{"version":3,"file":"vectorstores.cjs","names":["obj: Record","flattenedObject: Record","el: unknown","Document","VectorStore","embeddings: EmbeddingsInterface","args: WeaviateLibArgs","config: WeaviateLibArgs & { dimensions?: number }","configure","vectors: number[][]","documents: Document[]","options?: { ids?: string[] }","batch: DataObject[]","params: {\n ids?: string[];\n filter?: FilterValue;\n }","query: string","options?: HybridOptions","query_vector: number[] | undefined","generate: GenerateOptions","options?: BaseHybridOptions","query: number[]","k: number","filter?: FilterValue","documents: [Document, number, number, number[]][]","options: MaxMarginalRelevanceSearchOptions","_callbacks?: undefined","queryEmbedding: number[]","allResults: [Document, number, number, number[]][]","texts: string[]","metadatas: object | object[]","docs: Document[]"],"sources":["../src/vectorstores.ts"],"sourcesContent":["import * as uuid from \"uuid\";\nimport {\n HybridOptions,\n CollectionConfigCreate,\n configure,\n type DataObject,\n type FilterValue,\n GenerateOptions,\n GenerativeConfigRuntime,\n Metadata,\n Vectors,\n WeaviateClient,\n type WeaviateField,\n BaseHybridOptions,\n MetadataKeys,\n} from \"weaviate-client\";\nimport {\n type MaxMarginalRelevanceSearchOptions,\n VectorStore,\n} from \"@langchain/core/vectorstores\";\nimport type { EmbeddingsInterface } from \"@langchain/core/embeddings\";\nimport { Document } from \"@langchain/core/documents\";\nimport { maximalMarginalRelevance } from \"@langchain/core/utils/math\";\n\n// Note this function is not generic, it is designed specifically for Weaviate\n// https://weaviate.io/developers/weaviate/config-refs/datatypes#introduction\nexport const flattenObjectForWeaviate = (obj: Record) => {\n const flattenedObject: Record = {};\n for (const key in obj) {\n if (!Object.hasOwn(obj, key)) {\n continue;\n }\n\n // Replace any character that is not a letter, a number, or an underscore with '_'\n const newKey = key.replace(/[^a-zA-Z0-9_]/g, \"_\");\n\n const value = obj[key];\n if (typeof value === \"object\" && !Array.isArray(value)) {\n const recursiveResult = flattenObjectForWeaviate(\n value as Record\n );\n for (const deepKey in recursiveResult) {\n if (Object.hasOwn(recursiveResult, deepKey)) {\n // Replace any character in the deepKey that is not a letter, a number, or an underscore with '_'\n const newDeepKey = deepKey.replace(/[^a-zA-Z0-9_]/g, \"_\");\n flattenedObject[`${newKey}_${newDeepKey}`] = recursiveResult[deepKey];\n }\n }\n } else if (Array.isArray(value)) {\n if (value.length === 0) {\n flattenedObject[newKey] = value;\n } else if (\n typeof value[0] !== \"object\" &&\n value.every((el: unknown) => typeof el === typeof value[0])\n ) {\n // Weaviate only supports arrays of primitive types,\n // where all elements are of the same type\n flattenedObject[newKey] = value;\n }\n } else {\n flattenedObject[newKey] = value;\n }\n }\n return flattenedObject;\n};\n\n/**\n * Interface that defines the arguments required to create a new instance\n * of the `WeaviateStore` class. It includes the Weaviate client, the name\n * of the class in Weaviate, and optional keys for text and metadata.\n */\nexport interface WeaviateLibArgs {\n client: WeaviateClient;\n /**\n * The name of the class in Weaviate. Must start with a capital letter.\n */\n indexName?: string;\n textKey?: string;\n metadataKeys?: string[];\n tenant?: string;\n schema?: CollectionConfigCreate;\n}\n\nexport class WeaviateDocument extends Document {\n generated?: string;\n\n additional: Partial;\n\n vectors: Vectors;\n}\n\n/**\n * Class that extends the `VectorStore` base class. It provides methods to\n * interact with a Weaviate index, including adding vectors and documents,\n * deleting data, and performing similarity searches.\n */\nexport class WeaviateStore extends VectorStore {\n declare FilterType: FilterValue;\n\n private client: WeaviateClient;\n\n private indexName: string;\n\n private textKey: string;\n\n private queryAttrs: string[];\n\n private tenant?: string;\n\n private schema?: CollectionConfigCreate;\n\n _vectorstoreType(): string {\n return \"weaviate\";\n }\n\n constructor(public embeddings: EmbeddingsInterface, args: WeaviateLibArgs) {\n super(embeddings, args);\n\n this.client = args.client;\n this.indexName = args.indexName || args.schema?.name || \"\";\n this.textKey = args.textKey || \"text\";\n this.queryAttrs = [this.textKey];\n this.tenant = args.tenant;\n this.schema = args.schema;\n\n if (args.metadataKeys) {\n this.queryAttrs = [\n ...new Set([\n ...this.queryAttrs,\n ...args.metadataKeys.filter((k) => {\n // https://spec.graphql.org/June2018/#sec-Names\n // queryAttrs need to be valid GraphQL Names\n const keyIsValid = /^[_A-Za-z][_0-9A-Za-z]*$/.test(k);\n if (!keyIsValid) {\n console.warn(\n `Skipping metadata key ${k} as it is not a valid GraphQL Name`\n );\n }\n return keyIsValid;\n }),\n ]),\n ];\n }\n }\n\n static async initialize(\n embeddings: EmbeddingsInterface,\n config: WeaviateLibArgs & { dimensions?: number }\n ): Promise {\n const weaviateStore = new this(embeddings, config);\n const collection = await weaviateStore.client.collections.exists(\n weaviateStore.indexName\n );\n if (!collection) {\n if (weaviateStore.schema) {\n await weaviateStore.client.collections.create(weaviateStore.schema);\n } else {\n if (config.tenant) {\n await weaviateStore.client.collections.create({\n name: weaviateStore.indexName,\n multiTenancy: configure.multiTenancy({\n enabled: true,\n autoTenantCreation: true,\n }),\n });\n } else {\n await weaviateStore.client.collections.create({\n name: weaviateStore.indexName,\n });\n }\n }\n }\n return weaviateStore;\n }\n\n /**\n * Method to add vectors and corresponding documents to the Weaviate\n * index.\n * @param vectors Array of vectors to be added.\n * @param documents Array of documents corresponding to the vectors.\n * @param options Optional parameter that can include specific IDs for the documents.\n * @returns An array of document IDs.\n */\n async addVectors(\n vectors: number[][],\n documents: Document[],\n options?: { ids?: string[] }\n ) {\n const documentIds = options?.ids ?? documents.map((_) => uuid.v4());\n const batch: DataObject[] = documents.map((document, index) => {\n if (Object.hasOwn(document.metadata, \"id\"))\n throw new Error(\n \"Document inserted to Weaviate vectorstore should not have `id` in their metadata.\"\n );\n const flattenedMetadata = flattenObjectForWeaviate(\n document.metadata\n ) as Record;\n return {\n id: documentIds[index],\n vectors: vectors[index],\n references: {},\n properties: {\n [this.textKey]: document.pageContent,\n ...flattenedMetadata,\n },\n };\n });\n\n try {\n const collection = this.client.collections.get(this.indexName);\n let response;\n if (this.tenant) {\n response = await collection\n .withTenant(this.tenant)\n .data.insertMany(batch);\n } else {\n response = await collection.data.insertMany(batch);\n }\n console.log(\n `Successfully imported batch of ${\n Object.values(response.uuids).length\n } items`\n );\n if (response.hasErrors) {\n console.log(\"this the error\", response.errors);\n throw new Error(\"Error in batch import!\");\n }\n return Object.values(response.uuids);\n } catch (error) {\n console.error(\"Error importing batch:\", error);\n throw error;\n }\n }\n\n /**\n * Method to add documents to the Weaviate index. It first generates\n * vectors for the documents using the embeddings, then adds the vectors\n * and documents to the index.\n * @param documents Array of documents to be added.\n * @param options Optional parameter that can include specific IDs for the documents.\n * @returns An array of document IDs.\n */\n async addDocuments(documents: Document[], options?: { ids?: string[] }) {\n return this.addVectors(\n await this.embeddings.embedDocuments(documents.map((d) => d.pageContent)),\n documents,\n options\n );\n }\n\n /**\n * Method to delete data from the Weaviate index. It can delete data based\n * on specific IDs or a filter.\n * @param params Object that includes either an array of IDs or a filter for the data to be deleted.\n * @returns Promise that resolves when the deletion is complete.\n */\n async delete(params: {\n ids?: string[];\n filter?: FilterValue;\n }): Promise {\n const { ids, filter } = params;\n const collection = this.client.collections.get(this.indexName);\n if (ids && ids.length > 0) {\n if (this.tenant) {\n await collection\n .withTenant(this.tenant)\n .data.deleteMany(collection.filter.byId().containsAny(ids));\n } else {\n await collection.data.deleteMany(\n collection.filter.byId().containsAny(ids)\n );\n }\n } else if (filter) {\n if (this.tenant) {\n await collection.withTenant(this.tenant).data.deleteMany(filter);\n } else {\n await collection.data.deleteMany(filter);\n }\n } else {\n throw new Error(\n `This method requires either \"ids\" or \"filter\" to be set in the input object`\n );\n }\n }\n\n /**\n * Hybrid search combines the results of a vector search and a\n * keyword (BM25F) search by fusing the two result sets.\n * @param query The query to search for.\n * @param options available options for the search. Check docs for complete list\n * @returns Promise that resolves the result of the search within the fetched collection.\n */\n async hybridSearch(\n query: string,\n options?: HybridOptions\n ): Promise {\n const collection = this.client.collections.get(this.indexName);\n let query_vector: number[] | undefined;\n if (!options?.vector) {\n query_vector = await this.embeddings.embedQuery(query);\n }\n\n const options_with_vector = {\n ...options,\n vector: options?.vector || query_vector,\n returnMetadata: [\n \"score\",\n ...((options?.returnMetadata as MetadataKeys) || []),\n ] as MetadataKeys,\n };\n let result;\n if (this.tenant) {\n result = await collection.withTenant(this.tenant).query.hybrid(query, {\n ...options_with_vector,\n });\n } else {\n result = await collection.query.hybrid(query, {\n ...options_with_vector,\n });\n }\n const documents: Document[] = [];\n\n for (const data of result.objects) {\n const { properties = {}, metadata = {} } = data ?? {};\n const { [this.textKey]: text, ...rest } = properties;\n\n documents.push(\n new Document({\n pageContent: String(text ?? \"\"),\n metadata: {\n ...rest,\n ...metadata,\n },\n id: data.uuid,\n })\n );\n }\n return documents;\n }\n\n /**\n * Weaviate's Retrieval Augmented Generation (RAG) combines information retrieval\n * with generative AI models. It first performs the search, then passes both\n * the search results and your prompt to a generative AI model before returning the generated response.\n * @param query The query to search for.\n * @param generate available options for the generation. Check docs for complete list\n * @param options available options for performing the hybrid search\n * @returns Promise that resolves the result of the search including the generated data.\n */\n async generate(\n query: string,\n generate: GenerateOptions,\n options?: BaseHybridOptions\n ): Promise {\n const collection = this.client.collections.get(this.indexName);\n let result;\n if (this.tenant) {\n result = await collection\n .withTenant(this.tenant)\n .generate.hybrid(\n query,\n { ...(generate || {}) },\n { ...(options || {}) }\n );\n } else {\n result = await collection.generate.hybrid(\n query,\n { ...(generate || {}) },\n { ...(options || {}) }\n );\n }\n const documents = [];\n for (const data of result.objects) {\n const { properties = {} } = data ?? {};\n const { [this.textKey]: text, ...rest } = properties;\n\n const doc = new WeaviateDocument({\n pageContent: String(text ?? \"\"),\n metadata: {\n ...rest,\n },\n id: data.uuid,\n });\n\n doc.generated = data.generative?.text;\n doc.vectors = data.vectors;\n doc.additional = data.metadata || {};\n\n documents.push(doc);\n }\n return documents;\n }\n\n /**\n * Method to perform a similarity search on the stored vectors in the\n * Weaviate index. It returns the top k most similar documents and their\n * similarity scores.\n * @param query The query vector.\n * @param k The number of most similar documents to return.\n * @param filter Optional filter to apply to the search.\n * @returns An array of tuples, where each tuple contains a document and its similarity score.\n */\n async similaritySearchVectorWithScore(\n query: number[],\n k: number,\n filter?: FilterValue\n ): Promise<[Document, number][]> {\n const resultsWithEmbedding =\n await this.similaritySearchVectorWithScoreAndEmbedding(query, k, filter);\n return resultsWithEmbedding.map(([document, score, _embedding]) => [\n document,\n score,\n ]);\n }\n\n /**\n * Method to perform a similarity search on the stored vectors in the\n * Weaviate index. It returns the top k most similar documents, their\n * similarity scores and embedding vectors.\n * @param query The query vector.\n * @param k The number of most similar documents to return.\n * @param filter Optional filter to apply to the search.\n * @returns An array of tuples, where each tuple contains a document, its similarity score and its embedding vector.\n */\n async similaritySearchVectorWithScoreAndEmbedding(\n query: number[],\n k: number,\n filter?: FilterValue\n ): Promise<[Document, number, number, number[]][]> {\n try {\n const collection = this.client.collections.get(this.indexName);\n // define query attributes to return\n // if no queryAttrs, show all properties\n const queryAttrs =\n this.queryAttrs.length > 1 ? this.queryAttrs : undefined;\n let result;\n if (this.tenant) {\n result = await collection\n .withTenant(this.tenant)\n .query.nearVector(query, {\n filters: filter,\n limit: k,\n returnMetadata: [\"distance\", \"score\"],\n returnProperties: queryAttrs,\n });\n } else {\n result = await collection.query.nearVector(query, {\n filters: filter,\n limit: k,\n includeVector: true,\n returnMetadata: [\"distance\", \"score\"],\n returnProperties: queryAttrs,\n });\n }\n\n const documents: [Document, number, number, number[]][] = [];\n\n for (const data of result.objects) {\n const { properties = {}, metadata = {} } = data ?? {};\n const { [this.textKey]: text, ...rest } = properties;\n\n documents.push([\n new Document({\n pageContent: String(text ?? \"\"),\n metadata: {\n ...rest,\n },\n id: data.uuid,\n }),\n metadata?.distance ?? 0,\n metadata?.score ?? 0,\n Object.values(data.vectors)[0],\n ]);\n }\n return documents;\n } catch (e) {\n throw Error(`Error in similaritySearch ${e}`);\n }\n }\n\n /**\n * Return documents selected using the maximal marginal relevance.\n * Maximal marginal relevance optimizes for similarity to the query AND diversity\n * among selected documents.\n *\n * @param {string} query - Text to look up documents similar to.\n * @param {number} options.k - Number of documents to return.\n * @param {number} options.fetchK - Number of documents to fetch before passing to the MMR algorithm.\n * @param {number} options.lambda - Number between 0 and 1 that determines the degree of diversity among the results,\n * where 0 corresponds to maximum diversity and 1 to minimum diversity.\n * @param {this[\"FilterType\"]} options.filter - Optional filter\n * @param _callbacks\n *\n * @returns {Promise} - List of documents selected by maximal marginal relevance.\n */\n override async maxMarginalRelevanceSearch(\n query: string,\n options: MaxMarginalRelevanceSearchOptions,\n _callbacks?: undefined\n ): Promise {\n const { k, fetchK = 20, lambda = 0.5, filter } = options;\n const queryEmbedding: number[] = await this.embeddings.embedQuery(query);\n const allResults: [Document, number, number, number[]][] =\n await this.similaritySearchVectorWithScoreAndEmbedding(\n queryEmbedding,\n fetchK,\n filter\n );\n const embeddingList = allResults.map(\n ([_doc, _distance, _score, embedding]) => embedding\n );\n const mmrIndexes = maximalMarginalRelevance(\n queryEmbedding,\n embeddingList,\n lambda,\n k\n );\n return mmrIndexes\n .filter((idx) => idx !== -1)\n .map((idx) => allResults[idx][0]);\n }\n\n /**\n * Static method to create a new `WeaviateStore` instance from a list of\n * texts. It first creates documents from the texts and metadata, then\n * adds the documents to the Weaviate index.\n * @param texts Array of texts.\n * @param metadatas Metadata for the texts. Can be a single object or an array of objects.\n * @param embeddings Embeddings to be used for the texts.\n * @param args Arguments required to create a new `WeaviateStore` instance.\n * @returns A new `WeaviateStore` instance.\n */\n static fromTexts(\n texts: string[],\n metadatas: object | object[],\n embeddings: EmbeddingsInterface,\n args: WeaviateLibArgs\n ): Promise {\n const docs: Document[] = [];\n for (let i = 0; i < texts.length; i += 1) {\n const metadata = Array.isArray(metadatas) ? metadatas[i] : metadatas;\n const newDoc = new Document({\n pageContent: texts[i],\n metadata,\n });\n docs.push(newDoc);\n }\n return WeaviateStore.fromDocuments(docs, embeddings, args);\n }\n\n /**\n * Static method to create a new `WeaviateStore` instance from a list of\n * documents. It adds the documents to the Weaviate index.\n * @param docs Array of documents.\n * @param embeddings Embeddings to be used for the documents.\n * @param args Arguments required to create a new `WeaviateStore` instance.\n * @returns A new `WeaviateStore` instance.\n */\n static async fromDocuments(\n docs: Document[],\n embeddings: EmbeddingsInterface,\n args: WeaviateLibArgs\n ): Promise {\n const instance = await this.initialize(embeddings, args);\n await instance.addDocuments(docs);\n return instance;\n }\n\n /**\n * Static method to create a new `WeaviateStore` instance from an existing\n * Weaviate index.\n * @param embeddings Embeddings to be used for the Weaviate index.\n * @param args Arguments required to create a new `WeaviateStore` instance.\n * @returns A new `WeaviateStore` instance.\n */\n static async fromExistingIndex(\n embeddings: EmbeddingsInterface,\n args: WeaviateLibArgs\n ): Promise {\n return new this(embeddings, args);\n }\n}\n"],"mappings":";;;;;;;;AA0BA,MAAa,2BAA2B,CAACA,QAAiC;CACxE,MAAMC,kBAA2C,CAAE;AACnD,MAAK,MAAM,OAAO,KAAK;AACrB,MAAI,CAAC,OAAO,OAAO,KAAK,IAAI,CAC1B;EAIF,MAAM,SAAS,IAAI,QAAQ,kBAAkB,IAAI;EAEjD,MAAM,QAAQ,IAAI;AAClB,MAAI,OAAO,UAAU,YAAY,CAAC,MAAM,QAAQ,MAAM,EAAE;GACtD,MAAM,kBAAkB,yBACtB,MACD;AACD,QAAK,MAAM,WAAW,gBACpB,KAAI,OAAO,OAAO,iBAAiB,QAAQ,EAAE;IAE3C,MAAM,aAAa,QAAQ,QAAQ,kBAAkB,IAAI;IACzD,gBAAgB,GAAG,OAAO,CAAC,EAAE,YAAY,IAAI,gBAAgB;GAC9D;EAEJ,WAAU,MAAM,QAAQ,MAAM,EAC7B;OAAI,MAAM,WAAW,GACnB,gBAAgB,UAAU;YAE1B,OAAO,MAAM,OAAO,YACpB,MAAM,MAAM,CAACC,OAAgB,OAAO,OAAO,OAAO,MAAM,GAAG,EAI3D,gBAAgB,UAAU;EAC3B,OAED,gBAAgB,UAAU;CAE7B;AACD,QAAO;AACR;AAmBD,IAAa,mBAAb,cAAsCC,oCAAS;CAC7C;CAEA;CAEA;AACD;;;;;;AAOD,IAAa,gBAAb,MAAa,sBAAsBC,0CAAY;CAG7C,AAAQ;CAER,AAAQ;CAER,AAAQ;CAER,AAAQ;CAER,AAAQ;CAER,AAAQ;CAER,mBAA2B;AACzB,SAAO;CACR;CAED,YAAmBC,YAAiCC,MAAuB;EACzE,MAAM,YAAY,KAAK;EADN;EAGjB,KAAK,SAAS,KAAK;EACnB,KAAK,YAAY,KAAK,aAAa,KAAK,QAAQ,QAAQ;EACxD,KAAK,UAAU,KAAK,WAAW;EAC/B,KAAK,aAAa,CAAC,KAAK,OAAQ;EAChC,KAAK,SAAS,KAAK;EACnB,KAAK,SAAS,KAAK;AAEnB,MAAI,KAAK,cACP,KAAK,aAAa,CAChB,GAAG,IAAI,IAAI,CACT,GAAG,KAAK,YACR,GAAG,KAAK,aAAa,OAAO,CAAC,MAAM;GAGjC,MAAM,aAAa,2BAA2B,KAAK,EAAE;AACrD,OAAI,CAAC,YACH,QAAQ,KACN,CAAC,sBAAsB,EAAE,EAAE,kCAAkC,CAAC,CAC/D;AAEH,UAAO;EACR,EAAC,AACH,EACF;CAEJ;CAED,aAAa,WACXD,YACAE,QACwB;EACxB,MAAM,gBAAgB,IAAI,KAAK,YAAY;EAC3C,MAAM,aAAa,MAAM,cAAc,OAAO,YAAY,OACxD,cAAc,UACf;AACD,MAAI,CAAC,WACH,KAAI,cAAc,QAChB,MAAM,cAAc,OAAO,YAAY,OAAO,cAAc,OAAO;WAE/D,OAAO,QACT,MAAM,cAAc,OAAO,YAAY,OAAO;GAC5C,MAAM,cAAc;GACpB,cAAcC,0BAAU,aAAa;IACnC,SAAS;IACT,oBAAoB;GACrB,EAAC;EACH,EAAC;OAEF,MAAM,cAAc,OAAO,YAAY,OAAO,EAC5C,MAAM,cAAc,UACrB,EAAC;AAIR,SAAO;CACR;;;;;;;;;CAUD,MAAM,WACJC,SACAC,WACAC,SACA;EACA,MAAM,cAAc,SAAS,OAAO,UAAU,IAAI,CAAC,MAAM,KAAK,IAAI,CAAC;EACnE,MAAMC,QAAiC,UAAU,IAAI,CAAC,UAAU,UAAU;AACxE,OAAI,OAAO,OAAO,SAAS,UAAU,KAAK,CACxC,OAAM,IAAI,MACR;GAEJ,MAAM,oBAAoB,yBACxB,SAAS,SACV;AACD,UAAO;IACL,IAAI,YAAY;IAChB,SAAS,QAAQ;IACjB,YAAY,CAAE;IACd,YAAY;MACT,KAAK,UAAU,SAAS;KACzB,GAAG;IACJ;GACF;EACF,EAAC;AAEF,MAAI;GACF,MAAM,aAAa,KAAK,OAAO,YAAY,IAAI,KAAK,UAAU;GAC9D,IAAI;AACJ,OAAI,KAAK,QACP,WAAW,MAAM,WACd,WAAW,KAAK,OAAO,CACvB,KAAK,WAAW,MAAM;QAEzB,WAAW,MAAM,WAAW,KAAK,WAAW,MAAM;GAEpD,QAAQ,IACN,CAAC,+BAA+B,EAC9B,OAAO,OAAO,SAAS,MAAM,CAAC,OAC/B,MAAM,CAAC,CACT;AACD,OAAI,SAAS,WAAW;IACtB,QAAQ,IAAI,kBAAkB,SAAS,OAAO;AAC9C,UAAM,IAAI,MAAM;GACjB;AACD,UAAO,OAAO,OAAO,SAAS,MAAM;EACrC,SAAQ,OAAO;GACd,QAAQ,MAAM,0BAA0B,MAAM;AAC9C,SAAM;EACP;CACF;;;;;;;;;CAUD,MAAM,aAAaF,WAAuBC,SAA8B;AACtE,SAAO,KAAK,WACV,MAAM,KAAK,WAAW,eAAe,UAAU,IAAI,CAAC,MAAM,EAAE,YAAY,CAAC,EACzE,WACA,QACD;CACF;;;;;;;CAQD,MAAM,OAAOE,QAGK;EAChB,MAAM,EAAE,KAAK,QAAQ,GAAG;EACxB,MAAM,aAAa,KAAK,OAAO,YAAY,IAAI,KAAK,UAAU;AAC9D,MAAI,OAAO,IAAI,SAAS,EACtB,KAAI,KAAK,QACP,MAAM,WACH,WAAW,KAAK,OAAO,CACvB,KAAK,WAAW,WAAW,OAAO,MAAM,CAAC,YAAY,IAAI,CAAC;OAE7D,MAAM,WAAW,KAAK,WACpB,WAAW,OAAO,MAAM,CAAC,YAAY,IAAI,CAC1C;WAEM,OACT,KAAI,KAAK,QACP,MAAM,WAAW,WAAW,KAAK,OAAO,CAAC,KAAK,WAAW,OAAO;OAEhE,MAAM,WAAW,KAAK,WAAW,OAAO;MAG1C,OAAM,IAAI,MACR,CAAC,2EAA2E,CAAC;CAGlF;;;;;;;;CASD,MAAM,aACJC,OACAC,SACqB;EACrB,MAAM,aAAa,KAAK,OAAO,YAAY,IAAI,KAAK,UAAU;EAC9D,IAAIC;AACJ,MAAI,CAAC,SAAS,QACZ,eAAe,MAAM,KAAK,WAAW,WAAW,MAAM;EAGxD,MAAM,sBAAsB;GAC1B,GAAG;GACH,QAAQ,SAAS,UAAU;GAC3B,gBAAgB,CACd,SACA,GAAK,SAAS,kBAAmC,CAAE,CACpD;EACF;EACD,IAAI;AACJ,MAAI,KAAK,QACP,SAAS,MAAM,WAAW,WAAW,KAAK,OAAO,CAAC,MAAM,OAAO,OAAO,EACpE,GAAG,oBACJ,EAAC;OAEF,SAAS,MAAM,WAAW,MAAM,OAAO,OAAO,EAC5C,GAAG,oBACJ,EAAC;EAEJ,MAAMN,YAAwB,CAAE;AAEhC,OAAK,MAAM,QAAQ,OAAO,SAAS;GACjC,MAAM,EAAE,aAAa,CAAE,GAAE,WAAW,CAAE,GAAE,GAAG,QAAQ,CAAE;GACrD,MAAM,EAAE,CAAC,KAAK,UAAU,KAAM,GAAG,MAAM,GAAG;GAE1C,UAAU,KACR,IAAIP,oCAAS;IACX,aAAa,OAAO,QAAQ,GAAG;IAC/B,UAAU;KACR,GAAG;KACH,GAAG;IACJ;IACD,IAAI,KAAK;GACV,GACF;EACF;AACD,SAAO;CACR;;;;;;;;;;CAWD,MAAM,SACJW,OACAG,UACAC,SAC6B;EAC7B,MAAM,aAAa,KAAK,OAAO,YAAY,IAAI,KAAK,UAAU;EAC9D,IAAI;AACJ,MAAI,KAAK,QACP,SAAS,MAAM,WACZ,WAAW,KAAK,OAAO,CACvB,SAAS,OACR,OACA,EAAE,GAAI,YAAY,CAAE,EAAG,GACvB,EAAE,GAAI,WAAW,CAAE,EAAG,EACvB;OAEH,SAAS,MAAM,WAAW,SAAS,OACjC,OACA,EAAE,GAAI,YAAY,CAAE,EAAG,GACvB,EAAE,GAAI,WAAW,CAAE,EAAG,EACvB;EAEH,MAAM,YAAY,CAAE;AACpB,OAAK,MAAM,QAAQ,OAAO,SAAS;GACjC,MAAM,EAAE,aAAa,CAAE,GAAE,GAAG,QAAQ,CAAE;GACtC,MAAM,EAAE,CAAC,KAAK,UAAU,KAAM,GAAG,MAAM,GAAG;GAE1C,MAAM,MAAM,IAAI,iBAAiB;IAC/B,aAAa,OAAO,QAAQ,GAAG;IAC/B,UAAU,EACR,GAAG,KACJ;IACD,IAAI,KAAK;GACV;GAED,IAAI,YAAY,KAAK,YAAY;GACjC,IAAI,UAAU,KAAK;GACnB,IAAI,aAAa,KAAK,YAAY,CAAE;GAEpC,UAAU,KAAK,IAAI;EACpB;AACD,SAAO;CACR;;;;;;;;;;CAWD,MAAM,gCACJC,OACAC,GACAC,QAC+B;EAC/B,MAAM,uBACJ,MAAM,KAAK,4CAA4C,OAAO,GAAG,OAAO;AAC1E,SAAO,qBAAqB,IAAI,CAAC,CAAC,UAAU,OAAO,WAAW,KAAK,CACjE,UACA,KACD,EAAC;CACH;;;;;;;;;;CAWD,MAAM,4CACJF,OACAC,GACAC,QACiD;AACjD,MAAI;GACF,MAAM,aAAa,KAAK,OAAO,YAAY,IAAI,KAAK,UAAU;GAG9D,MAAM,aACJ,KAAK,WAAW,SAAS,IAAI,KAAK,aAAa;GACjD,IAAI;AACJ,OAAI,KAAK,QACP,SAAS,MAAM,WACZ,WAAW,KAAK,OAAO,CACvB,MAAM,WAAW,OAAO;IACvB,SAAS;IACT,OAAO;IACP,gBAAgB,CAAC,YAAY,OAAQ;IACrC,kBAAkB;GACnB,EAAC;QAEJ,SAAS,MAAM,WAAW,MAAM,WAAW,OAAO;IAChD,SAAS;IACT,OAAO;IACP,eAAe;IACf,gBAAgB,CAAC,YAAY,OAAQ;IACrC,kBAAkB;GACnB,EAAC;GAGJ,MAAMC,YAAoD,CAAE;AAE5D,QAAK,MAAM,QAAQ,OAAO,SAAS;IACjC,MAAM,EAAE,aAAa,CAAE,GAAE,WAAW,CAAE,GAAE,GAAG,QAAQ,CAAE;IACrD,MAAM,EAAE,CAAC,KAAK,UAAU,KAAM,GAAG,MAAM,GAAG;IAE1C,UAAU,KAAK;KACb,IAAInB,oCAAS;MACX,aAAa,OAAO,QAAQ,GAAG;MAC/B,UAAU,EACR,GAAG,KACJ;MACD,IAAI,KAAK;KACV;KACD,UAAU,YAAY;KACtB,UAAU,SAAS;KACnB,OAAO,OAAO,KAAK,QAAQ,CAAC;IAC7B,EAAC;GACH;AACD,UAAO;EACR,SAAQ,GAAG;AACV,SAAM,MAAM,CAAC,0BAA0B,EAAE,GAAG,CAAC;EAC9C;CACF;;;;;;;;;;;;;;;;CAiBD,MAAe,2BACbW,OACAS,SACAC,YACqB;EACrB,MAAM,EAAE,GAAG,SAAS,IAAI,SAAS,IAAK,QAAQ,GAAG;EACjD,MAAMC,iBAA2B,MAAM,KAAK,WAAW,WAAW,MAAM;EACxE,MAAMC,aACJ,MAAM,KAAK,4CACT,gBACA,QACA,OACD;EACH,MAAM,gBAAgB,WAAW,IAC/B,CAAC,CAAC,MAAM,WAAW,QAAQ,UAAU,KAAK,UAC3C;EACD,MAAM,uEACJ,gBACA,eACA,QACA,EACD;AACD,SAAO,WACJ,OAAO,CAAC,QAAQ,QAAQ,GAAG,CAC3B,IAAI,CAAC,QAAQ,WAAW,KAAK,GAAG;CACpC;;;;;;;;;;;CAYD,OAAO,UACLC,OACAC,WACAvB,YACAC,MACwB;EACxB,MAAMuB,OAAmB,CAAE;AAC3B,OAAK,IAAI,IAAI,GAAG,IAAI,MAAM,QAAQ,KAAK,GAAG;GACxC,MAAM,WAAW,MAAM,QAAQ,UAAU,GAAG,UAAU,KAAK;GAC3D,MAAM,SAAS,IAAI1B,oCAAS;IAC1B,aAAa,MAAM;IACnB;GACD;GACD,KAAK,KAAK,OAAO;EAClB;AACD,SAAO,cAAc,cAAc,MAAM,YAAY,KAAK;CAC3D;;;;;;;;;CAUD,aAAa,cACX0B,MACAxB,YACAC,MACwB;EACxB,MAAM,WAAW,MAAM,KAAK,WAAW,YAAY,KAAK;EACxD,MAAM,SAAS,aAAa,KAAK;AACjC,SAAO;CACR;;;;;;;;CASD,aAAa,kBACXD,YACAC,MACwB;AACxB,SAAO,IAAI,KAAK,YAAY;CAC7B;AACF"}