@liuhange/dsh-gov-data-inspector 3.1.0 → 3.1.2

package/lib/semanticRuleEngine.js

@@ -0,0 +1,160 @@
+import { ErrorDetailBuilder } from './errorDetailBuilder.js';
+function matchItemType(guide, itemTypeMatching) {
+    if (!itemTypeMatching)
+        return null;
+    const itemName = String(guide['事项名称'] ?? '');
+    const itemCode = String(guide['事项编码'] ?? '');
+    for (const [itemType, rule] of Object.entries(itemTypeMatching)) {
+        if (rule.keywords.some((kw) => itemName.includes(kw)))
+            return itemType;
+        if (itemCode && itemCode.startsWith(rule.codePrefix))
+            return itemType;
+    }
+    return null;
+}
+function extractTimeLimitDays(timeLimit) {
+    const str = String(timeLimit ?? '');
+    const match = str.match(/(\d+)\s*个?\s*工作日/);
+    if (!match || !match[1])
+        return null;
+    return parseInt(match[1], 10);
+}
+function detectTimeLimit(guide, guideId, itemType, kb, severityMapping) {
+    const timeLimitStr = String(guide['办理时限'] ?? '');
+    const days = extractTimeLimitDays(timeLimitStr);
+    if (days === null)
+        return [];
+    const stdEntry = kb.timeLimits.find((t) => t.itemType === itemType);
+    if (!stdEntry)
+        return [];
+    if (days > stdEntry.legalUpperLimit) {
+        return [
+            ErrorDetailBuilder.build({
+                guideId,
+                field: '办理时限',
+                errorType: 'semantic',
+                description: `办理时限"${timeLimitStr}"（${days}个工作日）超出法定上限${stdEntry.legalUpperLimit}个工作日。[数据源:${stdEntry.dataSource},条款:${stdEntry.standardClause}]`,
+                suggestion: `建议将办理时限调整为不超过${stdEntry.legalUpperLimit}个工作日，法定下限为${stdEntry.legalLowerLimit}个工作日`,
+                dataSource: stdEntry.dataSource,
+                standardClause: stdEntry.standardClause,
+            }, severityMapping),
+        ];
+    }
+    return [];
+}
+function detectMaterials(guide, guideId, itemType, kb, severityMapping) {
+    const materialStr = String(guide['申请材料'] ?? '');
+    if (!materialStr)
+        return [];
+    const materials = materialStr.split(/[、,，;；\n]/).map((s) => s.trim()).filter(Boolean);
+    const stdMaterials = kb.materials.filter((m) => m.itemType === itemType);
+    if (stdMaterials.length === 0)
+        return [];
+    const details = [];
+    for (const mat of materials) {
+        const matched = stdMaterials.some((std) => {
+            const stdName = std.standardName;
+            return mat === stdName || mat.includes(stdName) || stdName.includes(mat) || levenshteinSimilarity(mat, stdName) >= 0.95;
+        });
+        if (!matched) {
+            const closest = findClosestMaterial(mat, stdMaterials);
+            details.push(ErrorDetailBuilder.build({
+                guideId,
+                field: '申请材料',
+                errorType: 'semantic',
+                description: `申请材料"${mat}"与标准材料清单不匹配。标准材料清单中该事项类型应包含：${stdMaterials.map((m) => m.standardName).join('、')}。[数据源:${closest?.dataSource ?? 'unknown'},条款:${closest?.standardClause ?? 'unknown'}]`,
+                suggestion: closest
+                    ? `建议将"${mat}"修改为标准名称"${closest.standardName}"。${closest.basisClause}`
+                    : `建议核实"${mat}"是否为该事项标准材料，或补充至标准材料清单`,
+                dataSource: closest?.dataSource,
+                standardClause: closest?.standardClause,
+            }, severityMapping));
+        }
+    }
+    return details;
+}
+function levenshteinSimilarity(a, b) {
+    if (a.length === 0 && b.length === 0)
+        return 1;
+    const maxLen = Math.max(a.length, b.length);
+    if (maxLen === 0)
+        return 1;
+    const dist = levenshtein(a, b);
+    return 1 - dist / maxLen;
+}
+function levenshtein(a, b) {
+    const m = a.length;
+    const n = b.length;
+    const dp = Array.from({ length: m + 1 }, () => new Array(n + 1).fill(0));
+    for (let i = 0; i <= m; i++) {
+        const row = dp[i];
+        row[0] = i;
+    }
+    for (let j = 0; j <= n; j++) {
+        dp[0][j] = j;
+    }
+    for (let i = 1; i <= m; i++) {
+        for (let j = 1; j <= n; j++) {
+            const prevRow = dp[i - 1];
+            const currRow = dp[i];
+            if (a[i - 1] === b[j - 1])
+                currRow[j] = prevRow[j - 1];
+            else
+                currRow[j] = Math.min(prevRow[j], currRow[j - 1], prevRow[j - 1]) + 1;
+        }
+    }
+    return dp[m][n];
+}
+function findClosestMaterial(mat, stdMaterials) {
+    let best = null;
+    let bestSim = 0;
+    for (const std of stdMaterials) {
+        const sim = levenshteinSimilarity(mat, std.standardName);
+        if (sim > bestSim) {
+            bestSim = sim;
+            best = std;
+        }
+    }
+    return best;
+}
+function detectConditions(guide, guideId, itemType, kb, severityMapping) {
+    const conditionStr = String(guide['办理条件'] ?? '');
+    if (!conditionStr)
+        return [];
+    const stdConditions = kb.conditions.filter((c) => c.itemType === itemType);
+    if (stdConditions.length === 0)
+        return [];
+    const details = [];
+    for (const std of stdConditions) {
+        if (!conditionStr.includes(std.elementName) && !conditionStr.includes(std.standardValue)) {
+            details.push(ErrorDetailBuilder.build({
+                guideId,
+                field: '办理条件',
+                errorType: 'semantic',
+                description: `办理条件中缺失标准要素"${std.elementName}"，标准值应为"${std.standardValue}"。[数据源:${std.dataSource},条款:${std.standardClause}]`,
+                suggestion: `建议在办理条件中补充"${std.elementName}"要求，标准值为"${std.standardValue}"`,
+                dataSource: std.dataSource,
+                standardClause: std.standardClause,
+            }, severityMapping));
+        }
+    }
+    return details;
+}
+export const SemanticRuleEngine = {
+    detect(guide, guideId, kb, itemTypeMatching, severityMapping) {
+        if (!kb || !kb.timeLimits || !kb.materials || !kb.conditions) {
+            throw new Error('GOV_DATA_KB_MISSING: 标准知识库不可用，语义错误检测已中止');
+        }
+        const itemType = matchItemType(guide, itemTypeMatching);
+        if (!itemType)
+            return [];
+        const details = [];
+        details.push(...detectTimeLimit(guide, guideId, itemType, kb, severityMapping));
+        details.push(...detectMaterials(guide, guideId, itemType, kb, severityMapping));
+        details.push(...detectConditions(guide, guideId, itemType, kb, severityMapping));
+        return details;
+    },
+    matchItemType,
+    extractTimeLimitDays,
+};
+//# sourceMappingURL=semanticRuleEngine.js.map

package/lib/standardRuleSource.js

@@ -0,0 +1,62 @@
+const FAILED_RESULT = {
+    rules: [],
+    status: { status: 'failed', fetchedAt: '', recordCount: 0 },
+};
+export const StandardRuleSource = {
+    async fetch(credentials) {
+        const docPath = credentials?.docPath ?? process.env.STANDARD_DOC_PATH;
+        const endpoint = credentials?.endpoint ?? process.env.STANDARD_RULES_ENDPOINT;
+        if (!docPath && !endpoint) {
+            return { ...FAILED_RESULT, status: { ...FAILED_RESULT.status } };
+        }
+        try {
+            if (endpoint && endpoint.startsWith('https://')) {
+                const response = await fetch(endpoint, {
+                    method: 'GET',
+                    headers: { 'Accept': 'application/json' },
+                    signal: AbortSignal.timeout(10000),
+                });
+                if (!response.ok)
+                    return { ...FAILED_RESULT, status: { ...FAILED_RESULT.status } };
+                const raw = await response.json();
+                if (!Array.isArray(raw))
+                    return { ...FAILED_RESULT, status: { ...FAILED_RESULT.status } };
+                const rules = parseRules(raw);
+                return {
+                    rules,
+                    status: { status: 'success', fetchedAt: new Date().toISOString(), recordCount: rules.length },
+                };
+            }
+            if (docPath) {
+                const { readFileSync } = await import('node:fs');
+                const raw = JSON.parse(readFileSync(docPath, 'utf-8'));
+                if (!Array.isArray(raw))
+                    return { ...FAILED_RESULT, status: { ...FAILED_RESULT.status } };
+                const rules = parseRules(raw);
+                return {
+                    rules,
+                    status: { status: 'success', fetchedAt: new Date().toISOString(), recordCount: rules.length },
+                };
+            }
+            return { ...FAILED_RESULT, status: { ...FAILED_RESULT.status } };
+        }
+        catch {
+            return { ...FAILED_RESULT, status: { ...FAILED_RESULT.status } };
+        }
+    },
+};
+function parseRules(raw) {
+    return raw
+        .filter((item) => item !== null && typeof item === 'object')
+        .map((item) => ({
+        ruleId: String(item.ruleId ?? ''),
+        standardClause: String(item.standardClause ?? ''),
+        triggerFields: Array.isArray(item.triggerFields) ? item.triggerFields : [],
+        condition: String(item.condition ?? ''),
+        threshold: item.threshold,
+        suggestionTemplate: String(item.suggestionTemplate ?? ''),
+        triggerKeywords: Array.isArray(item.triggerKeywords) ? item.triggerKeywords : [],
+    }))
+        .filter((r) => r.ruleId && r.standardClause && r.triggerFields.length > 0);
+}
+//# sourceMappingURL=standardRuleSource.js.map

package/lib/types.js

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=types.js.map

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@liuhange/dsh-gov-data-inspector",
   "description": "Government data inspection plugin: service guide quality check and public data asset classification",
-  "version": "3.1.0",
+  "version": "3.1.2",
   "publishConfig": {
     "access": "public"
   },
@@ -31,10 +31,12 @@
     "test": "vitest run",
     "build": "tsc"
   },
+  "dependencies": {
+    "@liuhange/dsh-data-asset-shared": "^3.1.0"
+  },
   "peerDependencies": {
     "@deepseek-ai/cordis": "^4.0.0",
-    "@deepseek-ai/dsh-tools": "0.0.1-rc.1",
-    "@liuhange/dsh-data-asset-shared": "^3.0.0"
+    "@deepseek-ai/dsh-tools": "0.0.1-rc.1"
   },
   "dsh": {
     "bundle": {
@@ -57,4 +59,4 @@
     "办事指南"
   ],
   "homepage": "https://github.com/liuhange789/data-asset-inspector#readme"
-}
+}