feat: offline OCR (Tesseract) + embedding category classifier (@xenova/transformers)

Tesseract OCR (PHP, server-side): - Dockerfile: adds tesseract-ocr + tesseract-ocr-ita + libgd-dev (gd extension) - api/index.php: new tesseractReadExpiry() — decodes base64 image, pre-processes with GD (2× upscale, greyscale, auto-contrast, sharpen), runs tesseract CLI with ita+eng PSM-6, extracts date with multi-pattern regex (DD/MM/YYYY, MM/YYYY, ISO, named-month), returns YYYY-MM-DD + confidence - geminiReadExpiry() now: (1) tries Tesseract first; (2) falls back to Gemini Vision if OCR returns null or no date found; (3) passes source ('ocr'|'gemini') in response @xenova/transformers embedding classifier (browser-side): - index.html: ES-module bootstrap that lazy-loads 'Xenova/all-MiniLM-L6-v2' quantized (~23 MB, cached in browser) via window._getCategoryPipeline(); pre-warms on first scan page visit - assets/js/app.js: classifyCategoryByEmbedding(name) — embeds product name + 16 category anchor descriptions, cosine similarity, threshold 0.30; results cached in _embeddingCache Map - autoDetectCategory(): after keyword map misses, fires classifyCategoryByEmbedding async and updates select when resolved (respects manuallySet flag) - createQuickProduct(): if regex returned 'altro', silently patches category with embedding result via a background api call
2026-05-03 13:17:14 +00:00
parent c814d99d1f
commit a6c2fb93cf
4 changed files with 363 additions and 13 deletions
@@ -14,6 +14,39 @@
    <link rel="stylesheet" href="assets/css/style.css?v=20260421a">
    <!-- QuaggaJS for barcode scanning -->
    <script src="https://cdn.jsdelivr.net/npm/@ericblade/quagga2@1.8.4/dist/quagga.min.js"></script>
+    <!-- @xenova/transformers: ES-module bootstrap that exposes a lazy category-classifier as window._categoryPipelinePromise -->
+    <script type="module">
+        // Lazy-load the embedding pipeline only when first needed.
+        // Using a dynamic import so the ~2 MB WASM is not fetched on page load.
+        window._categoryPipelineReady = false;
+        window._categoryPipelinePromise = null;
+
+        window._getCategoryPipeline = async function() {
+            if (window._categoryPipelinePromise) return window._categoryPipelinePromise;
+            window._categoryPipelinePromise = (async () => {
+                try {
+                    const { pipeline, env } = await import(
+                        'https://cdn.jsdelivr.net/npm/@xenova/transformers@2/src/transformers.min.js'
+                    );
+                    // Keep WASM/model files in the browser cache; disable remote model check
+                    // to avoid CORS issues with the self-hosted instance.
+                    env.allowRemoteModels = true;
+                    env.useBrowserCache   = true;
+                    const pipe = await pipeline(
+                        'feature-extraction',
+                        'Xenova/all-MiniLM-L6-v2',
+                        { quantized: true }
+                    );
+                    window._categoryPipelineReady = true;
+                    return pipe;
+                } catch (e) {
+                    console.warn('[EverShelf] Embedding model unavailable, regex fallback only:', e);
+                    return null;
+                }
+            })();
+            return window._categoryPipelinePromise;
+        };
+    </script>
 </head>
 <body>