sku排序

tangwang
1 parent deccd68a
Showing 6 changed files with 200 additions and 172 deletions Show diff stats
config/config.yaml
frontend/index.html
frontend/static/css/style.css
frontend/static/js/app.js
search/searcher.py
tests/test_search_rerank_window.py
@@ -90,7 +90,34 @@ query_config:
  
   # 返回字段配置（_source includes）
   # null表示返回所有字段，[]表示不返回任何字段，列表表示只返回指定字段
-  source_fields: null
+  # 下列字段与 api/result_formatter.py（SpuResult 填充）及 search/searcher.py（SKU 排序/主图替换）一致
+  source_fields:
+    - spu_id
+    - handle
+    - title
+    - brief
+    - description
+    - vendor
+    - category_name
+    - category_name_text
+    - category_path
+    - category_id
+    - category_level
+    - category1_name
+    - category2_name
+    - category3_name
+    - tags
+    - min_price
+    - compare_at_price
+    - image_url
+    - sku_prices
+    - sku_weights
+    - sku_weight_units
+    - total_inventory
+    - option1_name
+    - option1_values
+    - specifications
+    - skus
  
   # KNN boost配置（向量召回的boost值）
   knn_boost: 0.25  # Lower boost for embedding recall
@@ -199,7 +199,7 @@
     </footer>
  
     <script src="/static/js/tenant_facets_config.js?v=1.4"></script>
-    <script src="/static/js/app.js?v=1.1"></script>
+    <script src="/static/js/app.js?v=1.0"></script>
     <script>
         // 自动补全功能（使用后端 /search/suggestions 接口）
         const SUGGEST_API = API_BASE_URL + '/search/suggestions';
@@ -375,108 +375,64 @@ body {
     display: flex;
     flex-wrap: wrap;
     align-items: center;
-    gap: 10px;
+    gap: 10px 14px;
     margin-top: 8px;
 }
  
-.product-debug-btn-api-result {
+.product-debug-inline-es-btn {
     font-family: inherit;
     font-size: 12px;
     padding: 4px 10px;
-    border: 1px solid #d35400;
+    border: 1px solid #ccc;
     border-radius: 4px;
-    background: #fff8f3;
-    color: #d35400;
+    background: #fafafa;
+    color: #333;
     cursor: pointer;
 }
  
-.product-debug-btn-api-result:hover {
-    background: #fdebd0;
-}
-
-.product-debug-link {
-    display: inline-block;
-    font-size: 12px;
-    color: #e67e22;
-    text-decoration: none;
-}
-
-.product-debug-link:hover {
-    text-decoration: underline;
-}
-
-.api-result-viewer-backdrop {
-    display: none;
-    position: fixed;
-    inset: 0;
-    z-index: 2000;
-    background: rgba(0, 0, 0, 0.45);
-    align-items: center;
-    justify-content: center;
-    padding: 24px;
-    box-sizing: border-box;
+.product-debug-inline-es-btn:hover {
+    background: #f0f0f0;
+    border-color: #bbb;
 }
  
-.api-result-viewer-panel {
-    background: #fff;
-    border-radius: 8px;
-    max-width: min(920px, 100%);
-    max-height: min(85vh, 100%);
-    display: flex;
-    flex-direction: column;
-    box-shadow: 0 8px 32px rgba(0, 0, 0, 0.2);
-    overflow: hidden;
+.product-debug--es-expanded {
+    max-height: min(70vh, 720px);
 }
  
-.api-result-viewer-header {
-    display: flex;
-    align-items: center;
-    justify-content: space-between;
-    gap: 12px;
-    padding: 12px 14px;
-    border-bottom: 1px solid #eee;
-    flex-shrink: 0;
-}
-
-.api-result-viewer-title {
-    font-size: 14px;
-    font-weight: 600;
-    color: #333;
+.product-es-doc-panel {
+    margin-top: 10px;
+    padding-top: 8px;
+    border-top: 1px dashed #e8e8e8;
 }
  
-.api-result-viewer-title code {
+.product-es-doc-panel-status {
     font-size: 12px;
-    font-weight: normal;
-    background: #f4f4f4;
-    padding: 1px 6px;
-    border-radius: 3px;
-}
-
-.api-result-viewer-close {
-    border: none;
-    background: transparent;
-    font-size: 22px;
-    line-height: 1;
     color: #888;
-    cursor: pointer;
-    padding: 0 4px;
 }
  
-.api-result-viewer-close:hover {
-    color: #333;
+.product-es-doc-pre {
+    margin: 6px 0 0;
+    padding: 10px;
+    background: #f5f5f5;
+    border-radius: 4px;
+    overflow: auto;
+    max-height: 50vh;
+    font-size: 11px;
+    line-height: 1.35;
+    white-space: pre-wrap;
+    word-break: break-word;
 }
  
-.api-result-viewer-pre {
-    margin: 0;
-    padding: 14px;
-    overflow: auto;
-    flex: 1;
+.product-debug-link {
+    display: inline-block;
+    margin-top: 0;
     font-size: 12px;
-    line-height: 1.45;
-    background: #fafafa;
-    border: none;
-    font-family: Menlo, Consolas, "Courier New", monospace;
-    white-space: pre;
+    color: #e67e22;
+    text-decoration: none;
+}
+
+.product-debug-link:hover {
+    text-decoration: underline;
 }
  
 .product-card:hover {
@@ -63,85 +63,88 @@ let state = {
     debug: true  // Always enable debug mode for test frontend
 };
  
-// 弹层：展示 /search/ 返回的 results[] 单条元素（非 ES 原始文档）
-function openApiResultViewer(item) {
-    let backdrop = document.getElementById('apiResultViewerBackdrop');
-    if (!backdrop) {
-        backdrop = document.createElement('div');
-        backdrop.id = 'apiResultViewerBackdrop';
-        backdrop.className = 'api-result-viewer-backdrop';
-        backdrop.innerHTML = `
-            <div class="api-result-viewer-panel" role="dialog" aria-modal="true" aria-labelledby="apiResultViewerTitle">
-                <div class="api-result-viewer-header">
-                    <span id="apiResultViewerTitle" class="api-result-viewer-title">搜索结果项（API <code>results[]</code>）</span>
-                    <button type="button" class="api-result-viewer-close" aria-label="关闭">&times;</button>
-                </div>
-                <pre class="api-result-viewer-pre"></pre>
-            </div>
-        `;
-        document.body.appendChild(backdrop);
-        backdrop.addEventListener('click', (e) => {
-            if (e.target === backdrop) {
-                closeApiResultViewer();
-            }
-        });
-        backdrop.querySelector('.api-result-viewer-close').addEventListener('click', closeApiResultViewer);
-        document.addEventListener('keydown', (e) => {
-            if (e.key === 'Escape') {
-                closeApiResultViewer();
-            }
-        });
-    }
-    const pre = backdrop.querySelector('.api-result-viewer-pre');
-    try {
-        pre.textContent = JSON.stringify(item, null, 2);
-    } catch (err) {
-        pre.textContent = String(item);
+// Initialize
+function initializeApp() {
+    // 初始化租户下拉框和分面面板
+    console.log('Initializing app...');
+    initTenantSelect();
+    setupProductGridEsDocToggle();
+    const searchInput = document.getElementById('searchInput');
+    if (searchInput) {
+        searchInput.focus();
     }
-    backdrop.style.display = 'flex';
 }
  
-function closeApiResultViewer() {
-    const backdrop = document.getElementById('apiResultViewerBackdrop');
-    if (backdrop) {
-        backdrop.style.display = 'none';
+/** Delegated handler: toggle inline ES raw response under each result card (survives innerHTML refresh on re-search). */
+function setupProductGridEsDocToggle() {
+    const grid = document.getElementById('productGrid');
+    if (!grid || grid.dataset.esDocToggleBound === '1') {
+        return;
     }
+    grid.dataset.esDocToggleBound = '1';
+    grid.addEventListener('click', onProductGridEsDocToggleClick);
 }
  
-function initProductGridResultViewer() {
-    const grid = document.getElementById('productGrid');
-    if (!grid || grid.dataset.apiResultViewerBound === '1') {
+async function onProductGridEsDocToggleClick(event) {
+    const btn = event.target.closest('[data-action="toggle-es-inline-doc"]');
+    if (!btn) {
+        return;
+    }
+    event.preventDefault();
+    const debugRoot = btn.closest('.product-debug');
+    if (!debugRoot) {
+        return;
+    }
+    const panel = debugRoot.querySelector('.product-es-doc-panel');
+    const pre = debugRoot.querySelector('.product-es-doc-pre');
+    const statusEl = debugRoot.querySelector('.product-es-doc-panel-status');
+    if (!panel || !pre || !statusEl) {
         return;
     }
-    grid.dataset.apiResultViewerBound = '1';
-    grid.addEventListener('click', (e) => {
-        const btn = e.target.closest('.product-debug-btn-api-result');
-        if (!btn) {
-            return;
-        }
-        e.preventDefault();
-        const idx = parseInt(btn.getAttribute('data-result-index'), 10);
-        if (Number.isNaN(idx)) {
-            return;
-        }
-        const results = state.lastSearchData && state.lastSearchData.results;
-        if (!results || idx < 0 || idx >= results.length) {
-            return;
-        }
-        openApiResultViewer(results[idx]);
-    });
-}
  
-// Initialize
-function initializeApp() {
-    // 初始化租户下拉框和分面面板
-    console.log('Initializing app...');
-    initTenantSelect();
-    initProductGridResultViewer();
-    const searchInput = document.getElementById('searchInput');
-    if (searchInput) {
-        searchInput.focus();
+    const spuId = btn.getAttribute('data-spu-id') || '';
+    const tenantId = getTenantId();
+    const url = `${API_BASE_URL}/search/es-doc/${encodeURIComponent(spuId)}?tenant_id=${encodeURIComponent(tenantId)}`;
+
+    if (debugRoot.dataset.esInlineOpen === '1') {
+        panel.setAttribute('hidden', '');
+        debugRoot.classList.remove('product-debug--es-expanded');
+        debugRoot.dataset.esInlineOpen = '0';
+        btn.textContent = '在结果中显示 ES 文档';
+        return;
     }
+
+    panel.removeAttribute('hidden');
+    debugRoot.classList.add('product-debug--es-expanded');
+    debugRoot.dataset.esInlineOpen = '1';
+    btn.textContent = '隐藏 ES 文档';
+
+    if (pre.textContent.length > 0) {
+        panel.scrollIntoView({ behavior: 'smooth', block: 'nearest' });
+        return;
+    }
+
+    statusEl.style.display = '';
+    statusEl.textContent = '加载中…';
+    pre.style.display = 'none';
+
+    try {
+        const response = await fetch(url);
+        if (!response.ok) {
+            const errText = await response.text();
+            throw new Error(`HTTP ${response.status}: ${errText.slice(0, 200)}`);
+        }
+        const data = await response.json();
+        pre.textContent = customStringify(data);
+        statusEl.style.display = 'none';
+        pre.style.display = 'block';
+    } catch (err) {
+        console.error('ES doc fetch failed', err);
+        statusEl.textContent = `加载失败: ${err.message || err}`;
+        pre.style.display = 'none';
+    }
+
+    panel.scrollIntoView({ behavior: 'smooth', block: 'nearest' });
 }
  
 // 在 DOM 加载完成后初始化
@@ -420,7 +423,7 @@ function displayResults(data) {
  
     const tenantId = getTenantId();
  
-    data.results.forEach((result, resultIndex) => {
+    data.results.forEach((result) => {
         const product = result;
         const title = product.title || product.name || 'N/A';
         const price = product.min_price || product.price || 'N/A';
@@ -472,13 +475,19 @@ function displayResults(data) {
                     <div class="product-debug-line">Fused score: ${fusedScore}</div>
                     ${titleLines}
                     <div class="product-debug-actions">
-                        <button type="button" class="product-debug-btn-api-result" data-result-index="${resultIndex}">
-                            查看 API 结果项
+                        <button type="button" class="product-debug-inline-es-btn"
+                            data-action="toggle-es-inline-doc"
+                            data-spu-id="${escapeAttr(String(spuId || ''))}">
+                            在结果中显示 ES 文档
                         </button>
                         <a class="product-debug-link" href="${rawUrl}" target="_blank" rel="noopener noreferrer">
                             查看 ES 原始文档
                         </a>
                     </div>
+                    <div class="product-es-doc-panel" hidden>
+                        <div class="product-es-doc-panel-status"></div>
+                        <pre class="product-es-doc-pre"></pre>
+                    </div>
                 </div>
             `;
         }
@@ -232,6 +232,29 @@ class Searcher:
             return ""
         return " ".join(str(value).strip().casefold().split())
  
+    @staticmethod
+    def _sku_option1_embedding_key(
+        sku: Dict[str, Any],
+        spu_option1_name: Optional[Any] = None,
+    ) -> Optional[str]:
+        """
+        Text sent to the embedding service for option1 must be "name:value"
+        (option name from SKU row or SPU-level option1_name).
+        """
+        value_raw = sku.get("option1_value")
+        if value_raw is None:
+            return None
+        value = str(value_raw).strip()
+        if not value:
+            return None
+        name = sku.get("option1_name")
+        if name is None or not str(name).strip():
+            name = spu_option1_name
+        name_str = str(name).strip() if name is not None and str(name).strip() else ""
+        if name_str:
+            value = f"{name_str}:{value}"
+        return value.casefold()
+
     def _build_sku_query_texts(self, parsed_query: ParsedQuery) -> List[str]:
         """Collect original and translated query texts for SKU option matching."""
         candidates: List[str] = []
@@ -271,8 +294,9 @@ class Searcher:
         self,
         skus: List[Dict[str, Any]],
         query_texts: List[str],
+        spu_option1_name: Optional[Any] = None,
     ) -> Optional[int]:
-        """Return the first SKU whose option1_value appears in query texts."""
+        """Return the first SKU whose option1_value (or name:value) appears in query texts."""
         if not skus or not query_texts:
             return None
  
@@ -282,6 +306,13 @@ class Searcher:
                 continue
             if any(option1_value in query_text for query_text in query_texts):
                 return index
+            embed_key = self._sku_option1_embedding_key(sku, spu_option1_name)
+            if embed_key and embed_key != option1_value:
+                composite_norm = self._normalize_sku_match_text(embed_key.replace(":", " "))
+                if any(composite_norm in query_text for query_text in query_texts):
+                    return index
+                if any(embed_key.casefold() in query_text for query_text in query_texts):
+                    return index
         return None
  
     def _encode_query_vector_for_sku_matching(
@@ -323,19 +354,17 @@ class Searcher:
         skus: List[Dict[str, Any]],
         option1_vectors: Dict[str, np.ndarray],
         query_vector: np.ndarray,
+        spu_option1_name: Optional[Any] = None,
     ) -> Tuple[Optional[int], Optional[float]]:
-        """Select the SKU whose option1_value is most similar to the query."""
+        """Select the SKU whose option1 embedding key (name:value) is most similar to the query."""
         best_index: Optional[int] = None
         best_score: Optional[float] = None
  
         for index, sku in enumerate(skus):
-            option1_value_raw = sku.get("option1_value")
-            if option1_value_raw is None:
+            embed_key = self._sku_option1_embedding_key(sku, spu_option1_name)
+            if not embed_key:
                 continue
-            option1_value = str(option1_value_raw).strip()
-            if not option1_value:
-                continue
-            option_vector = option1_vectors.get(option1_value)
+            option_vector = option1_vectors.get(embed_key)
             if option_vector is None:
                 continue
             score = float(np.inner(query_vector, option_vector))
@@ -385,7 +414,10 @@ class Searcher:
             if not isinstance(skus, list) or not skus:
                 continue
  
-            match_index = self._find_query_matching_sku_index(skus, query_texts)
+            spu_option1_name = source.get("option1_name")
+            match_index = self._find_query_matching_sku_index(
+                skus, query_texts, spu_option1_name=spu_option1_name
+            )
             if match_index is not None:
                 self._promote_matching_sku(source, match_index)
                 text_matched += 1
@@ -393,14 +425,11 @@ class Searcher:
  
             unmatched_hits.append(hit)
             for sku in skus:
-                option1_value_raw = sku.get("option1_value")
-                if option1_value_raw is None:
+                embed_key = self._sku_option1_embedding_key(sku, spu_option1_name)
+                if not embed_key or embed_key in seen_option1_values:
                     continue
-                option1_value = str(option1_value_raw).strip()
-                if not option1_value or option1_value in seen_option1_values:
-                    continue
-                seen_option1_values.add(option1_value)
-                option1_values_to_encode.append(option1_value)
+                seen_option1_values.add(embed_key)
+                option1_values_to_encode.append(embed_key)
  
         if not unmatched_hits or not option1_values_to_encode:
             return
@@ -437,7 +466,12 @@ class Searcher:
             skus = source.get("skus")
             if not isinstance(skus, list) or not skus:
                 continue
-            match_index, _ = self._select_sku_by_embedding(skus, option1_vectors, query_vector_array)
+            match_index, _ = self._select_sku_by_embedding(
+                skus,
+                option1_vectors,
+                query_vector_array,
+                spu_option1_name=source.get("option1_name"),
+            )
             if match_index is None:
                 continue
             self._promote_matching_sku(source, match_index)
@@ -370,6 +370,7 @@ def test_searcher_promotes_sku_when_option1_matches_translated_query(monkeypatch
             "title": {"en": f"product-{doc_id}"},
             "brief": {"en": f"brief-{doc_id}"},
             "vendor": {"en": f"vendor-{doc_id}"},
+            "option1_name": "Color",
             "image_url": "https://img/default.jpg",
             "skus": [
                 {"sku_id": "sku-red", "option1_value": "Red", "image_src": "https://img/red.jpg"},
@@ -406,8 +407,8 @@ def test_searcher_promotes_sku_by_embedding_when_query_has_no_direct_option_matc
     encoder = _FakeTextEncoder(
         {
             "linen summer dress": [0.8, 0.2],
-            "Red": [1.0, 0.0],
-            "Blue": [0.0, 1.0],
+            "color:Red": [1.0, 0.0],
+            "color:Blue": [0.0, 1.0],
         }
     )
  
@@ -431,6 +432,7 @@ def test_searcher_promotes_sku_by_embedding_when_query_has_no_direct_option_matc
             "title": {"en": f"product-{doc_id}"},
             "brief": {"en": f"brief-{doc_id}"},
             "vendor": {"en": f"vendor-{doc_id}"},
+            "option1_name": "Color",
             "image_url": "https://img/default.jpg",
             "skus": [
                 {"sku_id": "sku-red", "option1_value": "Red", "image_src": "https://img/red.jpg"},