修复货源查询 404 行被过滤且状态列置空：Python 404 的 ASIN 保留导出并标记「ASIN未找到」

线上现象：货源采集任务 26402（1212.xlsx）结果文件中存在 Python 端 404 Not Found 的行没有图，且状态列空白——这些行实际已完成采集但被后端过滤。根因链： 1. Python 404 行 title/url/图全空 → isMissingRequired 必填校验拦截 → markRowsFailed（status=FAILED，error 保留 "404 Not Found"）； 2. isExportableResultRow 对「其它错误需带 title/url/图」的判断把这类 404 行过滤掉，即使存在也被 isSilentOutputFailure 把状态列置空。修复： - 新增 isAsinNotFound 判定（404 / not found / 商品未找到等文案，大小写不敏感）； - resolveResultStatus 在图片判定后、rawStatus→失败映射前插入「ASIN未找到」状态；isExportableResultRow 对 404 行直接放行； - writeResultWorkbook 状态列 gate 排除 404 行（不再被置空）； - 新增 2 条单元测试（404/中文文案命中、普通技术失败不误判）。
2026-09-02 13:47:58 +08:00
parent 478233ba7a
commit 52c59755ab
2 changed files with 64 additions and 1 deletions
@@ -4623,7 +4623,8 @@ public class SimilarAsinTaskService {
                String isConform = resultRow == null ? "" : userFacingLlmCellValue(resultRow, resultRow.getIsConform());
                String reason = resultRow == null ? "" : userFacingLlmCellValue(resultRow, resultRow.getReason());
                String category = resultRow == null ? "" : userFacingLlmCellValue(resultRow, resultRow.getCategory());
-                String status = resultRow == null || isSilentOutputFailure(resultRow)
+                String status = resultRow == null
+                        || (isSilentOutputFailure(resultRow) && !isAsinNotFound(resultRow.getError()))
                        ? ""
                        : resolveResultStatus(resultRow, isStock, similarity, isConform, reason, category);
                row.createCell(col++).setCellValue(isStock);
@@ -4858,6 +4859,12 @@ public class SimilarAsinTaskService {
        if (hasText(row.getMainUrl()) || hasText(row.getPuzzleImg1()) || hasText(row.getPuzzleImg2())) {
            return "\u6210\u529f";
        }
+        // Python \u7aef\u641c\u4e0d\u5230\u5546\u54c1\uff08404\uff09\u7684\u884c\uff1a\u72b6\u6001\u5217\u660e\u786e\u6807\u8bb0\uff0c\u4eba\u5de5\u590d\u6838\u65f6\u4e00\u773c\u53ef\u8fa8\u3002
+        // \u653e\u5728\u56fe\u7247/\u72b6\u6001\u5224\u65ad\u4e4b\u524d\uff1a404 \u884c\u5fc5\u7136\u65e0\u56fe\u4e14 status \u5df2\u88ab markRowFailed \u7f6e\u4e3a FAILED\uff0c
+        // \u4e0d\u80fd\u843d\u56de"\u5931\u8d25"\uff08\u90a3\u662f LLM/\u751f\u6210\u94fe\u8def\u5931\u8d25\u8bed\u4e49\uff0c\u4e0e"ASIN \u641c\u4e0d\u5230"\u662f\u4e24\u56de\u4e8b\uff09\u3002
+        if (isAsinNotFound(row.getError())) {
+            return "ASIN\u672a\u627e\u5230";
+        }
        String rawStatus = row.getStatus() == null ? "" : row.getStatus().trim();
        if (!rawStatus.isBlank()) {
            return isFailedStatus(rawStatus) ? "\u5931\u8d25" : "\u6210\u529f";
@@ -4872,6 +4879,22 @@ public class SimilarAsinTaskService {
        return hasText(row.getConclusion()) ? "\u6210\u529f" : "\u5931\u8d25";
    }

+    /**
+     * \u5224\u5b9a\u662f\u5426"ASIN \u672a\u627e\u5230"\uff1aPython \u7aef\u6293\u53d6 404\uff08error="404 Not Found"\uff09\u6216\u7b49\u4ef7\u6587\u6848\u3002
+     * \u547d\u4e2d\u8be5\u5224\u5b9a\u7684\u884c\u4e0d\u4f1a\u88ab\u5f53\u6210\u751f\u6210\u5931\u8d25\u8ba1\u91cd\u8bd5\uff0c\u4e5f\u4e0d\u4f1a\u88ab\u5bfc\u51fa\u8fc7\u6ee4\uff0c\u72b6\u6001\u5217\u663e\u793a"ASIN\u672a\u627e\u5230"\u3002
+     */
+    static boolean isAsinNotFound(String error) {
+        String normalized = error == null ? "" : error.toLowerCase(Locale.ROOT).trim();
+        if (normalized.isBlank()) {
+            return false;
+        }
+        return normalized.contains("404")
+                || normalized.contains("not found")
+                || normalized.contains("asin \u672a\u627e\u5230")
+                || normalized.contains("asin\u4e0d\u5b58\u5728")
+                || normalized.contains("\u5546\u54c1\u672a\u627e\u5230");
+    }
+
    private static boolean isFailedStatus(String status) {
        String normalized = status == null ? "" : status.trim().toLowerCase(Locale.ROOT);
        return normalized.contains("fail")
@@ -4895,6 +4918,11 @@ public class SimilarAsinTaskService {
        if (error.isBlank()) {
            return false;
        }
+        // Python 端搜不到商品的 404 行：虽然 title/url/图全空，也必须导出，
+        // 并在状态列标记"ASIN未找到"，供人工复核。在"其它错误需带 title/url/图"之前判断。
+        if (isAsinNotFound(error)) {
+            return true;
+        }
        return !normalize(row.getTitle()).isBlank()
                || !normalize(row.getUrl()).isBlank()
                || row.hasImageUrl();
@@ -54,6 +54,41 @@ class SimilarAsinTaskServiceTest {
        assertEquals("\u6210\u529f", SimilarAsinTaskService.resolveResultStatus(withVisibleResultData, "", "80%", "", "", ""));
    }

+    @Test
+    void resultStatusMarksAsinNotFoundRows() {
+        // Python \u7aef 404 \u884c\uff1aerror="404 Not Found"\uff0cstatus \u88ab markRowFailed \u7f6e\u4e3a FAILED\uff0c
+        // \u4f46\u72b6\u6001\u5217\u5fc5\u987b\u663e\u793a"ASIN\u672a\u627e\u5230"\uff0c\u4e0d\u80fd\u843d\u56de\u666e\u901a"\u5931\u8d25"\u3002
+        SimilarAsinResultRowDto notFound = new SimilarAsinResultRowDto();
+        notFound.setError("404 Not Found");
+        notFound.setStatus("FAILED");
+        notFound.setReason("required field missing (asin/title/url), skip LLM");
+        assertEquals("ASIN\u672a\u627e\u5230", SimilarAsinTaskService.resolveResultStatus(notFound));
+
+        // \u5176\u5b83\u7b49\u4ef7\u6587\u6848\u540c\u6837\u547d\u4e2d
+        SimilarAsinResultRowDto chinese = new SimilarAsinResultRowDto();
+        chinese.setError("\u5546\u54c1\u672a\u627e\u5230");
+        assertEquals("ASIN\u672a\u627e\u5230", SimilarAsinTaskService.resolveResultStatus(chinese));
+
+        // \u666e\u901a\u6280\u672f\u6027\u5931\u8d25\u4e0d\u8bef\u5224\uff1aerror \u4e0d\u542b 404/not found
+        SimilarAsinResultRowDto llmFailure = new SimilarAsinResultRowDto();
+        llmFailure.setError("LLM \u8c03\u7528\u8d85\u65f6");
+        llmFailure.setStatus("FAILED");
+        assertEquals("\u5931\u8d25", SimilarAsinTaskService.resolveResultStatus(llmFailure));
+    }
+
+    @Test
+    void asinNotFoundDetectionMatchesPythonErrors() {
+        assertTrue(SimilarAsinTaskService.isAsinNotFound("404 Not Found"));
+        assertTrue(SimilarAsinTaskService.isAsinNotFound("404 not found"));
+        assertTrue(SimilarAsinTaskService.isAsinNotFound("404"));
+        assertTrue(SimilarAsinTaskService.isAsinNotFound("\u5546\u54c1\u672a\u627e\u5230"));
+        assertTrue(SimilarAsinTaskService.isAsinNotFound("ASIN \u672a\u627e\u5230"));
+        assertFalse(SimilarAsinTaskService.isAsinNotFound(null));
+        assertFalse(SimilarAsinTaskService.isAsinNotFound(""));
+        assertFalse(SimilarAsinTaskService.isAsinNotFound("LLM \u8c03\u7528\u8d85\u65f6"));
+        assertFalse(SimilarAsinTaskService.isAsinNotFound("\u4e0a\u4f20\u5931\u8d25"));
+    }
+
    @Test
    void fileBuildProgressIsTerminalOnlyAfterTaskAndStageComplete() {
        assertTrue(SimilarAsinTaskService.isTerminalFileBuildProgress("SUCCESS", 3, 3));