gs-search 0.1.4 → 0.1.5

package/README.ja.md

@@ -134,7 +134,7 @@ const engine = new SearchEngine({
 
 ### SearchEngine
 
-- `constructor(options: ISearchEngineConfig)`: 新しいコアエンジンインスタンスを作成
+- `constructor(options: ISearchEngineOption)`: 新しいコアエンジンインスタンスを作成
 - `init(): Promise<void>`: エンジンを初期化
 - `addDocument(doc: IDocument): Promise<void>`: 単一ドキュメントを追加
 - `addDocuments(docs: IDocument[]): Promise<void>`: 複数ドキュメントを追加

package/README.ko.md

@@ -134,7 +134,7 @@ const engine = new SearchEngine({
 
 ### SearchEngine
 
-- `constructor(options: ISearchEngineConfig)`: 새로운 코어 엔진 인스턴스 생성
+- `constructor(options: ISearchEngineOption)`: 새로운 코어 엔진 인스턴스 생성
 - `init(): Promise<void>`: 엔진 초기화
 - `addDocument(doc: IDocument): Promise<void>`: 단일 문서 추가
 - `addDocuments(docs: IDocument[]): Promise<void>`: 다중 문서 추가

package/README.md

@@ -116,7 +116,7 @@ const customTokenizer = (text: string): string[] => {
 
 // Create engine with custom tokenizers
 const engine = new SearchEngine({
-  baseDir: 'search-data',
+  storage: new BrowserStorage('search-data'),
   indexingTokenizer: customTokenizer,
   searchTokenizer: customTokenizer
 });
@@ -127,7 +127,7 @@ const engine = new SearchEngine({
 ### SimpleSearch
 
 **Static Methods (No instance creation required):**
-- `configure(config: Partial<ISearchEngineConfig>): void`: Configure the search engine
+- `configure(config: Partial<ISearchEngineOption>): void`: Configure the search engine
 - `addDocument(doc: IDocument): Promise<void>`: Add a single document
 - `addDocuments(docs: IDocument[]): Promise<void>`: Add multiple documents
 - `addDocumentIfMissing(doc: IDocument): Promise<void>`: Add a single document if it doesn't exist
@@ -141,7 +141,7 @@ const engine = new SearchEngine({
 
 ### SearchEngine
 
-- `constructor(options: ISearchEngineConfig)`: Create a new core engine instance
+- `constructor(options: ISearchEngineOption)`: Create a new core engine instance
 - `init(): Promise<void>`: Initialize the engine
 - `addDocument(doc: IDocument): Promise<void>`: Add a single document
 - `addDocuments(docs: IDocument[]): Promise<void>`: Add multiple documents

package/README.zh-CN.md

@@ -225,7 +225,7 @@ SimpleSearch.configure({
 ### SimpleSearch
 
 **静态方法（无需实例创建）：**
-- `configure(config: Partial<ISearchEngineConfig>): void`: 配置搜索引擎
+- `configure(config: Partial<ISearchEngineOption>): void`: 配置搜索引擎
 - `addDocument(doc: IDocument): Promise<void>`: 添加单个文档
 - `addDocuments(docs: IDocument[]): Promise<void>`: 添加多个文档
 - `addDocumentIfMissing(doc: IDocument): Promise<void>`: 如果文档不存在则添加单个文档
@@ -239,7 +239,7 @@ SimpleSearch.configure({
 
 ### SearchEngine
 
-- `constructor(options: ISearchEngineConfig)`: 创建一个新的核心引擎实例
+- `constructor(options: ISearchEngineOption)`: 创建一个新的核心引擎实例
 - `init(): Promise<void>`: 初始化引擎
 - `addDocument(doc: IDocument): Promise<void>`: 添加单个文档
 - `addDocuments(docs: IDocument[]): Promise<void>`: 添加多个文档

package/lib/core.cjs

@@ -1 +1 @@
-"use strict";var e=require('./browser'),t=require('./node');const s="search_meta.json",n="deleted_ids.bin",i="added_ids.bin";class o{#e;#t={wordSegments:[],charSegments:[]};#s=new Set;#n=new Set;constructor(e){this.#e=e}async load(){const e=await this.#e.read(s);if(e){const t=(new TextDecoder).decode(e);this.#t=JSON.parse(t)}else this.#t={wordSegments:[],charSegments:[]};const t=await this.#e.read(n);if(t){const e=new DataView(t);let s=0;const n=t.byteLength;for(;s<n&&!(s+4>n);){const t=e.getUint32(s,!0);this.#s.add(t),s+=4,s<n&&30===e.getUint8(s)&&(s+=1)}}const o=await this.#e.read(i);if(o){const e=new DataView(o);let t=0;const s=o.byteLength;for(;t<s&&!(t+4>s);){const n=e.getUint32(t,!0);this.#n.add(n),t+=4,t<s&&30===e.getUint8(t)&&(t+=1)}}}async save(){const e=JSON.stringify(this.#t);if(await this.#e.write(s,(new TextEncoder).encode(e).buffer),0===this.#s.size)await this.#e.remove(n);else{const e=4*this.#s.size+this.#s.size,t=new ArrayBuffer(e),s=new DataView(t);let i=0;for(const e of this.#s)s.setUint32(i,e,!0),i+=4,s.setUint8(i,30),i+=1;await this.#e.write(n,t)}if(0===this.#n.size)await this.#e.remove(i);else{const e=4*this.#n.size+this.#n.size,t=new ArrayBuffer(e),s=new DataView(t);let n=0;for(const e of this.#n)s.setUint32(n,e,!0),n+=4,s.setUint8(n,30),n+=1;await this.#e.write(i,t)}}getSegments(e){return"word"===e?this.#t.wordSegments:this.#t.charSegments}getDeletedIds(){return this.#s}addDeletedId(e){this.#s.add(e)}isDeleted(e){return this.#s.has(e)}addAddedId(e){this.#n.add(e)}removeAddedId(e){this.#n.delete(e)}isAdded(e){return this.#n.has(e)}getAddedIds(){return this.#n}hasDocument(e){return this.#n.has(e)||this.#s.has(e)}getLastSegmentInfo(e){const t=this.getSegments(e);return 0===t.length?null:t[t.length-1]}updateSegment(e,t,s,n,i,o){const a="word"===e?this.#t.wordSegments:this.#t.charSegments;if(o)a.push({filename:t,start:s,end:n,tokenCount:i});else{const e=a[a.length-1];e&&e.filename===t&&(e.end=n,e.tokenCount=i)}}reset(){this.#t={wordSegments:[],charSegments:[]},this.#s.clear(),this.#n.clear()}}class a{static SEPARATOR=30;#e;constructor(e){this.#e=e}async appendBatch(e,t){if(0===t.length)return await this.#e.getFileSize(e);const s=new TextEncoder;let n=0;for(const e of t){n+=8;for(const t of e.tokens){n+=2+Math.min(s.encode(t).byteLength,65535)}n+=1}const i=new Uint8Array(n);let o=0;for(const e of t){const t=[];for(const n of e.tokens){const e=s.encode(n),i=e.byteLength>65535?e.slice(0,65535):e;t.push(i)}const n=new DataView(i.buffer,o);n.setUint32(0,e.id,!0),n.setUint32(4,t.length,!0),o+=8;for(const e of t)new DataView(i.buffer,o).setUint16(0,e.byteLength,!0),o+=2,i.set(e,o),o+=e.byteLength;i[o++]=a.SEPARATOR}return await this.#e.append(e,i.buffer),await this.#e.getFileSize(e)}async readRange(e,t,s){const n=await this.#e.readRange(e,t,s);if(!n||0===n.byteLength)return[];const i=new DataView(n),o=new Uint8Array(n),r=new TextDecoder,h=[];let d=0;const c=n.byteLength;for(;d<c&&!(d+8>c);){const e=i.getUint32(d,!0);d+=4;const t=i.getUint32(d,!0);d+=4;const s=[];for(let e=0;e<t&&!(d+2>c);e++){const e=i.getUint16(d,!0);if(d+=2,d+e>c)break;const t=new Uint8Array(n,d,e);s.push(r.decode(t)),d+=e}d<c&&o[d]===a.SEPARATOR&&(d+=1),h.push({id:e,tokens:s})}return h}async getCurrentSize(e){return await this.#e.getFileSize(e)}}function r(e,t=305419896){const s=e.length,n=s>>2;let i=0;for(;i<n;){let s=255&e.charCodeAt(i)|(255&e.charCodeAt(++i))<<8|(255&e.charCodeAt(++i))<<16|(255&e.charCodeAt(++i))<<24;++i,s=3432918353*(65535&s)+((3432918353*(s>>>16)&65535)<<16)&4294967295,s=s<<15|s>>>17,s=461845907*(65535&s)+((461845907*(s>>>16)&65535)<<16)&4294967295,t=27492+(65535&(t=5*(65535&(t=(t^=s)<<13|t>>>19))+((5*(t>>>16)&65535)<<16)&4294967295))+(((t>>>16)+58964&65535)<<16)}let o=0;const a=3&s;return a>0&&(a>=3&&(o^=(255&e.charCodeAt(i+2))<<16),a>=2&&(o^=(255&e.charCodeAt(i+1))<<8),a>=1&&(o^=255&e.charCodeAt(i)),o=3432918353*(65535&o)+((3432918353*(o>>>16)&65535)<<16)&4294967295,o=o<<15|o>>>17,o=461845907*(65535&o)+((461845907*(o>>>16)&65535)<<16)&4294967295,t^=o),t^=s,t=2246822507*(65535&(t^=t>>>16))+((2246822507*(t>>>16)&65535)<<16)&4294967295,t=3266489909*(65535&(t^=t>>>13))+((3266489909*(t>>>16)&65535)<<16)&4294967295,(t^=t>>>16)>>>0}class h{#i;#e;#o=null;#a=null;static hash(e){return r(e)}constructor(e,t){this.#i=e,this.#e=t}async loadIndex(){return!!this.#o||(this.#o=await this.#e.read(this.#i),!!this.#o&&(this.#a=new DataView(this.#o),!0))}async buildAndSave(e){const t=new Map;for(const s of e){const e=new Map;for(const n of s.tokens)e.has(n)||(e.set(n,!0),t.has(n)||t.set(n,{hash:h.hash(n),postings:[]}),t.get(n).postings.push(s.id))}const s=Array.from(t.entries());s.sort(([e,{hash:t}],[s,{hash:n}])=>t!==n?t-n:e.localeCompare(s));let n=0,i=0;for(const[e,{postings:t}]of s)n+=t.length,i+=e.length+1;const o=20*s.length,a=12+o+4*n,r=new ArrayBuffer(a+i),d=new DataView(r);d.setUint32(0,1229866072),d.setUint32(4,s.length),d.setUint32(8,a);let c=12,g=12+o,f=a;for(const[e,{hash:t,postings:n}]of s){d.setUint32(c,t),d.setUint32(c+4,e.length),d.setUint32(c+8,f),d.setUint32(c+12,g),d.setUint32(c+16,n.length),c+=20;for(let e=0;e<n.length;e++)d.setUint32(g,n[e],!0),g+=4;const s=(new TextEncoder).encode(e);for(let e=0;e<s.length;e++)d.setUint8(f++,s[e]);d.setUint8(f++,0)}await this.#e.write(this.#i,r),this.#o=r,this.#a=d}search(e){if(!this.#a||!this.#o)return[];const t=h.hash(e),s=this.#a.getUint32(4);let n=0,i=s-1;const o=12,a=20,r=new TextDecoder;for(;n<=i;){const h=n+i>>>1,d=o+h*a,c=this.#a.getUint32(d);if(c<t)n=h+1;else{if(!(c>t)){if(!(h>0&&this.#a.getUint32(o+(h-1)*a)===t||h<s-1&&this.#a.getUint32(o+(h+1)*a)===t)){const e=this.#a.getUint32(o+h*a+12),t=this.#a.getUint32(o+h*a+16),s=[];for(let n=0;n<t;n++)s.push(this.#a.getUint32(e+4*n,!0));return s}let n=h;for(;n>0;){const e=o+(n-1)*a;if(this.#a.getUint32(e)!==t)break;n--}for(let i=n;i<s;i++){const s=o+i*a;if(this.#a.getUint32(s)!==t)break;const n=this.#a.getUint32(s+4),h=this.#a.getUint32(s+8),d=new Uint8Array(this.#o,h,n);if(r.decode(d)===e){const e=this.#a.getUint32(s+12),t=this.#a.getUint32(s+16),n=[];for(let s=0;s<t;s++)n.push(this.#a.getUint32(e+4*s,!0));return n}}return[]}i=h-1}}return[]}}const d="word_cache.bin",c="char_cache.bin";exports.SearchEngine=class{#e;#t;#r;#h;#d=!1;#c;#g=!1;#f={word:0,char:0};constructor(s){if(!s.baseDir)throw new Error("SearchEngine requires 'baseDir' in config.");if(this.#c={wordSegmentTokenThreshold:1e5,charSegmentTokenThreshold:5e5,minWordTokenSave:0,minCharTokenSave:0,...s},(this.#c.minWordTokenSave||0)>=(this.#c.wordSegmentTokenThreshold||1e5))throw new Error("minWordTokenSave must be less than wordSegmentTokenThreshold");if((this.#c.minCharTokenSave||0)>=(this.#c.charSegmentTokenThreshold||5e5))throw new Error("minCharTokenSave must be less than charSegmentTokenThreshold");let n=null;if(this.#c.storage&&("object"==typeof this.#c.storage?n=this.#c.storage:"browser"===this.#c.storage?n=new e.BrowserStorage(this.#c.baseDir):"node"===this.#c.storage&&(n=new t.NodeStorage(this.#c.baseDir))),!n){const s=typeof navigator<"u"&&navigator?.storage?.getDirectory instanceof Function,i=typeof process<"u"&&null!=process.versions&&null!=process.versions.node;s?n=new e.BrowserStorage(this.#c.baseDir):i&&(n=new t.NodeStorage(this.#c.baseDir))}if(!n)throw new Error('Storage initialization failed. Please configure "storage" explicitly or ensure you are in a supported environment (Browser/Node).');this.#e=n,this.#t=new o(this.#e),this.#r=new a(this.#e),this.#h=new Map}async init(){if(this.#d)return;await this.#t.load();const e=[...this.#t.getSegments("word"),...this.#t.getSegments("char")];for(const t of e)this.#h.has(t.filename)||this.#h.set(t.filename,new h(t.filename,this.#e)),await this.#h.get(t.filename).loadIndex();this.#d=!0}startBatch(){this.#g=!0,this.#f={word:0,char:0}}async endBatch(){this.#g=!1,this.#f.word>0&&await this.#l("word",this.#f.word),this.#f.char>0&&await this.#l("char",this.#f.char),this.#f={word:0,char:0},await this.#t.save()}#w(e){try{if(typeof Intl<"u"&&"function"==typeof Intl.Segmenter&&"function"==typeof Array.from){const t=new Intl.Segmenter([],{granularity:"word"}).segment(e);if("object"==typeof t&&null!==t)return Array.from(t).filter(e=>e?.isWordLike).map(e=>e?.segment?.toLowerCase()||"")}}catch{}return e.toLowerCase().split(/[^a-z0-9\u4e00-\u9fa5]+/g).filter(e=>e.length>0)}#m(e){return this.#c.indexingTokenizer?this.#c.indexingTokenizer(e):this.#w(e.text)}#u(e){return this.#c.searchTokenizer?this.#c.searchTokenizer(e):this.#m(e)}async addDocument(e){return this.addDocuments([e])}async addDocumentIfMissing(e){return this.addDocumentsIfMissing([e])}async addDocumentsIfMissing(e){if(this.#d||await this.init(),0===e.length)return;const t=this.#t.getDeletedIds(),s=[],n=[],i=[];for(const o of e){if(t.has(o.id)||this.#t.isAdded(o.id))continue;const e=this.#m(o),a=[],r=[];for(const t of e)t.length>1?a.push(t):1===t.length&&r.push(t);a.length>0&&s.push({id:o.id,tokens:a}),r.length>0&&n.push({id:o.id,tokens:r}),i.push(o)}if(0===i.length)return;let o=0,a=0;if(s.length>0){await this.#r.appendBatch(d,s);for(const e of s)o+=e.tokens.length}if(n.length>0){await this.#r.appendBatch(c,n);for(const e of n)a+=e.tokens.length}for(const e of i)this.#t.addAddedId(e.id);this.#g?(this.#f.word+=o,this.#f.char+=a):(o>0&&await this.#l("word",o),a>0&&await this.#l("char",a),await this.#t.save())}async addDocuments(e){if(this.#d||await this.init(),0===e.length)return;const t=this.#t.getDeletedIds(),s=[],n=[];for(const i of e){if(t.has(i.id))throw new Error(`Document ID ${i.id} has been deleted and cannot be re-added.`);if(this.#t.isAdded(i.id))throw new Error(`Document ID ${i.id} already exists.`);const e=this.#m(i),o=[],a=[];for(const t of e)t.length>1?o.push(t):1===t.length&&a.push(t);o.length>0&&s.push({id:i.id,tokens:o}),a.length>0&&n.push({id:i.id,tokens:a})}let i=0,o=0;if(s.length>0){await this.#r.appendBatch(d,s);for(const e of s)i+=e.tokens.length}if(n.length>0){await this.#r.appendBatch(c,n);for(const e of n)o+=e.tokens.length}for(const t of e)this.#t.addAddedId(t.id);this.#g?(this.#f.word+=i,this.#f.char+=o):(i>0&&await this.#l("word",i),o>0&&await this.#l("char",o),await this.#t.save())}async#l(e,t){const s="word"===e?d:c,n=await this.#r.getCurrentSize(s),i="word"===e?this.#c.wordSegmentTokenThreshold||1e5:this.#c.charSegmentTokenThreshold||5e5,o="word"===e?this.#c.minWordTokenSave||0:this.#c.minCharTokenSave||0,a=this.#t.getLastSegmentInfo(e);let r,g,f,l;const w=()=>{const t=this.#t.getSegments(e).length+1;return`${e}_seg_${t}.bin`};if(a){const e=a.tokenCount;e>=i||e+t>=i?(r=w(),f=!0,g=a.end,l=t):(r=a.filename,f=!1,g=a.start,l=e+t)}else r=w(),f=!0,g=0,l=t;if(l<o)return void this.#t.updateSegment(e,r,g,n,l,f);const m=await this.#r.readRange(s,g,n);let u=this.#h.get(r);u||(u=new h(r,this.#e),this.#h.set(r,u)),await u.buildAndSave(m),this.#t.updateSegment(e,r,g,n,l,f)}async search(e,t){this.#d||await this.init();const s="string"==typeof e?{text:e}:e,n=this.#u(s),i=n.filter(e=>e.length>1),o=n.filter(e=>1===e.length),a=this.#t.getDeletedIds(),r=new Map,d=new Map,c=e=>{const t=this.#t.getSegments(e);for(const e of t){const t=e.filename;!this.#h.has(t)&&!d.has(t)&&d.set(t,new h(t,this.#e))}};c("word"),c("char"),await Promise.all(Array.from(d.entries()).map(([e,t])=>t.loadIndex().then(s=>{s&&this.#h.set(e,t)})));const g=async(e,t)=>{if(0===t.length)return;const s=this.#t.getSegments(e);for(const e of s){const s=e.filename,n=this.#h.get(s);if(n)for(const e of t){const t=n.search(e),s=1+.1*e.length;for(const n of t)if(!a.has(n))if(r.has(n)){const t=r.get(n);t.score+=s,t.tokens.add(e)}else r.set(n,{score:0,tokens:new Set([e])})}}};await g("word",i),await g("char",o);const f=[];return r.forEach((e,t)=>{f.push({id:t,score:e.score,tokens:Array.from(e.tokens)})}),f.sort((e,t)=>t.score-e.score),"number"==typeof t&&t>0?f.slice(0,t):f}async removeDocument(e){this.#d||await this.init(),this.#t.addDeletedId(e),this.#t.removeAddedId(e),await this.#t.save()}async clearAll(){await this.#e.clearAll(),this.#h.clear(),this.#t.reset(),this.#d=!1,this.#g=!1,this.#f={word:0,char:0}}async getStatus(){return this.#d||await this.init(),{wordSegments:this.#t.getSegments("word").length,charSegments:this.#t.getSegments("char").length,deleted:this.#t.getDeletedIds().size,wordCacheSize:await this.#r.getCurrentSize(d),charCacheSize:await this.#r.getCurrentSize(c),inBatch:this.#g}}async hasDocument(e){return this.#d||await this.init(),this.#t.hasDocument(e)}},exports.hash=r,exports.murmur3_32=r;
+"use strict";const e="search_meta.json",t="deleted_ids.bin",s="added_ids.bin";class n{#e;#t={wordSegments:[],charSegments:[]};#s=new Set;#n=new Set;constructor(e){this.#e=e}async load(){const n=await this.#e.read(e);if(n){const e=(new TextDecoder).decode(n);this.#t=JSON.parse(e)}else this.#t={wordSegments:[],charSegments:[]};const i=await this.#e.read(t);if(i){const e=new DataView(i);let t=0;const s=i.byteLength;for(;t<s&&!(t+4>s);){const n=e.getUint32(t,!0);this.#s.add(n),t+=4,t<s&&30===e.getUint8(t)&&(t+=1)}}const a=await this.#e.read(s);if(a){const e=new DataView(a);let t=0;const s=a.byteLength;for(;t<s&&!(t+4>s);){const n=e.getUint32(t,!0);this.#n.add(n),t+=4,t<s&&30===e.getUint8(t)&&(t+=1)}}}async save(){const n=JSON.stringify(this.#t);if(await this.#e.write(e,(new TextEncoder).encode(n).buffer),0===this.#s.size)await this.#e.remove(t);else{const e=4*this.#s.size+this.#s.size,s=new ArrayBuffer(e),n=new DataView(s);let i=0;for(const e of this.#s)n.setUint32(i,e,!0),i+=4,n.setUint8(i,30),i+=1;await this.#e.write(t,s)}if(0===this.#n.size)await this.#e.remove(s);else{const e=4*this.#n.size+this.#n.size,t=new ArrayBuffer(e),n=new DataView(t);let i=0;for(const e of this.#n)n.setUint32(i,e,!0),i+=4,n.setUint8(i,30),i+=1;await this.#e.write(s,t)}}getSegments(e){return"word"===e?this.#t.wordSegments:this.#t.charSegments}getDeletedIds(){return this.#s}addDeletedId(e){this.#s.add(e)}isDeleted(e){return this.#s.has(e)}addAddedId(e){this.#n.add(e)}removeAddedId(e){this.#n.delete(e)}isAdded(e){return this.#n.has(e)}getAddedIds(){return this.#n}hasDocument(e){return this.#n.has(e)||this.#s.has(e)}getLastSegmentInfo(e){const t=this.getSegments(e);return 0===t.length?null:t[t.length-1]}updateSegment(e,t,s,n,i,a){const o="word"===e?this.#t.wordSegments:this.#t.charSegments;if(a)o.push({filename:t,start:s,end:n,tokenCount:i});else{const e=o[o.length-1];e&&e.filename===t&&(e.end=n,e.tokenCount=i)}}reset(){this.#t={wordSegments:[],charSegments:[]},this.#s.clear(),this.#n.clear()}}class i{static SEPARATOR=30;#e;constructor(e){this.#e=e}async appendBatch(e,t){if(0===t.length)return await this.#e.getFileSize(e);const s=new TextEncoder;let n=0;for(const e of t){n+=8;for(const t of e.tokens){n+=2+Math.min(s.encode(t).byteLength,65535)}n+=1}const a=new Uint8Array(n);let o=0;for(const e of t){const t=[];for(const n of e.tokens){const e=s.encode(n),i=e.byteLength>65535?e.slice(0,65535):e;t.push(i)}const n=new DataView(a.buffer,o);n.setUint32(0,e.id,!0),n.setUint32(4,t.length,!0),o+=8;for(const e of t)new DataView(a.buffer,o).setUint16(0,e.byteLength,!0),o+=2,a.set(e,o),o+=e.byteLength;a[o++]=i.SEPARATOR}return await this.#e.append(e,a.buffer),await this.#e.getFileSize(e)}async readRange(e,t,s){const n=await this.#e.readRange(e,t,s);if(!n||0===n.byteLength)return[];const a=new DataView(n),o=new Uint8Array(n),r=new TextDecoder,h=[];let d=0;const c=n.byteLength;for(;d<c&&!(d+8>c);){const e=a.getUint32(d,!0);d+=4;const t=a.getUint32(d,!0);d+=4;const s=[];for(let e=0;e<t&&!(d+2>c);e++){const e=a.getUint16(d,!0);if(d+=2,d+e>c)break;const t=new Uint8Array(n,d,e);s.push(r.decode(t)),d+=e}d<c&&o[d]===i.SEPARATOR&&(d+=1),h.push({id:e,tokens:s})}return h}async getCurrentSize(e){return await this.#e.getFileSize(e)}}function a(e,t=305419896){const s=e.length,n=s>>2;let i=0;for(;i<n;){let s=255&e.charCodeAt(i)|(255&e.charCodeAt(++i))<<8|(255&e.charCodeAt(++i))<<16|(255&e.charCodeAt(++i))<<24;++i,s=3432918353*(65535&s)+((3432918353*(s>>>16)&65535)<<16)&4294967295,s=s<<15|s>>>17,s=461845907*(65535&s)+((461845907*(s>>>16)&65535)<<16)&4294967295,t=27492+(65535&(t=5*(65535&(t=(t^=s)<<13|t>>>19))+((5*(t>>>16)&65535)<<16)&4294967295))+(((t>>>16)+58964&65535)<<16)}let a=0;const o=3&s;return o>0&&(o>=3&&(a^=(255&e.charCodeAt(i+2))<<16),o>=2&&(a^=(255&e.charCodeAt(i+1))<<8),o>=1&&(a^=255&e.charCodeAt(i)),a=3432918353*(65535&a)+((3432918353*(a>>>16)&65535)<<16)&4294967295,a=a<<15|a>>>17,a=461845907*(65535&a)+((461845907*(a>>>16)&65535)<<16)&4294967295,t^=a),t^=s,t=2246822507*(65535&(t^=t>>>16))+((2246822507*(t>>>16)&65535)<<16)&4294967295,t=3266489909*(65535&(t^=t>>>13))+((3266489909*(t>>>16)&65535)<<16)&4294967295,(t^=t>>>16)>>>0}class o{#i;#e;#a=null;#o=null;static hash(e){return a(e)}constructor(e,t){this.#i=e,this.#e=t}async loadIndex(){return!!this.#a||(this.#a=await this.#e.read(this.#i),!!this.#a&&(this.#o=new DataView(this.#a),!0))}async buildAndSave(e){const t=new Map;for(const s of e){const e=new Map;for(const n of s.tokens)e.has(n)||(e.set(n,!0),t.has(n)||t.set(n,{hash:o.hash(n),postings:[]}),t.get(n).postings.push(s.id))}const s=Array.from(t.entries());s.sort(([e,{hash:t}],[s,{hash:n}])=>t!==n?t-n:e.localeCompare(s));let n=0,i=0;for(const[e,{postings:t}]of s)n+=t.length,i+=e.length+1;const a=20*s.length,r=12+a+4*n,h=new ArrayBuffer(r+i),d=new DataView(h);d.setUint32(0,1229866072),d.setUint32(4,s.length),d.setUint32(8,r);let c=12,g=12+a,f=r;for(const[e,{hash:t,postings:n}]of s){d.setUint32(c,t),d.setUint32(c+4,e.length),d.setUint32(c+8,f),d.setUint32(c+12,g),d.setUint32(c+16,n.length),c+=20;for(let e=0;e<n.length;e++)d.setUint32(g,n[e],!0),g+=4;const s=(new TextEncoder).encode(e);for(let e=0;e<s.length;e++)d.setUint8(f++,s[e]);d.setUint8(f++,0)}await this.#e.write(this.#i,h),this.#a=h,this.#o=d}search(e){if(!this.#o||!this.#a)return[];const t=o.hash(e),s=this.#o.getUint32(4);let n=0,i=s-1;const a=12,r=20,h=new TextDecoder;for(;n<=i;){const o=n+i>>>1,d=a+o*r,c=this.#o.getUint32(d);if(c<t)n=o+1;else{if(!(c>t)){if(!(o>0&&this.#o.getUint32(a+(o-1)*r)===t||o<s-1&&this.#o.getUint32(a+(o+1)*r)===t)){const e=this.#o.getUint32(a+o*r+12),t=this.#o.getUint32(a+o*r+16),s=[];for(let n=0;n<t;n++)s.push(this.#o.getUint32(e+4*n,!0));return s}let n=o;for(;n>0;){const e=a+(n-1)*r;if(this.#o.getUint32(e)!==t)break;n--}for(let i=n;i<s;i++){const s=a+i*r;if(this.#o.getUint32(s)!==t)break;const n=this.#o.getUint32(s+4),o=this.#o.getUint32(s+8),d=new Uint8Array(this.#a,o,n);if(h.decode(d)===e){const e=this.#o.getUint32(s+12),t=this.#o.getUint32(s+16),n=[];for(let s=0;s<t;s++)n.push(this.#o.getUint32(e+4*s,!0));return n}}return[]}i=o-1}}return[]}}const r=({text:e})=>{try{if(typeof Intl<"u"&&"function"==typeof Intl.Segmenter&&"function"==typeof Array.from){const t=new Intl.Segmenter([],{granularity:"word"}).segment(e);if("object"==typeof t&&null!==t)return Array.from(t).filter(e=>e?.isWordLike).map(e=>e?.segment?.toLowerCase()||"")}}catch{}return e.toLowerCase().split(/[^a-z0-9\u4e00-\u9fa5]+/g).filter(e=>e.length>0)},h="word_cache.bin",d="char_cache.bin";exports.SearchEngine=class{#e;#t;#r;#h;#d=!1;#c;#g=!1;#f={word:0,char:0};constructor(e){if(this.#c={wordSegmentTokenThreshold:1e5,charSegmentTokenThreshold:5e5,minWordTokenSave:0,minCharTokenSave:0,indexingTokenizer:e.indexingTokenizer||r,...e},(this.#c.minWordTokenSave||0)>=(this.#c.wordSegmentTokenThreshold||1e5))throw new Error("minWordTokenSave must be less than wordSegmentTokenThreshold");if((this.#c.minCharTokenSave||0)>=(this.#c.charSegmentTokenThreshold||5e5))throw new Error("minCharTokenSave must be less than charSegmentTokenThreshold");this.#e=e.storage,this.#t=new n(this.#e),this.#r=new i(this.#e),this.#h=new Map}startBatch(){this.#g=!0,this.#f={word:0,char:0}}async endBatch(){this.#g=!1,this.#f.word>0&&await this.#l("word",this.#f.word),this.#f.char>0&&await this.#l("char",this.#f.char),this.#f={word:0,char:0},await this.#t.save()}async addDocument(e){return this.addDocuments([e])}async addDocumentIfMissing(e){return this.addDocumentsIfMissing([e])}async addDocumentsIfMissing(e){if(this.#d||await this.#w(),0===e.length)return;const t=this.#t.getDeletedIds(),s=[],n=[],i=[];for(const a of e){if(t.has(a.id)||this.#t.isAdded(a.id))continue;const e=this.#m(a),o=[],r=[];for(const t of e)t.length>1?o.push(t):1===t.length&&r.push(t);o.length>0&&s.push({id:a.id,tokens:o}),r.length>0&&n.push({id:a.id,tokens:r}),i.push(a)}if(0===i.length)return;let a=0,o=0;if(s.length>0){await this.#r.appendBatch(h,s);for(const e of s)a+=e.tokens.length}if(n.length>0){await this.#r.appendBatch(d,n);for(const e of n)o+=e.tokens.length}for(const e of i)this.#t.addAddedId(e.id);this.#g?(this.#f.word+=a,this.#f.char+=o):(a>0&&await this.#l("word",a),o>0&&await this.#l("char",o),await this.#t.save())}async addDocuments(e){if(this.#d||await this.#w(),0===e.length)return;const t=this.#t.getDeletedIds(),s=[],n=[];for(const i of e){if(t.has(i.id))throw new Error(`Document ID ${i.id} has been deleted and cannot be re-added.`);if(this.#t.isAdded(i.id))throw new Error(`Document ID ${i.id} already exists.`);const e=this.#m(i),a=[],o=[];for(const t of e)t.length>1?a.push(t):1===t.length&&o.push(t);a.length>0&&s.push({id:i.id,tokens:a}),o.length>0&&n.push({id:i.id,tokens:o})}let i=0,a=0;if(s.length>0){await this.#r.appendBatch(h,s);for(const e of s)i+=e.tokens.length}if(n.length>0){await this.#r.appendBatch(d,n);for(const e of n)a+=e.tokens.length}for(const t of e)this.#t.addAddedId(t.id);this.#g?(this.#f.word+=i,this.#f.char+=a):(i>0&&await this.#l("word",i),a>0&&await this.#l("char",a),await this.#t.save())}async search(e,t){this.#d||await this.#w();const s="string"==typeof e?{text:e}:e,n=this.#u(s),i=n.filter(e=>e.length>1),a=n.filter(e=>1===e.length),r=this.#t.getDeletedIds(),h=new Map,d=new Map,c=e=>{const t=this.#t.getSegments(e);for(const e of t){const t=e.filename;!this.#h.has(t)&&!d.has(t)&&d.set(t,new o(t,this.#e))}};c("word"),c("char"),await Promise.all(Array.from(d.entries()).map(([e,t])=>t.loadIndex().then(s=>{s&&this.#h.set(e,t)})));const g=async(e,t)=>{if(0===t.length)return;const s=this.#t.getSegments(e);for(const e of s){const s=e.filename,n=this.#h.get(s);if(n)for(const e of t){const t=n.search(e),s=1+.1*e.length;for(const n of t)if(!r.has(n))if(h.has(n)){const t=h.get(n);t.score+=s,t.tokens.add(e)}else h.set(n,{score:0,tokens:new Set([e])})}}};await g("word",i),await g("char",a);const f=[];return h.forEach((e,t)=>{f.push({id:t,score:e.score,tokens:Array.from(e.tokens)})}),f.sort((e,t)=>t.score-e.score),"number"==typeof t&&t>0?f.slice(0,t):f}async removeDocument(e){this.#d||await this.#w(),this.#t.addDeletedId(e),this.#t.removeAddedId(e),await this.#t.save()}async clearAll(){await this.#e.clearAll(),this.#h.clear(),this.#t.reset(),this.#d=!1,this.#g=!1,this.#f={word:0,char:0}}async getStatus(){return this.#d||await this.#w(),{wordSegments:this.#t.getSegments("word").length,charSegments:this.#t.getSegments("char").length,deleted:this.#t.getDeletedIds().size,wordCacheSize:await this.#r.getCurrentSize(h),charCacheSize:await this.#r.getCurrentSize(d),inBatch:this.#g}}async hasDocument(e){return this.#d||await this.#w(),this.#t.hasDocument(e)}async#w(){if(this.#d)return;await this.#t.load();const e=[...this.#t.getSegments("word"),...this.#t.getSegments("char")];for(const t of e)this.#h.has(t.filename)||this.#h.set(t.filename,new o(t.filename,this.#e)),await this.#h.get(t.filename).loadIndex();this.#d=!0}#m(e){return this.#c.indexingTokenizer(e)}#u(e){return this.#c.searchTokenizer?this.#c.searchTokenizer(e):this.#m(e)}async#l(e,t){const s="word"===e?h:d,n=await this.#r.getCurrentSize(s),i="word"===e?this.#c.wordSegmentTokenThreshold||1e5:this.#c.charSegmentTokenThreshold||5e5,a="word"===e?this.#c.minWordTokenSave||0:this.#c.minCharTokenSave||0,r=this.#t.getLastSegmentInfo(e);let c,g,f,l;const w=()=>{const t=this.#t.getSegments(e).length+1;return`${e}_seg_${t}.bin`};if(r){const e=r.tokenCount;e>=i||e+t>=i?(c=w(),f=!0,g=r.end,l=t):(c=r.filename,f=!1,g=r.start,l=e+t)}else c=w(),f=!0,g=0,l=t;if(l<a)return void this.#t.updateSegment(e,c,g,n,l,f);const m=await this.#r.readRange(s,g,n);let u=this.#h.get(c);u||(u=new o(c,this.#e),this.#h.set(c,u)),await u.buildAndSave(m),this.#t.updateSegment(e,c,g,n,l,f)}},exports.hash=a,exports.murmur3_32=a;

package/lib/core.d.ts

@@ -1,12 +1,11 @@
-import { ISearchEngineConfig, IDocument, IDocumentBase, IResult } from './type';
+import { ISearchEngine, ISearchEngineOption, IDocument, IDocumentBase, IResult, ISearchEngineStatus } from './type';
 
 /**
  * 核心搜索引擎类 (多实例支持)
  */
-declare class SearchEngine {
+declare class SearchEngine implements ISearchEngine {
     #private;
-    constructor(config: ISearchEngineConfig);
-    init(): Promise<void>;
+    constructor(config: ISearchEngineOption);
     /**
      * 开启批处理
      * 批处理期间 addDocuments 只写入缓存，不触发索引段构建
@@ -29,17 +28,10 @@ declare class SearchEngine {
      */
     addDocumentsIfMissing<T extends IDocument = IDocument>(docs: T[]): Promise<void>;
     addDocuments<T extends IDocument = IDocument>(docs: T[]): Promise<void>;
-    search<T extends IDocumentBase = IDocumentBase>(query: T | string, limit?: number): Promise<IResult[]>;
+    search<T extends IDocumentBase | string = any>(query: T, limit?: number): Promise<IResult[]>;
     removeDocument(id: number): Promise<void>;
     clearAll(): Promise<void>;
-    getStatus(): Promise<{
-        wordSegments: number;
-        charSegments: number;
-        deleted: number;
-        wordCacheSize: number;
-        charCacheSize: number;
-        inBatch: boolean;
-    }>;
+    getStatus(): Promise<ISearchEngineStatus>;
     /**
      * 检查文档ID是否曾经添加过（包括已删除的）
      * @param id 文档ID

package/lib/core.js

@@ -1 +1 @@
-import{BrowserStorage as e}from'./browser';import{NodeStorage as t}from'./node';const s="search_meta.json",n="deleted_ids.bin",i="added_ids.bin";class o{#e;#t={wordSegments:[],charSegments:[]};#s=new Set;#n=new Set;constructor(e){this.#e=e}async load(){const e=await this.#e.read(s);if(e){const t=(new TextDecoder).decode(e);this.#t=JSON.parse(t)}else this.#t={wordSegments:[],charSegments:[]};const t=await this.#e.read(n);if(t){const e=new DataView(t);let s=0;const n=t.byteLength;for(;s<n&&!(s+4>n);){const t=e.getUint32(s,!0);this.#s.add(t),s+=4,s<n&&30===e.getUint8(s)&&(s+=1)}}const o=await this.#e.read(i);if(o){const e=new DataView(o);let t=0;const s=o.byteLength;for(;t<s&&!(t+4>s);){const n=e.getUint32(t,!0);this.#n.add(n),t+=4,t<s&&30===e.getUint8(t)&&(t+=1)}}}async save(){const e=JSON.stringify(this.#t);if(await this.#e.write(s,(new TextEncoder).encode(e).buffer),0===this.#s.size)await this.#e.remove(n);else{const e=4*this.#s.size+this.#s.size,t=new ArrayBuffer(e),s=new DataView(t);let i=0;for(const e of this.#s)s.setUint32(i,e,!0),i+=4,s.setUint8(i,30),i+=1;await this.#e.write(n,t)}if(0===this.#n.size)await this.#e.remove(i);else{const e=4*this.#n.size+this.#n.size,t=new ArrayBuffer(e),s=new DataView(t);let n=0;for(const e of this.#n)s.setUint32(n,e,!0),n+=4,s.setUint8(n,30),n+=1;await this.#e.write(i,t)}}getSegments(e){return"word"===e?this.#t.wordSegments:this.#t.charSegments}getDeletedIds(){return this.#s}addDeletedId(e){this.#s.add(e)}isDeleted(e){return this.#s.has(e)}addAddedId(e){this.#n.add(e)}removeAddedId(e){this.#n.delete(e)}isAdded(e){return this.#n.has(e)}getAddedIds(){return this.#n}hasDocument(e){return this.#n.has(e)||this.#s.has(e)}getLastSegmentInfo(e){const t=this.getSegments(e);return 0===t.length?null:t[t.length-1]}updateSegment(e,t,s,n,i,o){const a="word"===e?this.#t.wordSegments:this.#t.charSegments;if(o)a.push({filename:t,start:s,end:n,tokenCount:i});else{const e=a[a.length-1];e&&e.filename===t&&(e.end=n,e.tokenCount=i)}}reset(){this.#t={wordSegments:[],charSegments:[]},this.#s.clear(),this.#n.clear()}}class a{static SEPARATOR=30;#e;constructor(e){this.#e=e}async appendBatch(e,t){if(0===t.length)return await this.#e.getFileSize(e);const s=new TextEncoder;let n=0;for(const e of t){n+=8;for(const t of e.tokens){n+=2+Math.min(s.encode(t).byteLength,65535)}n+=1}const i=new Uint8Array(n);let o=0;for(const e of t){const t=[];for(const n of e.tokens){const e=s.encode(n),i=e.byteLength>65535?e.slice(0,65535):e;t.push(i)}const n=new DataView(i.buffer,o);n.setUint32(0,e.id,!0),n.setUint32(4,t.length,!0),o+=8;for(const e of t)new DataView(i.buffer,o).setUint16(0,e.byteLength,!0),o+=2,i.set(e,o),o+=e.byteLength;i[o++]=a.SEPARATOR}return await this.#e.append(e,i.buffer),await this.#e.getFileSize(e)}async readRange(e,t,s){const n=await this.#e.readRange(e,t,s);if(!n||0===n.byteLength)return[];const i=new DataView(n),o=new Uint8Array(n),r=new TextDecoder,h=[];let d=0;const c=n.byteLength;for(;d<c&&!(d+8>c);){const e=i.getUint32(d,!0);d+=4;const t=i.getUint32(d,!0);d+=4;const s=[];for(let e=0;e<t&&!(d+2>c);e++){const e=i.getUint16(d,!0);if(d+=2,d+e>c)break;const t=new Uint8Array(n,d,e);s.push(r.decode(t)),d+=e}d<c&&o[d]===a.SEPARATOR&&(d+=1),h.push({id:e,tokens:s})}return h}async getCurrentSize(e){return await this.#e.getFileSize(e)}}function r(e,t=305419896){const s=e.length,n=s>>2;let i=0;for(;i<n;){let s=255&e.charCodeAt(i)|(255&e.charCodeAt(++i))<<8|(255&e.charCodeAt(++i))<<16|(255&e.charCodeAt(++i))<<24;++i,s=3432918353*(65535&s)+((3432918353*(s>>>16)&65535)<<16)&4294967295,s=s<<15|s>>>17,s=461845907*(65535&s)+((461845907*(s>>>16)&65535)<<16)&4294967295,t=27492+(65535&(t=5*(65535&(t=(t^=s)<<13|t>>>19))+((5*(t>>>16)&65535)<<16)&4294967295))+(((t>>>16)+58964&65535)<<16)}let o=0;const a=3&s;return a>0&&(a>=3&&(o^=(255&e.charCodeAt(i+2))<<16),a>=2&&(o^=(255&e.charCodeAt(i+1))<<8),a>=1&&(o^=255&e.charCodeAt(i)),o=3432918353*(65535&o)+((3432918353*(o>>>16)&65535)<<16)&4294967295,o=o<<15|o>>>17,o=461845907*(65535&o)+((461845907*(o>>>16)&65535)<<16)&4294967295,t^=o),t^=s,t=2246822507*(65535&(t^=t>>>16))+((2246822507*(t>>>16)&65535)<<16)&4294967295,t=3266489909*(65535&(t^=t>>>13))+((3266489909*(t>>>16)&65535)<<16)&4294967295,(t^=t>>>16)>>>0}class h{#i;#e;#o=null;#a=null;static hash(e){return r(e)}constructor(e,t){this.#i=e,this.#e=t}async loadIndex(){return!!this.#o||(this.#o=await this.#e.read(this.#i),!!this.#o&&(this.#a=new DataView(this.#o),!0))}async buildAndSave(e){const t=new Map;for(const s of e){const e=new Map;for(const n of s.tokens)e.has(n)||(e.set(n,!0),t.has(n)||t.set(n,{hash:h.hash(n),postings:[]}),t.get(n).postings.push(s.id))}const s=Array.from(t.entries());s.sort(([e,{hash:t}],[s,{hash:n}])=>t!==n?t-n:e.localeCompare(s));let n=0,i=0;for(const[e,{postings:t}]of s)n+=t.length,i+=e.length+1;const o=20*s.length,a=12+o+4*n,r=new ArrayBuffer(a+i),d=new DataView(r);d.setUint32(0,1229866072),d.setUint32(4,s.length),d.setUint32(8,a);let c=12,g=12+o,f=a;for(const[e,{hash:t,postings:n}]of s){d.setUint32(c,t),d.setUint32(c+4,e.length),d.setUint32(c+8,f),d.setUint32(c+12,g),d.setUint32(c+16,n.length),c+=20;for(let e=0;e<n.length;e++)d.setUint32(g,n[e],!0),g+=4;const s=(new TextEncoder).encode(e);for(let e=0;e<s.length;e++)d.setUint8(f++,s[e]);d.setUint8(f++,0)}await this.#e.write(this.#i,r),this.#o=r,this.#a=d}search(e){if(!this.#a||!this.#o)return[];const t=h.hash(e),s=this.#a.getUint32(4);let n=0,i=s-1;const o=12,a=20,r=new TextDecoder;for(;n<=i;){const h=n+i>>>1,d=o+h*a,c=this.#a.getUint32(d);if(c<t)n=h+1;else{if(!(c>t)){if(!(h>0&&this.#a.getUint32(o+(h-1)*a)===t||h<s-1&&this.#a.getUint32(o+(h+1)*a)===t)){const e=this.#a.getUint32(o+h*a+12),t=this.#a.getUint32(o+h*a+16),s=[];for(let n=0;n<t;n++)s.push(this.#a.getUint32(e+4*n,!0));return s}let n=h;for(;n>0;){const e=o+(n-1)*a;if(this.#a.getUint32(e)!==t)break;n--}for(let i=n;i<s;i++){const s=o+i*a;if(this.#a.getUint32(s)!==t)break;const n=this.#a.getUint32(s+4),h=this.#a.getUint32(s+8),d=new Uint8Array(this.#o,h,n);if(r.decode(d)===e){const e=this.#a.getUint32(s+12),t=this.#a.getUint32(s+16),n=[];for(let s=0;s<t;s++)n.push(this.#a.getUint32(e+4*s,!0));return n}}return[]}i=h-1}}return[]}}const d="word_cache.bin",c="char_cache.bin";class g{#e;#t;#r;#h;#d=!1;#c;#g=!1;#f={word:0,char:0};constructor(s){if(!s.baseDir)throw new Error("SearchEngine requires 'baseDir' in config.");if(this.#c={wordSegmentTokenThreshold:1e5,charSegmentTokenThreshold:5e5,minWordTokenSave:0,minCharTokenSave:0,...s},(this.#c.minWordTokenSave||0)>=(this.#c.wordSegmentTokenThreshold||1e5))throw new Error("minWordTokenSave must be less than wordSegmentTokenThreshold");if((this.#c.minCharTokenSave||0)>=(this.#c.charSegmentTokenThreshold||5e5))throw new Error("minCharTokenSave must be less than charSegmentTokenThreshold");let n=null;if(this.#c.storage&&("object"==typeof this.#c.storage?n=this.#c.storage:"browser"===this.#c.storage?n=new e(this.#c.baseDir):"node"===this.#c.storage&&(n=new t(this.#c.baseDir))),!n){const s=typeof navigator<"u"&&navigator?.storage?.getDirectory instanceof Function,i=typeof process<"u"&&null!=process.versions&&null!=process.versions.node;s?n=new e(this.#c.baseDir):i&&(n=new t(this.#c.baseDir))}if(!n)throw new Error('Storage initialization failed. Please configure "storage" explicitly or ensure you are in a supported environment (Browser/Node).');this.#e=n,this.#t=new o(this.#e),this.#r=new a(this.#e),this.#h=new Map}async init(){if(this.#d)return;await this.#t.load();const e=[...this.#t.getSegments("word"),...this.#t.getSegments("char")];for(const t of e)this.#h.has(t.filename)||this.#h.set(t.filename,new h(t.filename,this.#e)),await this.#h.get(t.filename).loadIndex();this.#d=!0}startBatch(){this.#g=!0,this.#f={word:0,char:0}}async endBatch(){this.#g=!1,this.#f.word>0&&await this.#l("word",this.#f.word),this.#f.char>0&&await this.#l("char",this.#f.char),this.#f={word:0,char:0},await this.#t.save()}#w(e){try{if(typeof Intl<"u"&&"function"==typeof Intl.Segmenter&&"function"==typeof Array.from){const t=new Intl.Segmenter([],{granularity:"word"}).segment(e);if("object"==typeof t&&null!==t)return Array.from(t).filter(e=>e?.isWordLike).map(e=>e?.segment?.toLowerCase()||"")}}catch{}return e.toLowerCase().split(/[^a-z0-9\u4e00-\u9fa5]+/g).filter(e=>e.length>0)}#m(e){return this.#c.indexingTokenizer?this.#c.indexingTokenizer(e):this.#w(e.text)}#u(e){return this.#c.searchTokenizer?this.#c.searchTokenizer(e):this.#m(e)}async addDocument(e){return this.addDocuments([e])}async addDocumentIfMissing(e){return this.addDocumentsIfMissing([e])}async addDocumentsIfMissing(e){if(this.#d||await this.init(),0===e.length)return;const t=this.#t.getDeletedIds(),s=[],n=[],i=[];for(const o of e){if(t.has(o.id)||this.#t.isAdded(o.id))continue;const e=this.#m(o),a=[],r=[];for(const t of e)t.length>1?a.push(t):1===t.length&&r.push(t);a.length>0&&s.push({id:o.id,tokens:a}),r.length>0&&n.push({id:o.id,tokens:r}),i.push(o)}if(0===i.length)return;let o=0,a=0;if(s.length>0){await this.#r.appendBatch(d,s);for(const e of s)o+=e.tokens.length}if(n.length>0){await this.#r.appendBatch(c,n);for(const e of n)a+=e.tokens.length}for(const e of i)this.#t.addAddedId(e.id);this.#g?(this.#f.word+=o,this.#f.char+=a):(o>0&&await this.#l("word",o),a>0&&await this.#l("char",a),await this.#t.save())}async addDocuments(e){if(this.#d||await this.init(),0===e.length)return;const t=this.#t.getDeletedIds(),s=[],n=[];for(const i of e){if(t.has(i.id))throw new Error(`Document ID ${i.id} has been deleted and cannot be re-added.`);if(this.#t.isAdded(i.id))throw new Error(`Document ID ${i.id} already exists.`);const e=this.#m(i),o=[],a=[];for(const t of e)t.length>1?o.push(t):1===t.length&&a.push(t);o.length>0&&s.push({id:i.id,tokens:o}),a.length>0&&n.push({id:i.id,tokens:a})}let i=0,o=0;if(s.length>0){await this.#r.appendBatch(d,s);for(const e of s)i+=e.tokens.length}if(n.length>0){await this.#r.appendBatch(c,n);for(const e of n)o+=e.tokens.length}for(const t of e)this.#t.addAddedId(t.id);this.#g?(this.#f.word+=i,this.#f.char+=o):(i>0&&await this.#l("word",i),o>0&&await this.#l("char",o),await this.#t.save())}async#l(e,t){const s="word"===e?d:c,n=await this.#r.getCurrentSize(s),i="word"===e?this.#c.wordSegmentTokenThreshold||1e5:this.#c.charSegmentTokenThreshold||5e5,o="word"===e?this.#c.minWordTokenSave||0:this.#c.minCharTokenSave||0,a=this.#t.getLastSegmentInfo(e);let r,g,f,l;const w=()=>{const t=this.#t.getSegments(e).length+1;return`${e}_seg_${t}.bin`};if(a){const e=a.tokenCount;e>=i||e+t>=i?(r=w(),f=!0,g=a.end,l=t):(r=a.filename,f=!1,g=a.start,l=e+t)}else r=w(),f=!0,g=0,l=t;if(l<o)return void this.#t.updateSegment(e,r,g,n,l,f);const m=await this.#r.readRange(s,g,n);let u=this.#h.get(r);u||(u=new h(r,this.#e),this.#h.set(r,u)),await u.buildAndSave(m),this.#t.updateSegment(e,r,g,n,l,f)}async search(e,t){this.#d||await this.init();const s="string"==typeof e?{text:e}:e,n=this.#u(s),i=n.filter(e=>e.length>1),o=n.filter(e=>1===e.length),a=this.#t.getDeletedIds(),r=new Map,d=new Map,c=e=>{const t=this.#t.getSegments(e);for(const e of t){const t=e.filename;!this.#h.has(t)&&!d.has(t)&&d.set(t,new h(t,this.#e))}};c("word"),c("char"),await Promise.all(Array.from(d.entries()).map(([e,t])=>t.loadIndex().then(s=>{s&&this.#h.set(e,t)})));const g=async(e,t)=>{if(0===t.length)return;const s=this.#t.getSegments(e);for(const e of s){const s=e.filename,n=this.#h.get(s);if(n)for(const e of t){const t=n.search(e),s=1+.1*e.length;for(const n of t)if(!a.has(n))if(r.has(n)){const t=r.get(n);t.score+=s,t.tokens.add(e)}else r.set(n,{score:0,tokens:new Set([e])})}}};await g("word",i),await g("char",o);const f=[];return r.forEach((e,t)=>{f.push({id:t,score:e.score,tokens:Array.from(e.tokens)})}),f.sort((e,t)=>t.score-e.score),"number"==typeof t&&t>0?f.slice(0,t):f}async removeDocument(e){this.#d||await this.init(),this.#t.addDeletedId(e),this.#t.removeAddedId(e),await this.#t.save()}async clearAll(){await this.#e.clearAll(),this.#h.clear(),this.#t.reset(),this.#d=!1,this.#g=!1,this.#f={word:0,char:0}}async getStatus(){return this.#d||await this.init(),{wordSegments:this.#t.getSegments("word").length,charSegments:this.#t.getSegments("char").length,deleted:this.#t.getDeletedIds().size,wordCacheSize:await this.#r.getCurrentSize(d),charCacheSize:await this.#r.getCurrentSize(c),inBatch:this.#g}}async hasDocument(e){return this.#d||await this.init(),this.#t.hasDocument(e)}}export{g as SearchEngine,r as hash,r as murmur3_32};
+const e="search_meta.json",t="deleted_ids.bin",s="added_ids.bin";class n{#e;#t={wordSegments:[],charSegments:[]};#s=new Set;#n=new Set;constructor(e){this.#e=e}async load(){const n=await this.#e.read(e);if(n){const e=(new TextDecoder).decode(n);this.#t=JSON.parse(e)}else this.#t={wordSegments:[],charSegments:[]};const i=await this.#e.read(t);if(i){const e=new DataView(i);let t=0;const s=i.byteLength;for(;t<s&&!(t+4>s);){const n=e.getUint32(t,!0);this.#s.add(n),t+=4,t<s&&30===e.getUint8(t)&&(t+=1)}}const a=await this.#e.read(s);if(a){const e=new DataView(a);let t=0;const s=a.byteLength;for(;t<s&&!(t+4>s);){const n=e.getUint32(t,!0);this.#n.add(n),t+=4,t<s&&30===e.getUint8(t)&&(t+=1)}}}async save(){const n=JSON.stringify(this.#t);if(await this.#e.write(e,(new TextEncoder).encode(n).buffer),0===this.#s.size)await this.#e.remove(t);else{const e=4*this.#s.size+this.#s.size,s=new ArrayBuffer(e),n=new DataView(s);let i=0;for(const e of this.#s)n.setUint32(i,e,!0),i+=4,n.setUint8(i,30),i+=1;await this.#e.write(t,s)}if(0===this.#n.size)await this.#e.remove(s);else{const e=4*this.#n.size+this.#n.size,t=new ArrayBuffer(e),n=new DataView(t);let i=0;for(const e of this.#n)n.setUint32(i,e,!0),i+=4,n.setUint8(i,30),i+=1;await this.#e.write(s,t)}}getSegments(e){return"word"===e?this.#t.wordSegments:this.#t.charSegments}getDeletedIds(){return this.#s}addDeletedId(e){this.#s.add(e)}isDeleted(e){return this.#s.has(e)}addAddedId(e){this.#n.add(e)}removeAddedId(e){this.#n.delete(e)}isAdded(e){return this.#n.has(e)}getAddedIds(){return this.#n}hasDocument(e){return this.#n.has(e)||this.#s.has(e)}getLastSegmentInfo(e){const t=this.getSegments(e);return 0===t.length?null:t[t.length-1]}updateSegment(e,t,s,n,i,a){const o="word"===e?this.#t.wordSegments:this.#t.charSegments;if(a)o.push({filename:t,start:s,end:n,tokenCount:i});else{const e=o[o.length-1];e&&e.filename===t&&(e.end=n,e.tokenCount=i)}}reset(){this.#t={wordSegments:[],charSegments:[]},this.#s.clear(),this.#n.clear()}}class i{static SEPARATOR=30;#e;constructor(e){this.#e=e}async appendBatch(e,t){if(0===t.length)return await this.#e.getFileSize(e);const s=new TextEncoder;let n=0;for(const e of t){n+=8;for(const t of e.tokens){n+=2+Math.min(s.encode(t).byteLength,65535)}n+=1}const a=new Uint8Array(n);let o=0;for(const e of t){const t=[];for(const n of e.tokens){const e=s.encode(n),i=e.byteLength>65535?e.slice(0,65535):e;t.push(i)}const n=new DataView(a.buffer,o);n.setUint32(0,e.id,!0),n.setUint32(4,t.length,!0),o+=8;for(const e of t)new DataView(a.buffer,o).setUint16(0,e.byteLength,!0),o+=2,a.set(e,o),o+=e.byteLength;a[o++]=i.SEPARATOR}return await this.#e.append(e,a.buffer),await this.#e.getFileSize(e)}async readRange(e,t,s){const n=await this.#e.readRange(e,t,s);if(!n||0===n.byteLength)return[];const a=new DataView(n),o=new Uint8Array(n),r=new TextDecoder,h=[];let d=0;const c=n.byteLength;for(;d<c&&!(d+8>c);){const e=a.getUint32(d,!0);d+=4;const t=a.getUint32(d,!0);d+=4;const s=[];for(let e=0;e<t&&!(d+2>c);e++){const e=a.getUint16(d,!0);if(d+=2,d+e>c)break;const t=new Uint8Array(n,d,e);s.push(r.decode(t)),d+=e}d<c&&o[d]===i.SEPARATOR&&(d+=1),h.push({id:e,tokens:s})}return h}async getCurrentSize(e){return await this.#e.getFileSize(e)}}function a(e,t=305419896){const s=e.length,n=s>>2;let i=0;for(;i<n;){let s=255&e.charCodeAt(i)|(255&e.charCodeAt(++i))<<8|(255&e.charCodeAt(++i))<<16|(255&e.charCodeAt(++i))<<24;++i,s=3432918353*(65535&s)+((3432918353*(s>>>16)&65535)<<16)&4294967295,s=s<<15|s>>>17,s=461845907*(65535&s)+((461845907*(s>>>16)&65535)<<16)&4294967295,t=27492+(65535&(t=5*(65535&(t=(t^=s)<<13|t>>>19))+((5*(t>>>16)&65535)<<16)&4294967295))+(((t>>>16)+58964&65535)<<16)}let a=0;const o=3&s;return o>0&&(o>=3&&(a^=(255&e.charCodeAt(i+2))<<16),o>=2&&(a^=(255&e.charCodeAt(i+1))<<8),o>=1&&(a^=255&e.charCodeAt(i)),a=3432918353*(65535&a)+((3432918353*(a>>>16)&65535)<<16)&4294967295,a=a<<15|a>>>17,a=461845907*(65535&a)+((461845907*(a>>>16)&65535)<<16)&4294967295,t^=a),t^=s,t=2246822507*(65535&(t^=t>>>16))+((2246822507*(t>>>16)&65535)<<16)&4294967295,t=3266489909*(65535&(t^=t>>>13))+((3266489909*(t>>>16)&65535)<<16)&4294967295,(t^=t>>>16)>>>0}class o{#i;#e;#a=null;#o=null;static hash(e){return a(e)}constructor(e,t){this.#i=e,this.#e=t}async loadIndex(){return!!this.#a||(this.#a=await this.#e.read(this.#i),!!this.#a&&(this.#o=new DataView(this.#a),!0))}async buildAndSave(e){const t=new Map;for(const s of e){const e=new Map;for(const n of s.tokens)e.has(n)||(e.set(n,!0),t.has(n)||t.set(n,{hash:o.hash(n),postings:[]}),t.get(n).postings.push(s.id))}const s=Array.from(t.entries());s.sort(([e,{hash:t}],[s,{hash:n}])=>t!==n?t-n:e.localeCompare(s));let n=0,i=0;for(const[e,{postings:t}]of s)n+=t.length,i+=e.length+1;const a=20*s.length,r=12+a+4*n,h=new ArrayBuffer(r+i),d=new DataView(h);d.setUint32(0,1229866072),d.setUint32(4,s.length),d.setUint32(8,r);let c=12,g=12+a,f=r;for(const[e,{hash:t,postings:n}]of s){d.setUint32(c,t),d.setUint32(c+4,e.length),d.setUint32(c+8,f),d.setUint32(c+12,g),d.setUint32(c+16,n.length),c+=20;for(let e=0;e<n.length;e++)d.setUint32(g,n[e],!0),g+=4;const s=(new TextEncoder).encode(e);for(let e=0;e<s.length;e++)d.setUint8(f++,s[e]);d.setUint8(f++,0)}await this.#e.write(this.#i,h),this.#a=h,this.#o=d}search(e){if(!this.#o||!this.#a)return[];const t=o.hash(e),s=this.#o.getUint32(4);let n=0,i=s-1;const a=12,r=20,h=new TextDecoder;for(;n<=i;){const o=n+i>>>1,d=a+o*r,c=this.#o.getUint32(d);if(c<t)n=o+1;else{if(!(c>t)){if(!(o>0&&this.#o.getUint32(a+(o-1)*r)===t||o<s-1&&this.#o.getUint32(a+(o+1)*r)===t)){const e=this.#o.getUint32(a+o*r+12),t=this.#o.getUint32(a+o*r+16),s=[];for(let n=0;n<t;n++)s.push(this.#o.getUint32(e+4*n,!0));return s}let n=o;for(;n>0;){const e=a+(n-1)*r;if(this.#o.getUint32(e)!==t)break;n--}for(let i=n;i<s;i++){const s=a+i*r;if(this.#o.getUint32(s)!==t)break;const n=this.#o.getUint32(s+4),o=this.#o.getUint32(s+8),d=new Uint8Array(this.#a,o,n);if(h.decode(d)===e){const e=this.#o.getUint32(s+12),t=this.#o.getUint32(s+16),n=[];for(let s=0;s<t;s++)n.push(this.#o.getUint32(e+4*s,!0));return n}}return[]}i=o-1}}return[]}}const r=({text:e})=>{try{if(typeof Intl<"u"&&"function"==typeof Intl.Segmenter&&"function"==typeof Array.from){const t=new Intl.Segmenter([],{granularity:"word"}).segment(e);if("object"==typeof t&&null!==t)return Array.from(t).filter(e=>e?.isWordLike).map(e=>e?.segment?.toLowerCase()||"")}}catch{}return e.toLowerCase().split(/[^a-z0-9\u4e00-\u9fa5]+/g).filter(e=>e.length>0)},h="word_cache.bin",d="char_cache.bin";class c{#e;#t;#r;#h;#d=!1;#c;#g=!1;#f={word:0,char:0};constructor(e){if(this.#c={wordSegmentTokenThreshold:1e5,charSegmentTokenThreshold:5e5,minWordTokenSave:0,minCharTokenSave:0,indexingTokenizer:e.indexingTokenizer||r,...e},(this.#c.minWordTokenSave||0)>=(this.#c.wordSegmentTokenThreshold||1e5))throw new Error("minWordTokenSave must be less than wordSegmentTokenThreshold");if((this.#c.minCharTokenSave||0)>=(this.#c.charSegmentTokenThreshold||5e5))throw new Error("minCharTokenSave must be less than charSegmentTokenThreshold");this.#e=e.storage,this.#t=new n(this.#e),this.#r=new i(this.#e),this.#h=new Map}startBatch(){this.#g=!0,this.#f={word:0,char:0}}async endBatch(){this.#g=!1,this.#f.word>0&&await this.#l("word",this.#f.word),this.#f.char>0&&await this.#l("char",this.#f.char),this.#f={word:0,char:0},await this.#t.save()}async addDocument(e){return this.addDocuments([e])}async addDocumentIfMissing(e){return this.addDocumentsIfMissing([e])}async addDocumentsIfMissing(e){if(this.#d||await this.#w(),0===e.length)return;const t=this.#t.getDeletedIds(),s=[],n=[],i=[];for(const a of e){if(t.has(a.id)||this.#t.isAdded(a.id))continue;const e=this.#m(a),o=[],r=[];for(const t of e)t.length>1?o.push(t):1===t.length&&r.push(t);o.length>0&&s.push({id:a.id,tokens:o}),r.length>0&&n.push({id:a.id,tokens:r}),i.push(a)}if(0===i.length)return;let a=0,o=0;if(s.length>0){await this.#r.appendBatch(h,s);for(const e of s)a+=e.tokens.length}if(n.length>0){await this.#r.appendBatch(d,n);for(const e of n)o+=e.tokens.length}for(const e of i)this.#t.addAddedId(e.id);this.#g?(this.#f.word+=a,this.#f.char+=o):(a>0&&await this.#l("word",a),o>0&&await this.#l("char",o),await this.#t.save())}async addDocuments(e){if(this.#d||await this.#w(),0===e.length)return;const t=this.#t.getDeletedIds(),s=[],n=[];for(const i of e){if(t.has(i.id))throw new Error(`Document ID ${i.id} has been deleted and cannot be re-added.`);if(this.#t.isAdded(i.id))throw new Error(`Document ID ${i.id} already exists.`);const e=this.#m(i),a=[],o=[];for(const t of e)t.length>1?a.push(t):1===t.length&&o.push(t);a.length>0&&s.push({id:i.id,tokens:a}),o.length>0&&n.push({id:i.id,tokens:o})}let i=0,a=0;if(s.length>0){await this.#r.appendBatch(h,s);for(const e of s)i+=e.tokens.length}if(n.length>0){await this.#r.appendBatch(d,n);for(const e of n)a+=e.tokens.length}for(const t of e)this.#t.addAddedId(t.id);this.#g?(this.#f.word+=i,this.#f.char+=a):(i>0&&await this.#l("word",i),a>0&&await this.#l("char",a),await this.#t.save())}async search(e,t){this.#d||await this.#w();const s="string"==typeof e?{text:e}:e,n=this.#u(s),i=n.filter(e=>e.length>1),a=n.filter(e=>1===e.length),r=this.#t.getDeletedIds(),h=new Map,d=new Map,c=e=>{const t=this.#t.getSegments(e);for(const e of t){const t=e.filename;!this.#h.has(t)&&!d.has(t)&&d.set(t,new o(t,this.#e))}};c("word"),c("char"),await Promise.all(Array.from(d.entries()).map(([e,t])=>t.loadIndex().then(s=>{s&&this.#h.set(e,t)})));const g=async(e,t)=>{if(0===t.length)return;const s=this.#t.getSegments(e);for(const e of s){const s=e.filename,n=this.#h.get(s);if(n)for(const e of t){const t=n.search(e),s=1+.1*e.length;for(const n of t)if(!r.has(n))if(h.has(n)){const t=h.get(n);t.score+=s,t.tokens.add(e)}else h.set(n,{score:0,tokens:new Set([e])})}}};await g("word",i),await g("char",a);const f=[];return h.forEach((e,t)=>{f.push({id:t,score:e.score,tokens:Array.from(e.tokens)})}),f.sort((e,t)=>t.score-e.score),"number"==typeof t&&t>0?f.slice(0,t):f}async removeDocument(e){this.#d||await this.#w(),this.#t.addDeletedId(e),this.#t.removeAddedId(e),await this.#t.save()}async clearAll(){await this.#e.clearAll(),this.#h.clear(),this.#t.reset(),this.#d=!1,this.#g=!1,this.#f={word:0,char:0}}async getStatus(){return this.#d||await this.#w(),{wordSegments:this.#t.getSegments("word").length,charSegments:this.#t.getSegments("char").length,deleted:this.#t.getDeletedIds().size,wordCacheSize:await this.#r.getCurrentSize(h),charCacheSize:await this.#r.getCurrentSize(d),inBatch:this.#g}}async hasDocument(e){return this.#d||await this.#w(),this.#t.hasDocument(e)}async#w(){if(this.#d)return;await this.#t.load();const e=[...this.#t.getSegments("word"),...this.#t.getSegments("char")];for(const t of e)this.#h.has(t.filename)||this.#h.set(t.filename,new o(t.filename,this.#e)),await this.#h.get(t.filename).loadIndex();this.#d=!0}#m(e){return this.#c.indexingTokenizer(e)}#u(e){return this.#c.searchTokenizer?this.#c.searchTokenizer(e):this.#m(e)}async#l(e,t){const s="word"===e?h:d,n=await this.#r.getCurrentSize(s),i="word"===e?this.#c.wordSegmentTokenThreshold||1e5:this.#c.charSegmentTokenThreshold||5e5,a="word"===e?this.#c.minWordTokenSave||0:this.#c.minCharTokenSave||0,r=this.#t.getLastSegmentInfo(e);let c,g,f,l;const w=()=>{const t=this.#t.getSegments(e).length+1;return`${e}_seg_${t}.bin`};if(r){const e=r.tokenCount;e>=i||e+t>=i?(c=w(),f=!0,g=r.end,l=t):(c=r.filename,f=!1,g=r.start,l=e+t)}else c=w(),f=!0,g=0,l=t;if(l<a)return void this.#t.updateSegment(e,c,g,n,l,f);const m=await this.#r.readRange(s,g,n);let u=this.#h.get(c);u||(u=new o(c,this.#e),this.#h.set(c,u)),await u.buildAndSave(m),this.#t.updateSegment(e,c,g,n,l,f)}}export{c as SearchEngine,a as hash,a as murmur3_32};

package/lib/simple.cjs

@@ -1 +1 @@
-"use strict";var t=require('./core');exports.SimpleSearch=class{static#t=null;static#e={baseDir:"simple_search_data",wordSegmentTokenThreshold:1e5,minWordTokenSave:0};static configure(e){const n={...this.#e,...e};this.#t=new t.SearchEngine(n)}static#n(){return this.#t||(this.#t=new t.SearchEngine(this.#e)),this.#t}static async startBatch(){this.#n().startBatch()}static async endBatch(){return this.#n().endBatch()}static async addDocument(t){return this.#n().addDocument(t)}static async addDocumentIfMissing(t){return this.#n().addDocumentIfMissing(t)}static async addDocuments(t){return this.#n().addDocuments(t)}static async addDocumentsIfMissing(t){return this.#n().addDocumentsIfMissing(t)}static async search(t,e){return this.#n().search(t,e)}static async removeDocument(t){return this.#n().removeDocument(t)}static async clearAll(){return this.#n().clearAll()}static async getStatus(){return this.#n().getStatus()}static async hasDocument(t){return this.#n().hasDocument(t)}};
+"use strict";var t=require('./core'),e=require('./browser'),n=require('./node');const s=Object.freeze({wordSegmentTokenThreshold:1e5,minWordTokenSave:0}),c="simple-search";exports.SimpleSearch=class{static#t=null;static#e;static get config(){if(this.#e)return this.#e;const t={...s};return typeof navigator<"u"&&navigator?.storage?.getDirectory instanceof Function?t.storage=new e.BrowserStorage(c):t.storage=new n.NodeStorage(c),this.#e=t}static configure(e){this.#e={...this.config,...e},this.#t&&(this.#t=new t.SearchEngine(this.config))}static async startBatch(){this.#n().startBatch()}static async endBatch(){return this.#n().endBatch()}static async addDocument(t){return this.#n().addDocument(t)}static async addDocumentIfMissing(t){return this.#n().addDocumentIfMissing(t)}static async addDocuments(t){return this.#n().addDocuments(t)}static async addDocumentsIfMissing(t){return this.#n().addDocumentsIfMissing(t)}static async search(t,e){return this.#n().search(t,e)}static async removeDocument(t){return this.#n().removeDocument(t)}static async clearAll(){return this.#n().clearAll()}static async getStatus(){return this.#n().getStatus()}static async hasDocument(t){return this.#n().hasDocument(t)}static#n(){return this.#t||(this.#t=new t.SearchEngine(this.config)),this.#t}};

package/lib/simple.d.ts

@@ -1,5 +1,5 @@
 import * as ___type from './type';
-import { ISearchEngineConfig, IDocument, IDocumentBase } from './type';
+import { ISearchEngineOption, IDocument, IDocumentBase } from './type';
 
 /**
  * 快速使用封装
@@ -7,27 +7,21 @@ import { ISearchEngineConfig, IDocument, IDocumentBase } from './type';
  */
 declare class SimpleSearch {
     #private;
+    static get config(): ISearchEngineOption;
     /**
      * 配置并初始化单例
      */
-    static configure(config: Partial<ISearchEngineConfig>): void;
+    static configure(config: Partial<ISearchEngineOption>): void;
     static startBatch(): Promise<void>;
     static endBatch(): Promise<void>;
     static addDocument<T extends IDocument = IDocument>(doc: T): Promise<void>;
     static addDocumentIfMissing<T extends IDocument = IDocument>(doc: T): Promise<void>;
     static addDocuments<T extends IDocument = IDocument>(docs: T[]): Promise<void>;
     static addDocumentsIfMissing<T extends IDocument = IDocument>(docs: T[]): Promise<void>;
-    static search<T extends IDocumentBase = IDocumentBase>(query: T | string, limit?: number): Promise<___type.IResult[]>;
+    static search<T extends IDocumentBase = any>(query: T | string, limit?: number): Promise<___type.IResult[]>;
     static removeDocument(id: number): Promise<void>;
     static clearAll(): Promise<void>;
-    static getStatus(): Promise<{
-        wordSegments: number;
-        charSegments: number;
-        deleted: number;
-        wordCacheSize: number;
-        charCacheSize: number;
-        inBatch: boolean;
-    }>;
+    static getStatus(): Promise<___type.ISearchEngineStatus>;
     /**
      * 检查文档ID是否曾经添加过（包括已删除的）
      * @param id 文档ID

package/lib/simple.js

@@ -1 +1 @@
-import{SearchEngine as t}from'./core';class s{static#t=null;static#s={baseDir:"simple_search_data",wordSegmentTokenThreshold:1e5,minWordTokenSave:0};static configure(s){const n={...this.#s,...s};this.#t=new t(n)}static#n(){return this.#t||(this.#t=new t(this.#s)),this.#t}static async startBatch(){this.#n().startBatch()}static async endBatch(){return this.#n().endBatch()}static async addDocument(t){return this.#n().addDocument(t)}static async addDocumentIfMissing(t){return this.#n().addDocumentIfMissing(t)}static async addDocuments(t){return this.#n().addDocuments(t)}static async addDocumentsIfMissing(t){return this.#n().addDocumentsIfMissing(t)}static async search(t,s){return this.#n().search(t,s)}static async removeDocument(t){return this.#n().removeDocument(t)}static async clearAll(){return this.#n().clearAll()}static async getStatus(){return this.#n().getStatus()}static async hasDocument(t){return this.#n().hasDocument(t)}}export{s as SimpleSearch};
+import{SearchEngine as t}from'./core';import{BrowserStorage as n}from'./browser';import{NodeStorage as s}from'./node';const e=Object.freeze({wordSegmentTokenThreshold:1e5,minWordTokenSave:0}),c="simple-search";class a{static#t=null;static#n;static get config(){if(this.#n)return this.#n;const t={...e};return typeof navigator<"u"&&navigator?.storage?.getDirectory instanceof Function?t.storage=new n(c):t.storage=new s(c),this.#n=t}static configure(n){this.#n={...this.config,...n},this.#t&&(this.#t=new t(this.config))}static async startBatch(){this.#s().startBatch()}static async endBatch(){return this.#s().endBatch()}static async addDocument(t){return this.#s().addDocument(t)}static async addDocumentIfMissing(t){return this.#s().addDocumentIfMissing(t)}static async addDocuments(t){return this.#s().addDocuments(t)}static async addDocumentsIfMissing(t){return this.#s().addDocumentsIfMissing(t)}static async search(t,n){return this.#s().search(t,n)}static async removeDocument(t){return this.#s().removeDocument(t)}static async clearAll(){return this.#s().clearAll()}static async getStatus(){return this.#s().getStatus()}static async hasDocument(t){return this.#s().hasDocument(t)}static#s(){return this.#t||(this.#t=new t(this.config)),this.#t}}export{a as SimpleSearch};

package/lib/type.d.ts

@@ -66,12 +66,9 @@ interface IStorage {
  */
 type IndexType = 'word' | 'char';
 
-interface ISearchEngineConfig {
-    /**
-     * 数据存储的基础目录 (必填)
-     * 用于区分不同的搜索引擎实例
-     */
-    baseDir: string;
+type IndexingTokenizer = <T extends IDocument = IDocument>(doc: T) => string[];
+type SearchTokenizer = <T extends IDocumentBase = IDocumentBase>(doc: T) => string[];
+interface ISearchEngineOption {
     /**
      * 存储实现配置 (可选)
      * - 'browser': 强制使用 OPFS (BrowserStorage)
@@ -79,7 +76,7 @@ interface ISearchEngineConfig {
      * - IStorage: 传入自定义的存储实例
      * - undefined: 自动检测环境
      */
-    storage?: 'browser' | 'node' | IStorage;
+    storage: IStorage;
     /**
      * 索引时使用的分词器 (算法核心配置)
      * - 作用: 将文档文本转换为索引用的token序列
@@ -87,7 +84,7 @@ interface ISearchEngineConfig {
      * - 建议: 针对不同语言(中文/英文/日文等)使用专门的分词实现
      * - 影响: 直接决定索引的粒度和搜索的准确性
      */
-    indexingTokenizer?: <T extends IDocument = IDocument>(doc: T) => string[];
+    indexingTokenizer?: IndexingTokenizer;
     /**
      * 搜索时使用的分词器 (算法核心配置)
      * - 作用: 将查询文本转换为搜索用的token序列
@@ -95,7 +92,7 @@ interface ISearchEngineConfig {
      * - 建议: 与indexingTokenizer保持一致的分词策略以确保搜索准确性
      * - 影响: 直接决定搜索匹配的范围和结果的相关性
      */
-    searchTokenizer?: <T extends IDocumentBase = IDocumentBase>(doc: T) => string[];
+    searchTokenizer?: SearchTokenizer;
     /**
      * 词索引分段阈值 (Token数) - 分段算法配置
      * - 作用: 控制词索引文件的大小，超过阈值时创建新的索引段
@@ -130,6 +127,52 @@ interface ISearchEngineConfig {
     minCharTokenSave?: number;
 }
 
+interface ISearchEngineStatus {
+    wordSegments: number;
+    charSegments: number;
+    deleted: number;
+    wordCacheSize: number;
+    charCacheSize: number;
+    inBatch: boolean;
+}
+/**
+ * 核心搜索引擎
+ */
+interface ISearchEngine {
+    /**
+     * 开启批处理
+     * 批处理期间 addDocuments 只写入缓存，不触发索引段构建
+     */
+    startBatch(): void;
+    /**
+     * 结束批处理
+     * 触发索引构建检查并保存元数据
+     */
+    endBatch(): Promise<void>;
+    addDocument<T extends IDocument = IDocument>(doc: T): Promise<void>;
+    /**
+     * 添加单个文档，如果文档ID已存在则跳过
+     * 用于在批量添加中途出错后的恢复添加行为，也可直接用于单个文档添加
+     */
+    addDocumentIfMissing<T extends IDocument = IDocument>(doc: T): Promise<void>;
+    /**
+     * 添加多个文档，跳过已存在的文档ID
+     * 用于在批量添加中途出错后的恢复添加行为，也可直接用于批量添加
+     */
+    addDocumentsIfMissing<T extends IDocument = IDocument>(docs: T[]): Promise<void>;
+    addDocuments<T extends IDocument = IDocument>(docs: T[]): Promise<void>;
+    search<T extends IDocumentBase | string = any>(query: string, limit?: number): Promise<IResult[]>;
+    removeDocument(id: number): Promise<void>;
+    clearAll(): Promise<void>;
+    getStatus(): Promise<ISearchEngineStatus>;
+    /**
+     * 检查文档ID是否曾经添加过（包括已删除的）
+     * @param id 文档ID
+     * @returns 文档是否曾经添加过的布尔值
+     */
+    hasDocument(id: number): Promise<boolean>;
+}
+
 
 
-export type { IDocument, IDocumentBase, IIndexMeta, IResult, ISearchEngineConfig, ISegmentMeta, IStorage, ITokenizedDoc, IndexType };
+export type { IDocument, IDocumentBase, IIndexMeta, IResult, ISearchEngine, ISearchEngineOption, ISearchEngineStatus, ISegmentMeta, IStorage, ITokenizedDoc, IndexType, IndexingTokenizer, SearchTokenizer };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "gs-search",
-  "version": "0.1.4",
+  "version": "0.1.5",
   "type": "module",
   "main": "lib/index.cjs",
   "module": "lib/index.js",