|
+
'? ',
|
|
20
|
+
'; ',
|
|
21
|
+
', ',
|
|
22
|
+
' ',
|
|
23
|
+
];
|
|
24
|
+
async function chunkManyDocuments(chunker, documents) {
|
|
25
|
+
const chunks = await Promise.all(documents.map((document) => chunker.chunk(document)));
|
|
26
|
+
return chunks.flat();
|
|
27
|
+
}
|
|
28
|
+
function splitTextSegments(text, options = {}) {
|
|
29
|
+
const maxChunkLength = toPositiveInteger(options.maxChunkLength, exports.DEFAULT_MAX_CHUNK_LENGTH);
|
|
30
|
+
const minChunkLength = Math.min(toPositiveInteger(options.minChunkLength, exports.DEFAULT_MIN_CHUNK_LENGTH), maxChunkLength);
|
|
31
|
+
const overlapLength = Math.min(Math.max(toPositiveInteger(options.overlapLength, exports.DEFAULT_OVERLAP_LENGTH), 0), Math.max(maxChunkLength - 1, 0));
|
|
32
|
+
const segments = [];
|
|
33
|
+
let start = skipWhitespace(text, 0);
|
|
34
|
+
while (start < text.length) {
|
|
35
|
+
const hardEnd = Math.min(start + maxChunkLength, text.length);
|
|
36
|
+
const end = hardEnd < text.length
|
|
37
|
+
? findBestBreak(text, start, hardEnd, minChunkLength)
|
|
38
|
+
: hardEnd;
|
|
39
|
+
const trimmed = trimRange(text, start, end);
|
|
40
|
+
if (trimmed.start < trimmed.end) {
|
|
41
|
+
segments.push({
|
|
42
|
+
content: text.slice(trimmed.start, trimmed.end),
|
|
43
|
+
startOffset: trimmed.start,
|
|
44
|
+
endOffset: trimmed.end,
|
|
45
|
+
});
|
|
46
|
+
}
|
|
47
|
+
if (end >= text.length) {
|
|
48
|
+
break;
|
|
49
|
+
}
|
|
50
|
+
const preferredStart = Math.max(end - overlapLength, start + 1);
|
|
51
|
+
const nextStart = overlapLength > 0
|
|
52
|
+
? findForwardBoundary(text, preferredStart, end)
|
|
53
|
+
: end;
|
|
54
|
+
start = skipWhitespace(text, Math.max(nextStart, start + 1));
|
|
55
|
+
}
|
|
56
|
+
return segments;
|
|
57
|
+
}
|
|
58
|
+
function createChunk(input) {
|
|
59
|
+
const content = input.content.trim();
|
|
60
|
+
const metadata = {
|
|
61
|
+
source: input.document.metadata.source,
|
|
62
|
+
chunkIndex: input.chunkIndex,
|
|
63
|
+
};
|
|
64
|
+
assignIfDefined(metadata, 'title', input.document.metadata.title);
|
|
65
|
+
assignIfDefined(metadata, 'tenantId', input.document.metadata.tenantId);
|
|
66
|
+
assignIfDefined(metadata, 'knowledgeBaseId', input.document.metadata.knowledgeBaseId);
|
|
67
|
+
assignIfDefined(metadata, 'startOffset', input.startOffset);
|
|
68
|
+
assignIfDefined(metadata, 'endOffset', input.endOffset);
|
|
69
|
+
if (input.document.metadata.acl !== undefined) {
|
|
70
|
+
metadata.acl = [...input.document.metadata.acl];
|
|
71
|
+
}
|
|
72
|
+
if (input.headingPath !== undefined && input.headingPath.length > 0) {
|
|
73
|
+
metadata.headingPath = [...input.headingPath];
|
|
74
|
+
}
|
|
75
|
+
const extra = mergeRecords(input.document.metadata.extra, input.extra);
|
|
76
|
+
if (extra !== undefined) {
|
|
77
|
+
metadata.extra = extra;
|
|
78
|
+
}
|
|
79
|
+
const chunkIdInput = {
|
|
80
|
+
documentId: input.document.id,
|
|
81
|
+
chunkIndex: input.chunkIndex,
|
|
82
|
+
content,
|
|
83
|
+
};
|
|
84
|
+
assignIfDefined(chunkIdInput, 'parentId', input.parentId);
|
|
85
|
+
const chunk = {
|
|
86
|
+
id: createChunkId(chunkIdInput),
|
|
87
|
+
documentId: input.document.id,
|
|
88
|
+
content,
|
|
89
|
+
kind: input.kind,
|
|
90
|
+
metadata,
|
|
91
|
+
};
|
|
92
|
+
assignIfDefined(chunk, 'parentId', input.parentId);
|
|
93
|
+
return chunk;
|
|
94
|
+
}
|
|
95
|
+
function createChunkId(input) {
|
|
96
|
+
const stableKey = [
|
|
97
|
+
input.documentId,
|
|
98
|
+
String(input.chunkIndex),
|
|
99
|
+
input.parentId ?? '',
|
|
100
|
+
input.content,
|
|
101
|
+
].join('\0');
|
|
102
|
+
return `chunk_${(0, node_crypto_1.createHash)('sha256').update(stableKey).digest('hex').slice(0, 16)}`;
|
|
103
|
+
}
|
|
104
|
+
function findBestBreak(text, start, hardEnd, minChunkLength) {
|
|
105
|
+
const minEnd = Math.min(start + minChunkLength, hardEnd);
|
|
106
|
+
for (const separator of DEFAULT_SEPARATORS) {
|
|
107
|
+
const index = text.lastIndexOf(separator, hardEnd);
|
|
108
|
+
if (index >= minEnd) {
|
|
109
|
+
return index + separator.length;
|
|
110
|
+
}
|
|
111
|
+
}
|
|
112
|
+
return hardEnd;
|
|
113
|
+
}
|
|
114
|
+
function skipWhitespace(text, start) {
|
|
115
|
+
let index = start;
|
|
116
|
+
while (index < text.length && /\s/.test(text[index] ?? '')) {
|
|
117
|
+
index += 1;
|
|
118
|
+
}
|
|
119
|
+
return index;
|
|
120
|
+
}
|
|
121
|
+
function trimRange(text, start, end) {
|
|
122
|
+
let trimmedStart = start;
|
|
123
|
+
let trimmedEnd = end;
|
|
124
|
+
while (trimmedStart < trimmedEnd && /\s/.test(text[trimmedStart] ?? '')) {
|
|
125
|
+
trimmedStart += 1;
|
|
126
|
+
}
|
|
127
|
+
while (trimmedEnd > trimmedStart && /\s/.test(text[trimmedEnd - 1] ?? '')) {
|
|
128
|
+
trimmedEnd -= 1;
|
|
129
|
+
}
|
|
130
|
+
return {
|
|
131
|
+
start: trimmedStart,
|
|
132
|
+
end: trimmedEnd,
|
|
133
|
+
};
|
|
134
|
+
}
|
|
135
|
+
function findForwardBoundary(text, preferredStart, end) {
|
|
136
|
+
let index = preferredStart;
|
|
137
|
+
while (index < end && !/\s/.test(text[index - 1] ?? '')) {
|
|
138
|
+
index += 1;
|
|
139
|
+
}
|
|
140
|
+
return Math.min(index, end);
|
|
141
|
+
}
|
|
142
|
+
function toPositiveInteger(value, fallback) {
|
|
143
|
+
if (value === undefined || !Number.isFinite(value) || value <= 0) {
|
|
144
|
+
return fallback;
|
|
145
|
+
}
|
|
146
|
+
return Math.floor(value);
|
|
147
|
+
}
|
|
148
|
+
function mergeRecords(left, right) {
|
|
149
|
+
if (left === undefined && right === undefined) {
|
|
150
|
+
return undefined;
|
|
151
|
+
}
|
|
152
|
+
return {
|
|
153
|
+
...(left ?? {}),
|
|
154
|
+
...(right ?? {}),
|
|
155
|
+
};
|
|
156
|
+
}
|
|
157
|
+
function assignIfDefined(target, key, value) {
|
|
158
|
+
if (value !== undefined) {
|
|
159
|
+
target[key] = value;
|
|
160
|
+
}
|
|
161
|
+
}
|
|
162
|
+
//# sourceMappingURL=chunker.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"chunker.js","sourceRoot":"","sources":["../../../src/rag/chunking/chunker.ts"],"names":[],"mappings":";;;AA0DA,gDAOC;AAED,8CA+CC;AAED,kCA2CC;AASD,sCASC;AAjLD,6CAAyC;AAyC5B,QAAA,wBAAwB,GAAG,IAAI,CAAC;AAChC,QAAA,wBAAwB,GAAG,GAAG,CAAC;AAC/B,QAAA,sBAAsB,GAAG,GAAG,CAAC;AAE1C,MAAM,kBAAkB,GAAG;IACvB,MAAM;IACN,IAAI;IACJ,IAAI;IACJ,GAAG,EAAE,sBAAsB;IAC3B,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,IAAI;IACJ,GAAG;CACG,CAAC;AAEJ,KAAK,UAAU,kBAAkB,CACpC,OAAgB,EAChB,SAAwB;IAExB,MAAM,MAAM,GAAG,MAAM,OAAO,CAAC,GAAG,CAAC,SAAS,CAAC,GAAG,CAAC,CAAC,QAAQ,EAAE,EAAE,CAAC,OAAO,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC;IAEvF,OAAO,MAAM,CAAC,IAAI,EAAE,CAAC;AACzB,CAAC;AAED,SAAgB,iBAAiB,CAC7B,IAAY,EACZ,UAA4B,EAAE;IAE9B,MAAM,cAAc,GAAG,iBAAiB,CACpC,OAAO,CAAC,cAAc,EACtB,gCAAwB,CAC3B,CAAC;IACF,MAAM,cAAc,GAAG,IAAI,CAAC,GAAG,CAC3B,iBAAiB,CAAC,OAAO,CAAC,cAAc,EAAE,gCAAwB,CAAC,EACnE,cAAc,CACjB,CAAC;IACF,MAAM,aAAa,GAAG,IAAI,CAAC,GAAG,CAC1B,IAAI,CAAC,GAAG,CAAC,iBAAiB,CAAC,OAAO,CAAC,aAAa,EAAE,8BAAsB,CAAC,EAAE,CAAC,CAAC,EAC7E,IAAI,CAAC,GAAG,CAAC,cAAc,GAAG,CAAC,EAAE,CAAC,CAAC,CAClC,CAAC;IAEF,MAAM,QAAQ,GAAkB,EAAE,CAAC;IACnC,IAAI,KAAK,GAAG,cAAc,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC;IAEpC,OAAO,KAAK,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC;QACzB,MAAM,OAAO,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,GAAG,cAAc,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC;QAC9D,MAAM,GAAG,GAAG,OAAO,GAAG,IAAI,CAAC,MAAM;YAC7B,CAAC,CAAC,aAAa,CAAC,IAAI,EAAE,KAAK,EAAE,OAAO,EAAE,cAAc,CAAC;YACrD,CAAC,CAAC,OAAO,CAAC;QACd,MAAM,OAAO,GAAG,SAAS,CAAC,IAAI,EAAE,KAAK,EAAE,GAAG,CAAC,CAAC;QAE5C,IAAI,OAAO,CAAC,KAAK,GAAG,OAAO,CAAC,GAAG,EAAE,CAAC;YAC9B,QAAQ,CAAC,IAAI,CAAC;gBACV,OAAO,EAAE,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,EAAE,OAAO,CAAC,GAAG,CAAC;gBAC/C,WAAW,EAAE,OAAO,CAAC,KAAK;gBAC1B,SAAS,EAAE,OAAO,CAAC,GAAG;aACzB,CAAC,CAAC;QACP,CAAC;QAED,IAAI,GAAG,IAAI,IAAI,CAAC,MAAM,EAAE,CAAC;YACrB,MAAM;QACV,CAAC;QAED,MAAM,cAAc,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,GAAG,aAAa,EAAE,KAAK,GAAG,CAAC,CAAC,CAAC;QAChE,MAAM,SAAS,GAAG,aAAa,GAAG,CAAC;YAC/B,CAAC,CAAC,mBAAmB,CAAC,IAAI,EAAE,cAAc,EAAE,GAAG,CAAC;YAChD,CAAC,CAAC,GAAG,CAAC;QACV,KAAK,GAAG,cAAc,CAAC,IAAI,EAAE,IAAI,CAAC,GAAG,CAAC,SAAS,EAAE,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC;IACjE,CAAC;IAED,OAAO,QAAQ,CAAC;AACpB,CAAC;AAED,SAAgB,WAAW,CAAC,KAAuB;IAC/C,MAAM,OAAO,GAAG,KAAK,CAAC,OAAO,CAAC,IAAI,EAAE,CAAC;IACrC,MAAM,QAAQ,GAAkB;QAC5B,MAAM,EAAE,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC,MAAM;QACtC,UAAU,EAAE,KAAK,CAAC,UAAU;KAC/B,CAAC;IAEF,eAAe,CAAC,QAAQ,EAAE,OAAO,EAAE,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC;IAClE,eAAe,CAAC,QAAQ,EAAE,UAAU,EAAE,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACxE,eAAe,CAAC,QAAQ,EAAE,iBAAiB,EAAE,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC,eAAe,CAAC,CAAC;IACtF,eAAe,CAAC,QAAQ,EAAE,aAAa,EAAE,KAAK,CAAC,WAAW,CAAC,CAAC;IAC5D,eAAe,CAAC,QAAQ,EAAE,WAAW,EAAE,KAAK,CAAC,SAAS,CAAC,CAAC;IAExD,IAAI,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC,GAAG,KAAK,SAAS,EAAE,CAAC;QAC5C,QAAQ,CAAC,GAAG,GAAG,CAAC,GAAG,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC;IACpD,CAAC;IACD,IAAI,KAAK,CAAC,WAAW,KAAK,SAAS,IAAI,KAAK,CAAC,WAAW,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAClE,QAAQ,CAAC,WAAW,GAAG,CAAC,GAAG,KAAK,CAAC,WAAW,CAAC,CAAC;IAClD,CAAC;IAED,MAAM,KAAK,GAAG,YAAY,CAAC,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC,KAAK,EAAE,KAAK,CAAC,KAAK,CAAC,CAAC;IACvE,IAAI,KAAK,KAAK,SAAS,EAAE,CAAC;QACtB,QAAQ,CAAC,KAAK,GAAG,KAAK,CAAC;IAC3B,CAAC;IAED,MAAM,YAAY,GAAiB;QAC/B,UAAU,EAAE,KAAK,CAAC,QAAQ,CAAC,EAAE;QAC7B,UAAU,EAAE,KAAK,CAAC,UAAU;QAC5B,OAAO;KACV,CAAC;IACF,eAAe,CAAC,YAAY,EAAE,UAAU,EAAE,KAAK,CAAC,QAAQ,CAAC,CAAC;IAE1D,MAAM,KAAK,GAAU;QACjB,EAAE,EAAE,aAAa,CAAC,YAAY,CAAC;QAC/B,UAAU,EAAE,KAAK,CAAC,QAAQ,CAAC,EAAE;QAC7B,OAAO;QACP,IAAI,EAAE,KAAK,CAAC,IAAI;QAChB,QAAQ;KACX,CAAC;IAEF,eAAe,CAAC,KAAK,EAAE,UAAU,EAAE,KAAK,CAAC,QAAQ,CAAC,CAAC;IAEnD,OAAO,KAAK,CAAC;AACjB,CAAC;AASD,SAAgB,aAAa,CAAC,KAAmB;IAC7C,MAAM,SAAS,GAAG;QACd,KAAK,CAAC,UAAU;QAChB,MAAM,CAAC,KAAK,CAAC,UAAU,CAAC;QACxB,KAAK,CAAC,QAAQ,IAAI,EAAE;QACpB,KAAK,CAAC,OAAO;KAChB,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAEb,OAAO,SAAS,IAAA,wBAAU,EAAC,QAAQ,CAAC,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,CAAC;AACxF,CAAC;AAED,SAAS,aAAa,CAClB,IAAY,EACZ,KAAa,EACb,OAAe,EACf,cAAsB;IAEtB,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,GAAG,cAAc,EAAE,OAAO,CAAC,CAAC;IAEzD,KAAK,MAAM,SAAS,IAAI,kBAAkB,EAAE,CAAC;QACzC,MAAM,KAAK,GAAG,IAAI,CAAC,WAAW,CAAC,SAAS,EAAE,OAAO,CAAC,CAAC;QACnD,IAAI,KAAK,IAAI,MAAM,EAAE,CAAC;YAClB,OAAO,KAAK,GAAG,SAAS,CAAC,MAAM,CAAC;QACpC,CAAC;IACL,CAAC;IAED,OAAO,OAAO,CAAC;AACnB,CAAC;AAED,SAAS,cAAc,CAAC,IAAY,EAAE,KAAa;IAC/C,IAAI,KAAK,GAAG,KAAK,CAAC;IAClB,OAAO,KAAK,GAAG,IAAI,CAAC,MAAM,IAAI,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC;QACzD,KAAK,IAAI,CAAC,CAAC;IACf,CAAC;IAED,OAAO,KAAK,CAAC;AACjB,CAAC;AAED,SAAS,SAAS,CAAC,IAAY,EAAE,KAAa,EAAE,GAAW;IACvD,IAAI,YAAY,GAAG,KAAK,CAAC;IACzB,IAAI,UAAU,GAAG,GAAG,CAAC;IAErB,OAAO,YAAY,GAAG,UAAU,IAAI,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,YAAY,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC;QACtE,YAAY,IAAI,CAAC,CAAC;IACtB,CAAC;IACD,OAAO,UAAU,GAAG,YAAY,IAAI,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,UAAU,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC;QACxE,UAAU,IAAI,CAAC,CAAC;IACpB,CAAC;IAED,OAAO;QACH,KAAK,EAAE,YAAY;QACnB,GAAG,EAAE,UAAU;KAClB,CAAC;AACN,CAAC;AAED,SAAS,mBAAmB,CAAC,IAAY,EAAE,cAAsB,EAAE,GAAW;IAC1E,IAAI,KAAK,GAAG,cAAc,CAAC;IAE3B,OAAO,KAAK,GAAG,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC;QACtD,KAAK,IAAI,CAAC,CAAC;IACf,CAAC;IAED,OAAO,IAAI,CAAC,GAAG,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;AAChC,CAAC;AAED,SAAS,iBAAiB,CAAC,KAAyB,EAAE,QAAgB;IAClE,IAAI,KAAK,KAAK,SAAS,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,KAAK,IAAI,CAAC,EAAE,CAAC;QAC/D,OAAO,QAAQ,CAAC;IACpB,CAAC;IAED,OAAO,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;AAC7B,CAAC;AAED,SAAS,YAAY,CACjB,IAAyC,EACzC,KAA0C;IAE1C,IAAI,IAAI,KAAK,SAAS,IAAI,KAAK,KAAK,SAAS,EAAE,CAAC;QAC5C,OAAO,SAAS,CAAC;IACrB,CAAC;IAED,OAAO;QACH,GAAG,CAAC,IAAI,IAAI,EAAE,CAAC;QACf,GAAG,CAAC,KAAK,IAAI,EAAE,CAAC;KACnB,CAAC;AACN,CAAC;AAED,SAAS,eAAe,CACpB,MAAS,EACT,GAAM,EACN,KAAuB;IAEvB,IAAI,KAAK,KAAK,SAAS,EAAE,CAAC;QACtB,MAAM,CAAC,GAAG,CAAC,GAAG,KAAK,CAAC;IACxB,CAAC;AACL,CAAC"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/rag/chunking/index.ts"],"names":[],"mappings":"AAAA,cAAc,WAAW,CAAC;AAC1B,cAAc,gBAAgB,CAAC;AAC/B,cAAc,oBAAoB,CAAC;AACnC,cAAc,gBAAgB,CAAC;AAC/B,cAAc,qBAAqB,CAAC"}
|
|
@@ -0,0 +1,22 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
|
|
3
|
+
if (k2 === undefined) k2 = k;
|
|
4
|
+
var desc = Object.getOwnPropertyDescriptor(m, k);
|
|
5
|
+
if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
|
|
6
|
+
desc = { enumerable: true, get: function() { return m[k]; } };
|
|
7
|
+
}
|
|
8
|
+
Object.defineProperty(o, k2, desc);
|
|
9
|
+
}) : (function(o, m, k, k2) {
|
|
10
|
+
if (k2 === undefined) k2 = k;
|
|
11
|
+
o[k2] = m[k];
|
|
12
|
+
}));
|
|
13
|
+
var __exportStar = (this && this.__exportStar) || function(m, exports) {
|
|
14
|
+
for (var p in m) if (p !== "default" && !Object.prototype.hasOwnProperty.call(exports, p)) __createBinding(exports, m, p);
|
|
15
|
+
};
|
|
16
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
17
|
+
__exportStar(require("./chunker"), exports);
|
|
18
|
+
__exportStar(require("./auto-chunker"), exports);
|
|
19
|
+
__exportStar(require("./markdown-chunker"), exports);
|
|
20
|
+
__exportStar(require("./parent-child"), exports);
|
|
21
|
+
__exportStar(require("./recursive-chunker"), exports);
|
|
22
|
+
//# sourceMappingURL=index.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.js","sourceRoot":"","sources":["../../../src/rag/chunking/index.ts"],"names":[],"mappings":";;;;;;;;;;;;;;;;AAAA,4CAA0B;AAC1B,iDAA+B;AAC/B,qDAAmC;AACnC,iDAA+B;AAC/B,sDAAoC"}
|
|
@@ -0,0 +1,13 @@
|
|
|
1
|
+
import type { Chunk, RawDocument } from '../types';
|
|
2
|
+
import type { Chunker, TextSplitOptions } from './chunker';
|
|
3
|
+
export interface MarkdownChunkerOptions extends TextSplitOptions {
|
|
4
|
+
includeHeadingInContent?: boolean;
|
|
5
|
+
}
|
|
6
|
+
export declare class MarkdownChunker implements Chunker {
|
|
7
|
+
private readonly options;
|
|
8
|
+
constructor(options?: MarkdownChunkerOptions);
|
|
9
|
+
chunk(document: RawDocument): Promise<Chunk[]>;
|
|
10
|
+
chunkMany(documents: RawDocument[]): Promise<Chunk[]>;
|
|
11
|
+
private chunkPlainContent;
|
|
12
|
+
}
|
|
13
|
+
//# sourceMappingURL=markdown-chunker.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"markdown-chunker.d.ts","sourceRoot":"","sources":["../../../src/rag/chunking/markdown-chunker.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACR,KAAK,EAEL,WAAW,EACd,MAAM,UAAU,CAAC;AAMlB,OAAO,KAAK,EACR,OAAO,EACP,gBAAgB,EACnB,MAAM,WAAW,CAAC;AAEnB,MAAM,WAAW,sBAAuB,SAAQ,gBAAgB;IAC5D,uBAAuB,CAAC,EAAE,OAAO,CAAC;CACrC;AAOD,qBAAa,eAAgB,YAAW,OAAO;IAC3C,OAAO,CAAC,QAAQ,CAAC,OAAO,CAAyB;gBAErC,OAAO,GAAE,sBAA2B;IAI1C,KAAK,CAAC,QAAQ,EAAE,WAAW,GAAG,OAAO,CAAC,KAAK,EAAE,CAAC;IAyC9C,SAAS,CAAC,SAAS,EAAE,WAAW,EAAE,GAAG,OAAO,CAAC,KAAK,EAAE,CAAC;IAI3D,OAAO,CAAC,iBAAiB;CAY5B"}
|