@geotechcli/core 0.4.21 → 0.4.23

package/dist/vision/preprocess.js

@@ -0,0 +1,194 @@
+import { createRequire } from 'node:module';
+import { dirname, join } from 'node:path';
+import { pathToFileURL } from 'node:url';
+const require = createRequire(import.meta.url);
+let cachedPdfRendererModule = null;
+let cachedCanvasFactory = null;
+let cachedStandardFontDataUrl = null;
+let cachedPdfRendererUrl = null;
+let cachedCanvasModuleUrl = null;
+async function loadSharp() {
+    try {
+        const mod = await import('sharp');
+        return mod.default ?? mod;
+    }
+    catch {
+        return null;
+    }
+}
+async function loadPdfRendererModule() {
+    if (!cachedPdfRendererUrl) {
+        const pdfjsPackagePath = require.resolve('pdfjs-dist/package.json');
+        cachedPdfRendererUrl = pathToFileURL(join(dirname(pdfjsPackagePath), 'legacy', 'build', 'pdf.mjs')).href;
+    }
+    cachedPdfRendererModule ??= import(cachedPdfRendererUrl);
+    return cachedPdfRendererModule;
+}
+async function loadCanvasFactory() {
+    if (!cachedCanvasModuleUrl) {
+        const canvasSpecifier = ['@napi-rs', 'canvas'].join('/');
+        cachedCanvasModuleUrl = pathToFileURL(require.resolve(canvasSpecifier)).href;
+    }
+    cachedCanvasFactory ??= import(/* webpackIgnore: true */ cachedCanvasModuleUrl)
+        .then((mod) => (mod.createCanvas ?? mod.default?.createCanvas));
+    return cachedCanvasFactory;
+}
+function resolvePdfStandardFontDataUrl() {
+    if (cachedStandardFontDataUrl) {
+        return cachedStandardFontDataUrl;
+    }
+    const pdfjsPackagePath = require.resolve('pdfjs-dist/package.json');
+    cachedStandardFontDataUrl = `${join(dirname(pdfjsPackagePath), 'standard_fonts').replaceAll('\\', '/')}/`;
+    return cachedStandardFontDataUrl;
+}
+export async function encodeRawRasterToPng(pixels, details) {
+    const sharp = await loadSharp();
+    if (!sharp) {
+        throw new Error('sharp is unavailable for raw raster encoding.');
+    }
+    return sharp(Buffer.from(pixels), {
+        raw: {
+            width: details.width,
+            height: details.height,
+            channels: details.channels,
+        },
+    })
+        .png()
+        .toBuffer();
+}
+export async function preprocessVisionImageBuffer(buffer, mimeType, policy = 'ocr-optimized') {
+    if (policy === 'none') {
+        return {
+            buffer: Buffer.from(buffer),
+            mimeType,
+            transformed: false,
+            warnings: [],
+        };
+    }
+    const sharp = await loadSharp();
+    if (!sharp) {
+        return {
+            buffer: Buffer.from(buffer),
+            mimeType,
+            transformed: false,
+            warnings: ['Image preprocessing skipped because sharp is unavailable in this runtime.'],
+        };
+    }
+    try {
+        const output = await sharp(Buffer.from(buffer), { pages: 1 })
+            .resize({
+            width: 1600,
+            height: 1600,
+            fit: 'inside',
+            withoutEnlargement: true,
+        })
+            .rotate()
+            .grayscale()
+            .normalize()
+            .sharpen()
+            .png()
+            .toBuffer();
+        return {
+            buffer: output,
+            mimeType: 'image/png',
+            transformed: true,
+            warnings: [],
+        };
+    }
+    catch (error) {
+        return {
+            buffer: Buffer.from(buffer),
+            mimeType,
+            transformed: false,
+            warnings: [
+                `Image preprocessing skipped: ${error instanceof Error ? error.message : String(error)}`,
+            ],
+        };
+    }
+}
+export async function renderPdfPageToImageBuffer(buffer, pageNumber = 1, options) {
+    const renderer = await loadPdfRendererModule();
+    const createCanvas = await loadCanvasFactory();
+    const loadingTask = renderer.getDocument({
+        data: new Uint8Array(buffer),
+        disableWorker: true,
+        disableFontFace: true,
+        useSystemFonts: false,
+        isEvalSupported: false,
+        standardFontDataUrl: resolvePdfStandardFontDataUrl(),
+    });
+    let pdfDocument = null;
+    try {
+        pdfDocument = await loadingTask.promise;
+        if (pageNumber < 1 || pageNumber > pdfDocument.numPages) {
+            return null;
+        }
+        const page = await pdfDocument.getPage(pageNumber);
+        const viewport = page.getViewport({ scale: options?.scale ?? 1.5 });
+        const canvas = createCanvas(Math.max(1, Math.ceil(viewport.width)), Math.max(1, Math.ceil(viewport.height)));
+        const context = canvas.getContext('2d');
+        await page.render({
+            canvasContext: context,
+            viewport,
+        }).promise;
+        const renderedBuffer = canvas.toBuffer('image/png');
+        const preprocessed = await preprocessVisionImageBuffer(renderedBuffer, 'image/png', options?.preprocessPolicy ?? 'ocr-optimized');
+        return {
+            buffer: preprocessed.buffer,
+            mimeType: 'image/png',
+            width: canvas.width,
+            height: canvas.height,
+            warnings: preprocessed.warnings,
+        };
+    }
+    finally {
+        try {
+            if (pdfDocument) {
+                await pdfDocument.destroy();
+            }
+            else {
+                loadingTask.destroy();
+            }
+        }
+        catch {
+            // Ignore renderer cleanup errors.
+        }
+    }
+}
+export async function extractPdfPageTextFromBuffer(buffer, pageNumber = 1) {
+    const renderer = await loadPdfRendererModule();
+    const loadingTask = renderer.getDocument({
+        data: new Uint8Array(buffer),
+        disableWorker: true,
+        disableFontFace: true,
+        useSystemFonts: false,
+        isEvalSupported: false,
+        standardFontDataUrl: resolvePdfStandardFontDataUrl(),
+    });
+    let pdfDocument = null;
+    try {
+        pdfDocument = await loadingTask.promise;
+        if (pageNumber < 1 || pageNumber > pdfDocument.numPages) {
+            return null;
+        }
+        const page = await pdfDocument.getPage(pageNumber);
+        const textContent = await page.getTextContent();
+        const items = Array.isArray(textContent.items) ? textContent.items : [];
+        const text = items
+            .map((item) => (typeof item?.str === 'string' ? item.str : ''))
+            .join(' ')
+            .replace(/\s+/g, ' ')
+            .trim();
+        return text || null;
+    }
+    finally {
+        try {
+            await pdfDocument?.destroy();
+        }
+        catch {
+            // Ignore cleanup failures from pdfjs during fallback extraction.
+        }
+        loadingTask.destroy();
+    }
+}
+//# sourceMappingURL=preprocess.js.map

package/dist/vision/preprocess.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"preprocess.js","sourceRoot":"","sources":["../../src/vision/preprocess.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,aAAa,EAAE,MAAM,aAAa,CAAC;AAC5C,OAAO,EAAE,OAAO,EAAE,IAAI,EAAE,MAAM,WAAW,CAAC;AAC1C,OAAO,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AA2CzC,MAAM,OAAO,GAAG,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;AAE/C,IAAI,uBAAuB,GAAsC,IAAI,CAAC;AACtE,IAAI,mBAAmB,GAAqC,IAAI,CAAC;AACjE,IAAI,yBAAyB,GAAkB,IAAI,CAAC;AACpD,IAAI,oBAAoB,GAAkB,IAAI,CAAC;AAC/C,IAAI,qBAAqB,GAAkB,IAAI,CAAC;AAEhD,KAAK,UAAU,SAAS;IACtB,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,MAAM,MAAM,CAAC,OAAO,CAAC,CAAC;QAClC,OAAO,GAAG,CAAC,OAAO,IAAI,GAAG,CAAC;IAC5B,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,CAAC;IACd,CAAC;AACH,CAAC;AAED,KAAK,UAAU,qBAAqB;IAClC,IAAI,CAAC,oBAAoB,EAAE,CAAC;QAC1B,MAAM,gBAAgB,GAAG,OAAO,CAAC,OAAO,CAAC,yBAAyB,CAAC,CAAC;QACpE,oBAAoB,GAAG,aAAa,CAClC,IAAI,CAAC,OAAO,CAAC,gBAAgB,CAAC,EAAE,QAAQ,EAAE,OAAO,EAAE,SAAS,CAAC,CAC9D,CAAC,IAAI,CAAC;IACT,CAAC;IAED,uBAAuB,KAAK,MAAM,CAAC,oBAAoB,CAA+B,CAAC;IACvF,OAAO,uBAAuB,CAAC;AACjC,CAAC;AAED,KAAK,UAAU,iBAAiB;IAC9B,IAAI,CAAC,qBAAqB,EAAE,CAAC;QAC3B,MAAM,eAAe,GAAG,CAAC,UAAU,EAAE,QAAQ,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACzD,qBAAqB,GAAG,aAAa,CAAC,OAAO,CAAC,OAAO,CAAC,eAAe,CAAC,CAAC,CAAC,IAAI,CAAC;IAC/E,CAAC;IAED,mBAAmB,KAAK,MAAM,CAAC,yBAAyB,CAAC,qBAAqB,CAAC;SAC5E,IAAI,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,YAAY,IAAI,GAAG,CAAC,OAAO,EAAE,YAAY,CAAqB,CAAC,CAAC;IACtF,OAAO,mBAAmB,CAAC;AAC7B,CAAC;AAED,SAAS,6BAA6B;IACpC,IAAI,yBAAyB,EAAE,CAAC;QAC9B,OAAO,yBAAyB,CAAC;IACnC,CAAC;IAED,MAAM,gBAAgB,GAAG,OAAO,CAAC,OAAO,CAAC,yBAAyB,CAAC,CAAC;IACpE,yBAAyB,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,gBAAgB,CAAC,EAAE,gBAAgB,CAAC,CAAC,UAAU,CAAC,IAAI,EAAE,GAAG,CAAC,GAAG,CAAC;IAC1G,OAAO,yBAAyB,CAAC;AACnC,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,oBAAoB,CACxC,MAAkB,EAClB,OAA4D;IAE5D,MAAM,KAAK,GAAG,MAAM,SAAS,EAAE,CAAC;IAChC,IAAI,CAAC,KAAK,EAAE,CAAC;QACX,MAAM,IAAI,KAAK,CAAC,+CAA+C,CAAC,CAAC;IACnE,CAAC;IAED,OAAO,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE;QAChC,GAAG,EAAE;YACH,KAAK,EAAE,OAAO,CAAC,KAAK;YACpB,MAAM,EAAE,OAAO,CAAC,MAAM;YACtB,QAAQ,EAAE,OAAO,CAAC,QAAQ;SAC3B;KACF,CAAC;SACC,GAAG,EAAE;SACL,QAAQ,EAAE,CAAC;AAChB,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,2BAA2B,CAC/C,MAAkB,EAClB,QAAgB,EAChB,SAAsC,eAAe;IAErD,IAAI,MAAM,KAAK,MAAM,EAAE,CAAC;QACtB,OAAO;YACL,MAAM,EAAE,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC;YAC3B,QAAQ;YACR,WAAW,EAAE,KAAK;YAClB,QAAQ,EAAE,EAAE;SACb,CAAC;IACJ,CAAC;IAED,MAAM,KAAK,GAAG,MAAM,SAAS,EAAE,CAAC;IAChC,IAAI,CAAC,KAAK,EAAE,CAAC;QACX,OAAO;YACL,MAAM,EAAE,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC;YAC3B,QAAQ;YACR,WAAW,EAAE,KAAK;YAClB,QAAQ,EAAE,CAAC,2EAA2E,CAAC;SACxF,CAAC;IACJ,CAAC;IAED,IAAI,CAAC;QACH,MAAM,MAAM,GAAG,MAAM,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC;aAC1D,MAAM,CAAC;YACN,KAAK,EAAE,IAAI;YACX,MAAM,EAAE,IAAI;YACZ,GAAG,EAAE,QAAQ;YACb,kBAAkB,EAAE,IAAI;SACzB,CAAC;aACD,MAAM,EAAE;aACR,SAAS,EAAE;aACX,SAAS,EAAE;aACX,OAAO,EAAE;aACT,GAAG,EAAE;aACL,QAAQ,EAAE,CAAC;QAEd,OAAO;YACL,MAAM,EAAE,MAAM;YACd,QAAQ,EAAE,WAAW;YACrB,WAAW,EAAE,IAAI;YACjB,QAAQ,EAAE,EAAE;SACb,CAAC;IACJ,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,OAAO;YACL,MAAM,EAAE,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC;YAC3B,QAAQ;YACR,WAAW,EAAE,KAAK;YAClB,QAAQ,EAAE;gBACR,gCAAgC,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,EAAE;aACzF;SACF,CAAC;IACJ,CAAC;AACH,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,0BAA0B,CAC9C,MAAkB,EAClB,UAAU,GAAG,CAAC,EACd,OAGC;IAED,MAAM,QAAQ,GAAG,MAAM,qBAAqB,EAAE,CAAC;IAC/C,MAAM,YAAY,GAAG,MAAM,iBAAiB,EAAE,CAAC;IAC/C,MAAM,WAAW,GAAG,QAAQ,CAAC,WAAW,CAAC;QACvC,IAAI,EAAE,IAAI,UAAU,CAAC,MAAM,CAAC;QAC5B,aAAa,EAAE,IAAI;QACnB,eAAe,EAAE,IAAI;QACrB,cAAc,EAAE,KAAK;QACrB,eAAe,EAAE,KAAK;QACtB,mBAAmB,EAAE,6BAA6B,EAAE;KACrD,CAAC,CAAC;IAEH,IAAI,WAAW,GASJ,IAAI,CAAC;IAEhB,IAAI,CAAC;QACH,WAAW,GAAG,MAAM,WAAW,CAAC,OAAO,CAAC;QACxC,IAAI,UAAU,GAAG,CAAC,IAAI,UAAU,GAAG,WAAW,CAAC,QAAQ,EAAE,CAAC;YACxD,OAAO,IAAI,CAAC;QACd,CAAC;QAED,MAAM,IAAI,GAAG,MAAM,WAAW,CAAC,OAAO,CAAC,UAAU,CAAC,CAAC;QACnD,MAAM,QAAQ,GAAG,IAAI,CAAC,WAAW,CAAC,EAAE,KAAK,EAAE,OAAO,EAAE,KAAK,IAAI,GAAG,EAAE,CAAC,CAAC;QACpE,MAAM,MAAM,GAAG,YAAY,CACzB,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,EACtC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC,CACxC,CAAC;QACF,MAAM,OAAO,GAAG,MAAM,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC;QAExC,MAAM,IAAI,CAAC,MAAM,CAAC;YAChB,aAAa,EAAE,OAAO;YACtB,QAAQ;SACT,CAAC,CAAC,OAAO,CAAC;QAEX,MAAM,cAAc,GAAG,MAAM,CAAC,QAAQ,CAAC,WAAW,CAAC,CAAC;QACpD,MAAM,YAAY,GAAG,MAAM,2BAA2B,CACpD,cAAc,EACd,WAAW,EACX,OAAO,EAAE,gBAAgB,IAAI,eAAe,CAC7C,CAAC;QAEF,OAAO;YACL,MAAM,EAAE,YAAY,CAAC,MAAM;YAC3B,QAAQ,EAAE,WAAW;YACrB,KAAK,EAAE,MAAM,CAAC,KAAK;YACnB,MAAM,EAAE,MAAM,CAAC,MAAM;YACrB,QAAQ,EAAE,YAAY,CAAC,QAAQ;SAChC,CAAC;IACJ,CAAC;YAAS,CAAC;QACT,IAAI,CAAC;YACH,IAAI,WAAW,EAAE,CAAC;gBAChB,MAAM,WAAW,CAAC,OAAO,EAAE,CAAC;YAC9B,CAAC;iBAAM,CAAC;gBACN,WAAW,CAAC,OAAO,EAAE,CAAC;YACxB,CAAC;QACH,CAAC;QAAC,MAAM,CAAC;YACP,kCAAkC;QACpC,CAAC;IACH,CAAC;AACH,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,4BAA4B,CAChD,MAAkB,EAClB,UAAU,GAAG,CAAC;IAEd,MAAM,QAAQ,GAAG,MAAM,qBAAqB,EAAE,CAAC;IAC/C,MAAM,WAAW,GAAG,QAAQ,CAAC,WAAW,CAAC;QACvC,IAAI,EAAE,IAAI,UAAU,CAAC,MAAM,CAAC;QAC5B,aAAa,EAAE,IAAI;QACnB,eAAe,EAAE,IAAI;QACrB,cAAc,EAAE,KAAK;QACrB,eAAe,EAAE,KAAK;QACtB,mBAAmB,EAAE,6BAA6B,EAAE;KACrD,CAAC,CAAC;IAEH,IAAI,WAAW,GAQJ,IAAI,CAAC;IAEhB,IAAI,CAAC;QACH,WAAW,GAAG,MAAM,WAAW,CAAC,OAM/B,CAAC;QACF,IAAI,UAAU,GAAG,CAAC,IAAI,UAAU,GAAG,WAAW,CAAC,QAAQ,EAAE,CAAC;YACxD,OAAO,IAAI,CAAC;QACd,CAAC;QAED,MAAM,IAAI,GAAG,MAAM,WAAW,CAAC,OAAO,CAAC,UAAU,CAAC,CAAC;QACnD,MAAM,WAAW,GAAG,MAAM,IAAI,CAAC,cAAc,EAAE,CAAC;QAChD,MAAM,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,WAAW,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,CAAC;QACxE,MAAM,IAAI,GAAG,KAAK;aACf,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC,OAAO,IAAI,EAAE,GAAG,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;aAC9D,IAAI,CAAC,GAAG,CAAC;aACT,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC;aACpB,IAAI,EAAE,CAAC;QAEV,OAAO,IAAI,IAAI,IAAI,CAAC;IACtB,CAAC;YAAS,CAAC;QACT,IAAI,CAAC;YACH,MAAM,WAAW,EAAE,OAAO,EAAE,CAAC;QAC/B,CAAC;QAAC,MAAM,CAAC;YACP,iEAAiE;QACnE,CAAC;QACD,WAAW,CAAC,OAAO,EAAE,CAAC;IACxB,CAAC;AACH,CAAC"}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@geotechcli/core",
-  "version": "0.4.21",
+  "version": "0.4.23",
   "description": "Shared computation and LLM engine for geotechCLI",
   "type": "module",
   "main": "./dist/index.js",
@@ -39,9 +39,13 @@
     "lint": "tsc --noEmit"
   },
   "dependencies": {
+    "@napi-rs/canvas": "^0.1.99",
     "docx": "^9.0.0",
     "fflate": "^0.8.2",
+    "pdf-lib": "^1.17.1",
+    "pdfjs-dist": "^5.6.205",
     "pdfkit": "^0.15.0",
+    "sharp": "^0.34.5",
     "zod": "^3.24.0"
   },
   "devDependencies": {