@orkg/scidquest 1.1.0 → 1.1.4

package/dist/utils/theme.d.ts

@@ -0,0 +1,6 @@
+import { PaletteMode, ThemeOptions } from '@mui/material/styles';
+declare const getDesignTokens: (mode: PaletteMode) => ThemeOptions;
+declare const theme: import("@mui/material").Theme;
+export declare const createAppTheme: (mode?: PaletteMode) => import("@mui/material").Theme;
+export { getDesignTokens };
+export default theme;

package/dist/utils/tokenEstimator.d.ts

@@ -0,0 +1,32 @@
+import type { ParentContext, SiblingContext, PreviousEntryContext, TokenLimits } from '../types/context';
+export declare class TokenEstimator {
+    private tokenLimits;
+    constructor(tokenLimits?: Partial<TokenLimits>);
+    estimateTokens(text: string): number;
+    estimateParentContextTokens(parentContext: ParentContext): number;
+    estimateSiblingContextTokens(siblingContext: SiblingContext[]): number;
+    estimatePreviousEntryContextTokens(previousEntryContext: PreviousEntryContext[]): number;
+    calculateTotalTokens(contexts: {
+        parentContext?: ParentContext;
+        siblingContext?: SiblingContext[];
+        previousEntryContext?: PreviousEntryContext[];
+    }): {
+        parentTokens: number;
+        siblingTokens: number;
+        previousEntryTokens: number;
+        totalTokens: number;
+    };
+    checkLimits(contexts: {
+        parentContext?: ParentContext;
+        siblingContext?: SiblingContext[];
+        previousEntryContext?: PreviousEntryContext[];
+    }): {
+        parentExceeded: boolean;
+        siblingExceeded: boolean;
+        previousEntryExceeded: boolean;
+        anyExceeded: boolean;
+    };
+    getTokenLimits(): TokenLimits;
+    updateTokenLimits(limits: Partial<TokenLimits>): void;
+}
+export declare const tokenEstimator: TokenEstimator;

package/dist/utils/tokenPdfMatcher.d.ts

@@ -0,0 +1,14 @@
+export interface TokenMatchResult {
+    found: boolean;
+    startIndex: number;
+    endIndex: number;
+    confidence: number;
+    matchedText: string;
+    skippedTokens: string[];
+}
+export declare function findTokenMatch(pdfText: string, searchText: string, options?: {
+    allowSkipReferences?: boolean;
+    minTokenMatch?: number;
+    maxSkippedTokens?: number;
+}): TokenMatchResult;
+export declare function stripReferences(text: string): string;

package/dist/utils/unicodeNormalizer.d.ts

@@ -0,0 +1,13 @@
+/**
+ * Unicode Normalization for PDF Text Matching
+ * Handles special characters that look the same but have different Unicode values
+ */
+/**
+ * Normalize Unicode characters to their ASCII equivalents
+ */
+export declare function normalizeUnicode(text: string): string;
+export declare function aggressiveUnicodeNormalize(text: string): string;
+export declare function debugUnicode(text: string): string;
+export declare function unicodeEquals(str1: string, str2: string): boolean;
+export declare function unicodeFindIndex(haystack: string, needle: string): number;
+export declare function unicodeIncludes(haystack: string, needle: string): boolean;

package/dist/utils/validationRules.d.ts

@@ -0,0 +1,51 @@
+export type ValidationRuleType = 'required' | 'format' | 'range' | 'custom' | 'minLength' | 'maxLength';
+export type FormatType = 'email' | 'url' | 'date' | 'phone' | 'number';
+export interface ValidationRule {
+    type: ValidationRuleType;
+    errorMessage?: string;
+}
+export interface RequiredRule extends ValidationRule {
+    type: 'required';
+}
+export interface FormatRule extends ValidationRule {
+    type: 'format';
+    format: FormatType;
+    pattern?: string;
+}
+export interface RangeRule extends ValidationRule {
+    type: 'range';
+    min?: number;
+    max?: number;
+}
+export interface MinLengthRule extends ValidationRule {
+    type: 'minLength';
+    value: number;
+}
+export interface MaxLengthRule extends ValidationRule {
+    type: 'maxLength';
+    value: number;
+}
+export interface CustomRule extends ValidationRule {
+    type: 'custom';
+    validator: (value: any) => boolean;
+    errorMessage: string;
+}
+export type AnyValidationRule = RequiredRule | FormatRule | RangeRule | MinLengthRule | MaxLengthRule | CustomRule;
+export interface ValidationResult {
+    isValid: boolean;
+    error?: string;
+}
+export declare function executeValidationRule(value: any, rule: AnyValidationRule): ValidationResult;
+export declare function validateValue(value: any, rules: AnyValidationRule[]): ValidationResult;
+export declare function parseValidationRules(questionConfig: any): AnyValidationRule[];
+export interface FieldValidationState {
+    isValid: boolean;
+    error?: string;
+    touched: boolean;
+}
+export interface ValidationState {
+    fields: Record<string, FieldValidationState>;
+}
+export declare function createValidationState(): ValidationState;
+export declare function updateFieldValidation(state: ValidationState, fieldId: string, result: ValidationResult, touched?: boolean): ValidationState;
+export declare function getFieldValidation(state: ValidationState, fieldId: string): FieldValidationState | undefined;

package/package.json

@@ -1,27 +1,42 @@
 {
   "name": "@orkg/scidquest",
-  "version": "1.1.0",
+  "version": "1.1.4",
   "type": "module",
   "license": "MIT",
   "description": "SciD-QuESt: From Scientific Documents to Knowledge - Questionnaire-Based Extraction and Structuring of Knowledge in the Open Research Knowledge Graph with LLMs and Human Validation",
   "main": "./dist/scidquest.es.js",
   "module": "./dist/scidquest.es.js",
-  "types": "./dist/index.d.ts",
+  "types": "./dist/lib/index.d.ts",
   "files": [
-    "dist"
+    "dist/scidquest.es.js",
+    "dist/scidquest.css",
+    "dist/contribute-standalone.css",
+    "dist/semanticChunker*.js",
+    "dist/assets/logo.png",
+    "dist/assets/pdf.worker.min.mjs",
+    "dist/lib",
+    "dist/lib/index.d.ts",
+    "dist/components",
+    "dist/context",
+    "dist/store",
+    "dist/services",
+    "dist/utils",
+    "dist/types",
+    "dist/hooks"
   ],
   "exports": {
     ".": {
-      "types": "./dist/index.d.ts",
+      "types": "./dist/lib/index.d.ts",
       "import": "./dist/scidquest.es.js",
       "default": "./dist/scidquest.es.js"
     },
     "./assets/logo.png": "./dist/assets/logo.png",
+    "./dist/scidquest.css": "./dist/scidquest.css",
     "./dist/contribute-standalone.css": "./dist/contribute-standalone.css"
   },
   "scripts": {
-    "dev": "vite",
-    "build": "tsc && vite build && node ./scripts/copy-logo.mjs && node ./scripts/copy-css.mjs",
+    "dev": "node ./scripts/copy-pdf-worker.mjs && vite",
+    "build": "vite build && tsc -p tsconfig.lib.json && node ./scripts/prune-dist-declarations.mjs && node ./scripts/copy-pdf-worker.mjs && node ./scripts/copy-logo.mjs && node ./scripts/copy-css.mjs",
     "preview": "vite preview",
     "prepublishOnly": "npm run build"
   },
@@ -30,7 +45,11 @@
     "@emotion/styled": "^11.14.0",
     "@mui/material": "^6.4.11",
     "react": "^18.3.1",
-    "react-dom": "^18.3.1"
+    "react-dom": "^18.3.1",
+    "react-redux": "^9.2.0",
+    "@reduxjs/toolkit": "^2.8.2",
+    "pdfjs-dist": "5.4.296",
+    "react-pdf": "^10.3.0"
   },
   "dependencies": {
     "@ai-sdk/groq": "^1.2.9",
@@ -43,7 +62,7 @@
     "@mui/material": "^6.4.11",
     "@reduxjs/toolkit": "^2.8.2",
     "ai": "^5.0.86",
-    "pdfjs-dist": "^5.4.296",
+    "pdfjs-dist": "5.4.296",
     "react": "^18.3.1",
     "react-dom": "^18.3.1",
     "react-pdf": "^10.3.0",
@@ -61,5 +80,8 @@
   },
   "publishConfig": {
     "access": "public"
+  },
+  "overrides": {
+    "pdfjs-dist": "5.4.296"
   }
 }

package/dist/semanticChunker-CEckDeUn.js

@@ -1,102 +0,0 @@
-var g = Object.defineProperty;
-var p = (a, e, n) => e in a ? g(a, e, { enumerable: !0, configurable: !0, writable: !0, value: n }) : a[e] = n;
-var u = (a, e, n) => p(a, typeof e != "symbol" ? e + "" : e, n);
-import { pipeline as k } from "@huggingface/transformers";
-class f {
-  constructor() {
-    u(this, "model", null);
-    u(this, "modelLoading", null);
-  }
-  async initialize() {
-    if (!this.model) {
-      if (this.modelLoading) return this.modelLoading;
-      console.log("[FrontendSemanticChunker] Initializing model..."), this.modelLoading = (async () => {
-        try {
-          this.model = await k(
-            "feature-extraction",
-            "Xenova/all-MiniLM-L6-v2",
-            { dtype: "q8" }
-          ), console.log("[FrontendSemanticChunker] Model loaded successfully");
-        } catch (e) {
-          throw console.error("[FrontendSemanticChunker] Model initialization failed:", e), this.modelLoading = null, e;
-        }
-      })(), await this.modelLoading;
-    }
-  }
-  splitIntoChunks(e) {
-    const n = [];
-    for (const s of e) {
-      if (!s.text || s.text.trim().length === 0) continue;
-      const o = s.text.split(new RegExp("(?<=[.!?])(?=\\s)"));
-      let t = "", r = 0;
-      for (const h of o) {
-        const m = h.trim();
-        m && (t && t.length + m.length + 1 > 2e3 ? (n.push({
-          text: t.trim(),
-          pageNumber: s.pageNumber,
-          startIndex: r,
-          endIndex: r + t.length,
-          tokenEstimate: Math.ceil(t.length / 4)
-        }), t = t.trim().split(/\s+/).slice(-Math.ceil(400 / 5)).join(" ") + " " + m, r += t.length - 400) : t += (t ? " " : "") + m);
-      }
-      t.trim() && (n.push({
-        text: t.trim(),
-        pageNumber: s.pageNumber,
-        startIndex: r,
-        endIndex: r + t.length,
-        tokenEstimate: Math.ceil(t.length / 4)
-      }), t = "", r = 0);
-    }
-    return console.log(
-      `[FrontendSemanticChunker] Total chunks: ${n.length}, Target size: 2000 chars`
-    ), n;
-  }
-  async embed(e) {
-    const n = await this.model(e, { pooling: "mean", normalize: !0 });
-    return Array.from(n.data);
-  }
-  cosineSimilarity(e, n) {
-    let c = 0, l = 0, s = 0;
-    for (let o = 0; o < e.length; o++)
-      c += e[o] * n[o], l += e[o] * e[o], s += n[o] * n[o];
-    return c / (Math.sqrt(l) * Math.sqrt(s));
-  }
-  async findRelevantChunks(e, n, c = 4e3) {
-    console.log(
-      `[FrontendSemanticChunker] Finding relevant chunks for question (max ${c} tokens)`
-    ), await this.initialize();
-    const l = this.splitIntoChunks(n);
-    console.log(
-      `[FrontendSemanticChunker] Created ${l.length} chunks from ${n.length} pages`
-    ), console.log("[FrontendSemanticChunker] Generating embeddings...");
-    const s = await Promise.all(
-      l.map((i) => this.embed(i.text))
-    ), o = await this.embed(e), t = l.map((i, d) => ({
-      ...i,
-      similarity: this.cosineSimilarity(o, s[d])
-    }));
-    t.sort((i, d) => d.similarity - i.similarity), console.log(
-      "[FrontendSemanticChunker] Top 5 similarity scores:",
-      t.slice(0, 5).map((i) => `page ${i.pageNumber}: ${i.similarity.toFixed(4)}`)
-    );
-    const r = [];
-    let h = 0;
-    for (const i of t)
-      h + i.tokenEstimate <= c && (r.push(i), h += i.tokenEstimate);
-    console.log(
-      `[FrontendSemanticChunker] Selected ${r.length} chunks (${h} tokens)`
-    ), r.sort((i, d) => i.pageNumber - d.pageNumber);
-    const m = Array.from(new Set(r.map((i) => i.pageNumber)));
-    return console.log(
-      "[FrontendSemanticChunker] Chunks span pages:",
-      m.join(", ")
-    ), r;
-  }
-  isInitialized() {
-    return this.model !== null;
-  }
-}
-const y = new f();
-export {
-  y as frontendSemanticChunker
-};