@retrivora-ai/rag-engine 0.2.6 → 0.2.8

package/dist/server.js

@@ -217,6 +217,12 @@ var PostgreSQLProvider_exports = {};
 __export(PostgreSQLProvider_exports, {
   PostgreSQLProvider: () => PostgreSQLProvider
 });
+function stripInternalFilterKeys(filter) {
+  if (!filter) return {};
+  return Object.fromEntries(
+    Object.entries(filter).filter(([key]) => !key.startsWith("__"))
+  );
+}
 var import_pg, PostgreSQLProvider;
 var init_PostgreSQLProvider = __esm({
   "src/providers/vectordb/PostgreSQLProvider.ts"() {
@@ -309,8 +315,9 @@ var init_PostgreSQLProvider = __esm({
         let whereClause = namespace ? `WHERE namespace = $3` : "";
         const params = [vectorLiteral, topK];
         if (namespace) params.push(namespace);
-        if (filter && Object.keys(filter).length > 0) {
-          const filterConditions = Object.entries(filter).map(([key, val]) => {
+        const publicFilter = stripInternalFilterKeys(filter);
+        if (Object.keys(publicFilter).length > 0) {
+          const filterConditions = Object.entries(publicFilter).map(([key, val]) => {
             const paramIdx = params.length + 1;
             params.push(JSON.stringify(val));
             return `metadata->>'${key}' = $${paramIdx}`;
@@ -360,6 +367,12 @@ var MongoDBProvider_exports = {};
 __export(MongoDBProvider_exports, {
   MongoDBProvider: () => MongoDBProvider
 });
+function stripInternalFilterKeys2(filter) {
+  if (!filter) return {};
+  return Object.fromEntries(
+    Object.entries(filter).filter(([key]) => !key.startsWith("__"))
+  );
+}
 var import_mongodb, MongoDBProvider;
 var init_MongoDBProvider = __esm({
   "src/providers/vectordb/MongoDBProvider.ts"() {
@@ -411,6 +424,7 @@ var init_MongoDBProvider = __esm({
         await this.collection.bulkWrite(operations);
       }
       async query(vector, topK, namespace, filter) {
+        const publicFilter = stripInternalFilterKeys2(filter);
         const pipeline = [
           {
             $vectorSearch: __spreadValues({
@@ -419,7 +433,7 @@ var init_MongoDBProvider = __esm({
               queryVector: vector,
               numCandidates: Math.max(topK * 10, 100),
               limit: topK
-            }, filter || namespace ? { filter: __spreadValues(__spreadValues({}, filter || {}), namespace ? { namespace } : {}) } : {})
+            }, Object.keys(publicFilter).length > 0 || namespace ? { filter: __spreadValues(__spreadValues({}, publicFilter), namespace ? { namespace } : {}) } : {})
           },
           {
             $project: {
@@ -2752,6 +2766,40 @@ var EmbeddingStrategyResolver = class {
 };
 
 // src/core/Pipeline.ts
+function normalizeHintValue(value) {
+  return value.replace(/\s+/g, " ").trim();
+}
+function extractQueryFieldHints(question) {
+  if (!question.trim()) return [];
+  const hints = /* @__PURE__ */ new Map();
+  const addHint = (value, field) => {
+    const normalizedValue = normalizeHintValue(value);
+    if (!normalizedValue) return;
+    const normalizedField = field ? field.toLowerCase().replace(/[^a-z0-9]+/g, " ").trim() : void 0;
+    const key = `${normalizedField != null ? normalizedField : "*"}::${normalizedValue.toLowerCase()}`;
+    if (!hints.has(key)) {
+      hints.set(key, __spreadValues({
+        value: normalizedValue
+      }, normalizedField ? { field: normalizedField } : {}));
+    }
+  };
+  for (const match of question.matchAll(/["']([^"']{2,100})["']/g)) {
+    addHint(match[1]);
+  }
+  const fieldPattern = `([^\\n:=?.!,]{1,60}?)`;
+  const valuePattern = `([^\\n?.!,]{1,120}?)`;
+  const fieldValuePatterns = [
+    new RegExp(`\\b${fieldPattern}\\s*(?:=|:)\\s*["']?${valuePattern}["']?(?=[?.!,]|$)`, "gi"),
+    new RegExp(`\\b${fieldPattern}\\s+(?:is|are|was|were|equals?|equal to|named|called)\\s+["']?${valuePattern}["']?(?=[?.!,]|$)`, "gi"),
+    new RegExp(`\\bwith\\s+${fieldPattern}\\s+["']?${valuePattern}["']?(?=[?.!,]|$)`, "gi")
+  ];
+  for (const pattern of fieldValuePatterns) {
+    for (const match of question.matchAll(pattern)) {
+      addHint(match[2], match[1]);
+    }
+  }
+  return [...hints.values()];
+}
 var Pipeline = class {
   constructor(config) {
     this.initialised = false;
@@ -2839,7 +2887,11 @@ var Pipeline = class {
     const scoreThreshold = (_d = (_c = this.config.rag) == null ? void 0 : _c.scoreThreshold) != null ? _d : 0;
     try {
       const queryVector = await this.embeddingProvider.embed(question, { taskType: "query" });
-      const rawMatches = await this.vectorDB.query(queryVector, topK, ns, { __queryText: question });
+      const fieldHints = extractQueryFieldHints(question);
+      const rawMatches = await this.vectorDB.query(queryVector, topK, ns, {
+        __queryText: question,
+        __fieldHints: fieldHints
+      });
       const sources = rawMatches.filter((m) => m.score >= scoreThreshold);
       const context = sources.length ? sources.map((m, i) => `[Source ${i + 1}]
 ${m.content}`).join("\n\n---\n\n") : "No relevant context found.";
@@ -3692,10 +3744,26 @@ init_PostgreSQLProvider();
 // src/providers/vectordb/MultiTablePostgresProvider.ts
 var import_pg2 = require("pg");
 init_BaseVectorProvider();
+function normalizeFieldName(value) {
+  return value.toLowerCase().replace(/[^a-z0-9]+/g, "");
+}
+function sanitizeFilterHints(rawFilter) {
+  if (!Array.isArray(rawFilter == null ? void 0 : rawFilter.__fieldHints)) return [];
+  return rawFilter.__fieldHints.filter((hint) => typeof hint === "object" && hint !== null).map((hint) => __spreadValues({
+    value: typeof hint.value === "string" ? hint.value.trim().toLowerCase() : ""
+  }, typeof hint.field === "string" && hint.field.trim() ? { field: normalizeFieldName(hint.field) } : {})).filter((hint) => Boolean(hint.value));
+}
+function stripInternalFilterKeys3(filter) {
+  if (!filter) return {};
+  return Object.fromEntries(
+    Object.entries(filter).filter(([key]) => !key.startsWith("__"))
+  );
+}
 var MultiTablePostgresProvider = class extends BaseVectorProvider {
   constructor(config) {
     var _a, _b, _c;
     super(config);
+    this.tableSearchConfig = /* @__PURE__ */ new Map();
     const opts = config.options || {};
     if (!opts.connectionString) {
       throw new Error("[MultiTablePostgresProvider] options.connectionString is required");
@@ -3734,6 +3802,7 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
       if (this.tables.length === 0) {
         console.warn('[MultiTablePostgresProvider] No tables with "embedding" columns found in the database.');
       } else {
+        this.tableSearchConfig = await this.loadTableSearchConfig(client, this.tables);
         console.log(
           `[MultiTablePostgresProvider] Connected. Searching across ${this.tables.length} table(s): ${this.tables.join(", ")}`
         );
@@ -3763,7 +3832,7 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
    * Query all configured tables and merge results, sorted by cosine similarity score.
    */
   async query(vector, topK, _namespace, _filter) {
-    var _a, _b, _c;
+    var _a, _b;
     if (!this.pool) {
       throw new Error("[MultiTablePostgresProvider] Provider not initialized. Call initialize() first.");
     }
@@ -3771,22 +3840,51 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
     const allResults = [];
     console.log(`[MultiTablePostgresProvider] --- Starting Multi-Table Search ---`);
     const queryText = _filter == null ? void 0 : _filter.__queryText;
+    const fieldHints = sanitizeFilterHints(_filter);
+    const explicitFilter = stripInternalFilterKeys3(_filter);
     const queryPromises = this.tables.map(async (table) => {
+      var _a2, _b2, _c;
       try {
         let sqlQuery = "";
         let params = [];
+        const tableConfig = this.tableSearchConfig.get(table);
+        const availableFields = (_a2 = tableConfig == null ? void 0 : tableConfig.availableFields) != null ? _a2 : [];
+        const searchableFields = (_b2 = tableConfig == null ? void 0 : tableConfig.searchableFields) != null ? _b2 : [];
+        const normalizedFieldMap = (_c = tableConfig == null ? void 0 : tableConfig.normalizedFieldMap) != null ? _c : /* @__PURE__ */ new Map();
+        const whereConditions = [];
         if (queryText) {
+          const genericValueHints = fieldHints.filter((hint) => !hint.field).map((hint) => hint.value);
+          params = genericValueHints.length > 0 ? [vectorLiteral, queryText, genericValueHints] : [vectorLiteral, queryText];
+          const genericValueParamIndex = genericValueHints.length > 0 ? 3 : -1;
+          const exactValueScoreExpr = genericValueHints.length > 0 && searchableFields.length > 0 ? `GREATEST(${searchableFields.map(
+            (field) => `CASE WHEN LOWER(COALESCE(to_jsonb(t)->>'${field}', '')) = ANY($${genericValueParamIndex}::text[]) THEN 1 ELSE 0 END`
+          ).join(", ")})` : "0";
+          const fieldSpecificChecks = fieldHints.filter((hint) => hint.field).map((hint) => {
+            const column = hint.field ? normalizedFieldMap.get(hint.field) : void 0;
+            if (!column) return null;
+            params.push(hint.value);
+            const paramIndex = params.length;
+            return `CASE WHEN LOWER(COALESCE(to_jsonb(t)->>'${column}', '')) = $${paramIndex}::text THEN 1 ELSE 0 END`;
+          }).filter((expr) => Boolean(expr));
+          const exactFieldScoreExpr = fieldSpecificChecks.length > 0 ? `GREATEST(${fieldSpecificChecks.join(", ")})` : "0";
           sqlQuery = `
             SELECT *, 
              (1 - (embedding <=> $1::vector)) AS vector_score,
              COALESCE(ts_rank(to_tsvector('english', t.*::text), NULLIF(REPLACE(plainto_tsquery('english', $2)::text, '&', '|'), '')::tsquery), 0) AS keyword_score,
-             ((1 - (embedding <=> $1::vector)) + (COALESCE(ts_rank(to_tsvector('english', t.*::text), NULLIF(REPLACE(plainto_tsquery('english', $2)::text, '&', '|'), '')::tsquery), 0) * 2.0)) AS hybrid_score
+             ${exactValueScoreExpr} AS exact_value_score,
+             ${exactFieldScoreExpr} AS exact_field_score,
+             (
+               (1 - (embedding <=> $1::vector)) +
+               (COALESCE(ts_rank(to_tsvector('english', t.*::text), NULLIF(REPLACE(plainto_tsquery('english', $2)::text, '&', '|'), '')::tsquery), 0) * 2.0) +
+               (${exactValueScoreExpr} * 3.0) +
+               (${exactFieldScoreExpr} * 5.0)
+             ) AS hybrid_score
             FROM "${table}" t
             ORDER BY hybrid_score DESC
             LIMIT 50
           `;
-          params = [vectorLiteral, queryText];
         } else {
+          params = [vectorLiteral];
           sqlQuery = `
             SELECT *, 
              (1 - (embedding <=> $1::vector)) AS hybrid_score
@@ -3794,7 +3892,16 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
             ORDER BY hybrid_score DESC
             LIMIT 50
           `;
-          params = [vectorLiteral];
+        }
+        for (const [key, value] of Object.entries(explicitFilter)) {
+          const column = availableFields.includes(key) ? key : normalizedFieldMap.get(normalizeFieldName(key));
+          if (!column) continue;
+          params.push(String(value));
+          whereConditions.push(`COALESCE(to_jsonb(t)->>'${column}', '') = $${params.length}::text`);
+        }
+        if (whereConditions.length > 0) {
+          sqlQuery = sqlQuery.replace(`FROM "${table}" t`, `FROM "${table}" t
+            WHERE ${whereConditions.join(" AND ")}`);
         }
         const result = await this.pool.query(sqlQuery, params);
         if (result.rowCount && result.rowCount > 0) {
@@ -3804,10 +3911,12 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
         }
         const tableResults = [];
         for (const row of result.rows) {
-          const _a2 = row, { hybrid_score, id } = _a2, rest = __objRest(_a2, ["hybrid_score", "id"]);
+          const _d = row, { hybrid_score, id } = _d, rest = __objRest(_d, ["hybrid_score", "id"]);
           delete rest.embedding;
           delete rest.vector_score;
           delete rest.keyword_score;
+          delete rest.exact_value_score;
+          delete rest.exact_field_score;
           const content = `[TYPE: ${table.replace(/s$/, "").toUpperCase()}]
 ` + Object.entries(rest).filter(([k, v]) => v !== null && typeof v !== "object" && k !== "id").map(([k, v]) => `${k}: ${v}`).join("\n");
           tableResults.push({
@@ -3830,23 +3939,12 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
     for (const tableResults of resultsArray) {
       allResults.push(...tableResults);
     }
-    const resultsByTable = {};
-    for (const res of allResults) {
-      const table = (_a = res.metadata) == null ? void 0 : _a.source_table;
-      if (!resultsByTable[table]) resultsByTable[table] = [];
-      resultsByTable[table].push(res);
-    }
-    const balancedResults = [];
-    const tables = Object.keys(resultsByTable);
-    for (const table of tables) {
-      balancedResults.push(...resultsByTable[table].slice(0, 3));
-    }
-    const finalSorted = balancedResults.sort((a, b) => b.score - a.score);
+    const finalSorted = allResults.sort((a, b) => b.score - a.score);
     if (finalSorted.length > 0) {
       console.log(`[MultiTablePostgresProvider] --- Search Complete ---`);
-      console.log(`[MultiTablePostgresProvider] Final top match from "${(_c = (_b = finalSorted[0].metadata) == null ? void 0 : _b.source_table) != null ? _c : "unknown"}" with score ${finalSorted[0].score.toFixed(4)}`);
+      console.log(`[MultiTablePostgresProvider] Final top match from "${(_b = (_a = finalSorted[0].metadata) == null ? void 0 : _a.source_table) != null ? _b : "unknown"}" with score ${finalSorted[0].score.toFixed(4)}`);
     }
-    return finalSorted.slice(0, Math.max(topK, 15));
+    return finalSorted.slice(0, topK);
   }
   async delete(_id, _namespace) {
     console.warn("[MultiTablePostgresProvider] delete() is a no-op for multi-table mode.");
@@ -3867,6 +3965,51 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
       await this.pool.end();
     }
   }
+  async loadTableSearchConfig(client, tables) {
+    var _a, _b;
+    if (tables.length === 0) return /* @__PURE__ */ new Map();
+    const configuredSearchFields = this.parseConfiguredSearchFields();
+    const result = await client.query(`
+      SELECT table_name, column_name, data_type
+      FROM information_schema.columns
+      WHERE table_schema = 'public'
+        AND table_name = ANY($1::text[])
+      ORDER BY ordinal_position
+    `, [tables]);
+    const rowsByTable = /* @__PURE__ */ new Map();
+    for (const row of result.rows) {
+      const rows = (_a = rowsByTable.get(row.table_name)) != null ? _a : [];
+      rows.push({ column_name: row.column_name, data_type: row.data_type });
+      rowsByTable.set(row.table_name, rows);
+    }
+    const configByTable = /* @__PURE__ */ new Map();
+    for (const table of tables) {
+      const columns = (_b = rowsByTable.get(table)) != null ? _b : [];
+      const availableFields = columns.filter(({ column_name }) => column_name !== "embedding").map(({ column_name }) => column_name);
+      const inferredFields = columns.filter(
+        ({ column_name, data_type }) => column_name !== "embedding" && !["ARRAY", "json", "jsonb", "bytea", "tsvector", "USER-DEFINED"].includes(data_type)
+      ).map(({ column_name }) => column_name);
+      const searchableFields = configuredSearchFields.length > 0 ? configuredSearchFields.filter((field) => columns.some((column) => column.column_name === field)) : inferredFields;
+      const normalizedFieldMap = /* @__PURE__ */ new Map();
+      for (const field of availableFields) {
+        normalizedFieldMap.set(normalizeFieldName(field), field);
+      }
+      configByTable.set(table, {
+        availableFields,
+        searchableFields,
+        normalizedFieldMap
+      });
+    }
+    return configByTable;
+  }
+  parseConfiguredSearchFields() {
+    var _a;
+    const raw = (_a = this.config.options) == null ? void 0 : _a.searchFields;
+    if (Array.isArray(raw)) {
+      return raw.filter((field) => typeof field === "string" && field.trim().length > 0);
+    }
+    return [];
+  }
 };
 
 // src/server.ts

package/dist/server.mjs

@@ -34,17 +34,17 @@ import {
   createIngestHandler,
   createUploadHandler,
   getRagConfig
-} from "./chunk-7K4KXB6G.mjs";
+} from "./chunk-CYVPACA7.mjs";
 import "./chunk-EDLTMSNY.mjs";
 import {
   PineconeProvider
 } from "./chunk-BMHJTWSU.mjs";
 import {
   PostgreSQLProvider
-} from "./chunk-IUTAZ7QR.mjs";
+} from "./chunk-6GSARSCP.mjs";
 import {
   MongoDBProvider
-} from "./chunk-5HXNKSCR.mjs";
+} from "./chunk-IFPISZ2S.mjs";
 import {
   MilvusProvider
 } from "./chunk-VEJNRS4B.mjs";
@@ -335,10 +335,26 @@ function createFromPreset(presetName) {
 
 // src/providers/vectordb/MultiTablePostgresProvider.ts
 import { Pool } from "pg";
+function normalizeFieldName(value) {
+  return value.toLowerCase().replace(/[^a-z0-9]+/g, "");
+}
+function sanitizeFilterHints(rawFilter) {
+  if (!Array.isArray(rawFilter == null ? void 0 : rawFilter.__fieldHints)) return [];
+  return rawFilter.__fieldHints.filter((hint) => typeof hint === "object" && hint !== null).map((hint) => __spreadValues({
+    value: typeof hint.value === "string" ? hint.value.trim().toLowerCase() : ""
+  }, typeof hint.field === "string" && hint.field.trim() ? { field: normalizeFieldName(hint.field) } : {})).filter((hint) => Boolean(hint.value));
+}
+function stripInternalFilterKeys(filter) {
+  if (!filter) return {};
+  return Object.fromEntries(
+    Object.entries(filter).filter(([key]) => !key.startsWith("__"))
+  );
+}
 var MultiTablePostgresProvider = class extends BaseVectorProvider {
   constructor(config) {
     var _a, _b, _c;
     super(config);
+    this.tableSearchConfig = /* @__PURE__ */ new Map();
     const opts = config.options || {};
     if (!opts.connectionString) {
       throw new Error("[MultiTablePostgresProvider] options.connectionString is required");
@@ -377,6 +393,7 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
       if (this.tables.length === 0) {
         console.warn('[MultiTablePostgresProvider] No tables with "embedding" columns found in the database.');
       } else {
+        this.tableSearchConfig = await this.loadTableSearchConfig(client, this.tables);
         console.log(
           `[MultiTablePostgresProvider] Connected. Searching across ${this.tables.length} table(s): ${this.tables.join(", ")}`
         );
@@ -406,7 +423,7 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
    * Query all configured tables and merge results, sorted by cosine similarity score.
    */
   async query(vector, topK, _namespace, _filter) {
-    var _a, _b, _c;
+    var _a, _b;
     if (!this.pool) {
       throw new Error("[MultiTablePostgresProvider] Provider not initialized. Call initialize() first.");
     }
@@ -414,22 +431,51 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
     const allResults = [];
     console.log(`[MultiTablePostgresProvider] --- Starting Multi-Table Search ---`);
     const queryText = _filter == null ? void 0 : _filter.__queryText;
+    const fieldHints = sanitizeFilterHints(_filter);
+    const explicitFilter = stripInternalFilterKeys(_filter);
     const queryPromises = this.tables.map(async (table) => {
+      var _a2, _b2, _c;
       try {
         let sqlQuery = "";
         let params = [];
+        const tableConfig = this.tableSearchConfig.get(table);
+        const availableFields = (_a2 = tableConfig == null ? void 0 : tableConfig.availableFields) != null ? _a2 : [];
+        const searchableFields = (_b2 = tableConfig == null ? void 0 : tableConfig.searchableFields) != null ? _b2 : [];
+        const normalizedFieldMap = (_c = tableConfig == null ? void 0 : tableConfig.normalizedFieldMap) != null ? _c : /* @__PURE__ */ new Map();
+        const whereConditions = [];
         if (queryText) {
+          const genericValueHints = fieldHints.filter((hint) => !hint.field).map((hint) => hint.value);
+          params = genericValueHints.length > 0 ? [vectorLiteral, queryText, genericValueHints] : [vectorLiteral, queryText];
+          const genericValueParamIndex = genericValueHints.length > 0 ? 3 : -1;
+          const exactValueScoreExpr = genericValueHints.length > 0 && searchableFields.length > 0 ? `GREATEST(${searchableFields.map(
+            (field) => `CASE WHEN LOWER(COALESCE(to_jsonb(t)->>'${field}', '')) = ANY($${genericValueParamIndex}::text[]) THEN 1 ELSE 0 END`
+          ).join(", ")})` : "0";
+          const fieldSpecificChecks = fieldHints.filter((hint) => hint.field).map((hint) => {
+            const column = hint.field ? normalizedFieldMap.get(hint.field) : void 0;
+            if (!column) return null;
+            params.push(hint.value);
+            const paramIndex = params.length;
+            return `CASE WHEN LOWER(COALESCE(to_jsonb(t)->>'${column}', '')) = $${paramIndex}::text THEN 1 ELSE 0 END`;
+          }).filter((expr) => Boolean(expr));
+          const exactFieldScoreExpr = fieldSpecificChecks.length > 0 ? `GREATEST(${fieldSpecificChecks.join(", ")})` : "0";
           sqlQuery = `
             SELECT *, 
              (1 - (embedding <=> $1::vector)) AS vector_score,
              COALESCE(ts_rank(to_tsvector('english', t.*::text), NULLIF(REPLACE(plainto_tsquery('english', $2)::text, '&', '|'), '')::tsquery), 0) AS keyword_score,
-             ((1 - (embedding <=> $1::vector)) + (COALESCE(ts_rank(to_tsvector('english', t.*::text), NULLIF(REPLACE(plainto_tsquery('english', $2)::text, '&', '|'), '')::tsquery), 0) * 2.0)) AS hybrid_score
+             ${exactValueScoreExpr} AS exact_value_score,
+             ${exactFieldScoreExpr} AS exact_field_score,
+             (
+               (1 - (embedding <=> $1::vector)) +
+               (COALESCE(ts_rank(to_tsvector('english', t.*::text), NULLIF(REPLACE(plainto_tsquery('english', $2)::text, '&', '|'), '')::tsquery), 0) * 2.0) +
+               (${exactValueScoreExpr} * 3.0) +
+               (${exactFieldScoreExpr} * 5.0)
+             ) AS hybrid_score
             FROM "${table}" t
             ORDER BY hybrid_score DESC
             LIMIT 50
           `;
-          params = [vectorLiteral, queryText];
         } else {
+          params = [vectorLiteral];
           sqlQuery = `
             SELECT *, 
              (1 - (embedding <=> $1::vector)) AS hybrid_score
@@ -437,7 +483,16 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
             ORDER BY hybrid_score DESC
             LIMIT 50
           `;
-          params = [vectorLiteral];
+        }
+        for (const [key, value] of Object.entries(explicitFilter)) {
+          const column = availableFields.includes(key) ? key : normalizedFieldMap.get(normalizeFieldName(key));
+          if (!column) continue;
+          params.push(String(value));
+          whereConditions.push(`COALESCE(to_jsonb(t)->>'${column}', '') = $${params.length}::text`);
+        }
+        if (whereConditions.length > 0) {
+          sqlQuery = sqlQuery.replace(`FROM "${table}" t`, `FROM "${table}" t
+            WHERE ${whereConditions.join(" AND ")}`);
         }
         const result = await this.pool.query(sqlQuery, params);
         if (result.rowCount && result.rowCount > 0) {
@@ -447,10 +502,12 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
         }
         const tableResults = [];
         for (const row of result.rows) {
-          const _a2 = row, { hybrid_score, id } = _a2, rest = __objRest(_a2, ["hybrid_score", "id"]);
+          const _d = row, { hybrid_score, id } = _d, rest = __objRest(_d, ["hybrid_score", "id"]);
           delete rest.embedding;
           delete rest.vector_score;
           delete rest.keyword_score;
+          delete rest.exact_value_score;
+          delete rest.exact_field_score;
           const content = `[TYPE: ${table.replace(/s$/, "").toUpperCase()}]
 ` + Object.entries(rest).filter(([k, v]) => v !== null && typeof v !== "object" && k !== "id").map(([k, v]) => `${k}: ${v}`).join("\n");
           tableResults.push({
@@ -473,23 +530,12 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
     for (const tableResults of resultsArray) {
       allResults.push(...tableResults);
     }
-    const resultsByTable = {};
-    for (const res of allResults) {
-      const table = (_a = res.metadata) == null ? void 0 : _a.source_table;
-      if (!resultsByTable[table]) resultsByTable[table] = [];
-      resultsByTable[table].push(res);
-    }
-    const balancedResults = [];
-    const tables = Object.keys(resultsByTable);
-    for (const table of tables) {
-      balancedResults.push(...resultsByTable[table].slice(0, 3));
-    }
-    const finalSorted = balancedResults.sort((a, b) => b.score - a.score);
+    const finalSorted = allResults.sort((a, b) => b.score - a.score);
     if (finalSorted.length > 0) {
       console.log(`[MultiTablePostgresProvider] --- Search Complete ---`);
-      console.log(`[MultiTablePostgresProvider] Final top match from "${(_c = (_b = finalSorted[0].metadata) == null ? void 0 : _b.source_table) != null ? _c : "unknown"}" with score ${finalSorted[0].score.toFixed(4)}`);
+      console.log(`[MultiTablePostgresProvider] Final top match from "${(_b = (_a = finalSorted[0].metadata) == null ? void 0 : _a.source_table) != null ? _b : "unknown"}" with score ${finalSorted[0].score.toFixed(4)}`);
     }
-    return finalSorted.slice(0, Math.max(topK, 15));
+    return finalSorted.slice(0, topK);
   }
   async delete(_id, _namespace) {
     console.warn("[MultiTablePostgresProvider] delete() is a no-op for multi-table mode.");
@@ -510,6 +556,51 @@ var MultiTablePostgresProvider = class extends BaseVectorProvider {
       await this.pool.end();
     }
   }
+  async loadTableSearchConfig(client, tables) {
+    var _a, _b;
+    if (tables.length === 0) return /* @__PURE__ */ new Map();
+    const configuredSearchFields = this.parseConfiguredSearchFields();
+    const result = await client.query(`
+      SELECT table_name, column_name, data_type
+      FROM information_schema.columns
+      WHERE table_schema = 'public'
+        AND table_name = ANY($1::text[])
+      ORDER BY ordinal_position
+    `, [tables]);
+    const rowsByTable = /* @__PURE__ */ new Map();
+    for (const row of result.rows) {
+      const rows = (_a = rowsByTable.get(row.table_name)) != null ? _a : [];
+      rows.push({ column_name: row.column_name, data_type: row.data_type });
+      rowsByTable.set(row.table_name, rows);
+    }
+    const configByTable = /* @__PURE__ */ new Map();
+    for (const table of tables) {
+      const columns = (_b = rowsByTable.get(table)) != null ? _b : [];
+      const availableFields = columns.filter(({ column_name }) => column_name !== "embedding").map(({ column_name }) => column_name);
+      const inferredFields = columns.filter(
+        ({ column_name, data_type }) => column_name !== "embedding" && !["ARRAY", "json", "jsonb", "bytea", "tsvector", "USER-DEFINED"].includes(data_type)
+      ).map(({ column_name }) => column_name);
+      const searchableFields = configuredSearchFields.length > 0 ? configuredSearchFields.filter((field) => columns.some((column) => column.column_name === field)) : inferredFields;
+      const normalizedFieldMap = /* @__PURE__ */ new Map();
+      for (const field of availableFields) {
+        normalizedFieldMap.set(normalizeFieldName(field), field);
+      }
+      configByTable.set(table, {
+        availableFields,
+        searchableFields,
+        normalizedFieldMap
+      });
+    }
+    return configByTable;
+  }
+  parseConfiguredSearchFields() {
+    var _a;
+    const raw = (_a = this.config.options) == null ? void 0 : _a.searchFields;
+    if (Array.isArray(raw)) {
+      return raw.filter((field) => typeof field === "string" && field.trim().length > 0);
+    }
+    return [];
+  }
 };
 export {
   AnthropicProvider,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@retrivora-ai/rag-engine",
-  "version": "0.2.6",
+  "version": "0.2.8",
   "description": "Retrivora AI is a plug-and-play AI engine for RAG chat experiences — generic vector DB + LLM provider, embeddable or standalone.",
   "author": "Abhinav Alkuchi",
   "license": "MIT",

package/src/config/RagConfig.ts

@@ -28,6 +28,10 @@ export interface VectorDBConfig {
    * - idPath?: string (e.g. '_id')
    * - scorePath?: string (e.g. 'similarity')
    * - contentPath?: string (e.g. 'text')
+   *
+   * For multi-table PostgreSQL search, the following options are also supported:
+   * - tables?: string[] | string
+   * - searchFields?: string[] // optional override for which columns receive exact-match boosts
    */
   options: Record<string, unknown>;
 }

package/src/core/Pipeline.ts

@@ -7,6 +7,61 @@ import { BatchProcessor, BatchOptions } from './BatchProcessor';
 import { EmbeddingStrategyResolver } from '../config/EmbeddingStrategy';
 import { IngestDocument, ChatResponse } from '../types';
 
+interface QueryFieldHint {
+  field?: string;
+  value: string;
+}
+
+function normalizeHintValue(value: string): string {
+  return value.replace(/\s+/g, ' ').trim();
+}
+
+function extractQueryFieldHints(question: string): QueryFieldHint[] {
+  if (!question.trim()) return [];
+
+  const hints = new Map<string, QueryFieldHint>();
+
+  const addHint = (value: string, field?: string) => {
+    const normalizedValue = normalizeHintValue(value);
+    if (!normalizedValue) return;
+
+    const normalizedField = field
+      ? field
+          .toLowerCase()
+          .replace(/[^a-z0-9]+/g, ' ')
+          .trim()
+      : undefined;
+
+    const key = `${normalizedField ?? '*'}::${normalizedValue.toLowerCase()}`;
+    if (!hints.has(key)) {
+      hints.set(key, {
+        value: normalizedValue,
+        ...(normalizedField ? { field: normalizedField } : {}),
+      });
+    }
+  };
+
+  for (const match of question.matchAll(/["']([^"']{2,100})["']/g)) {
+    addHint(match[1]);
+  }
+
+  const fieldPattern = `([^\\n:=?.!,]{1,60}?)`;
+  const valuePattern = `([^\\n?.!,]{1,120}?)`;
+  const fieldValuePatterns = [
+    new RegExp(`\\b${fieldPattern}\\s*(?:=|:)\\s*["']?${valuePattern}["']?(?=[?.!,]|$)`, 'gi'),
+    new RegExp(`\\b${fieldPattern}\\s+(?:is|are|was|were|equals?|equal to|named|called)\\s+["']?${valuePattern}["']?(?=[?.!,]|$)`, 'gi'),
+    new RegExp(`\\bwith\\s+${fieldPattern}\\s+["']?${valuePattern}["']?(?=[?.!,]|$)`, 'gi'),
+  ];
+
+  for (const pattern of fieldValuePatterns) {
+    for (const match of question.matchAll(pattern)) {
+      addHint(match[2], match[1]);
+    }
+  }
+
+  return [...hints.values()];
+}
+
 /**
  * Pipeline — orchestrates the RAG flow: Embed → Search → Augment → Generate.
  *
@@ -135,8 +190,12 @@ export class Pipeline {
 
     try {
       const queryVector = await this.embeddingProvider.embed(question, { taskType: 'query' });
+      const fieldHints = extractQueryFieldHints(question);
       // Pass the original text to the vector DB so it can perform hybrid (keyword) search if supported
-      const rawMatches = await this.vectorDB.query(queryVector, topK, ns, { __queryText: question });
+      const rawMatches = await this.vectorDB.query(queryVector, topK, ns, {
+        __queryText: question,
+        __fieldHints: fieldHints,
+      });
 
       const sources = rawMatches.filter((m) => m.score >= scoreThreshold);
       const context = sources.length

package/src/providers/vectordb/MongoDBProvider.ts

@@ -3,6 +3,14 @@ import { VectorDBConfig } from '../../config/RagConfig';
 import { BaseVectorProvider } from './BaseVectorProvider';
 import { VectorMatch, UpsertDocument } from '../../types';
 
+function stripInternalFilterKeys(filter?: Record<string, unknown>): Record<string, unknown> {
+  if (!filter) return {};
+
+  return Object.fromEntries(
+    Object.entries(filter).filter(([key]) => !key.startsWith('__'))
+  );
+}
+
 /**
  * MongoDBProvider — MongoDB Atlas Vector Search implementation.
  */
@@ -67,6 +75,7 @@ export class MongoDBProvider extends BaseVectorProvider {
   }
 
   async query(vector: number[], topK: number, namespace?: string, filter?: Record<string, unknown>): Promise<VectorMatch[]> {
+    const publicFilter = stripInternalFilterKeys(filter);
     const pipeline: Record<string, unknown>[] = [
       {
         $vectorSearch: {
@@ -75,7 +84,9 @@ export class MongoDBProvider extends BaseVectorProvider {
           queryVector: vector,
           numCandidates: Math.max(topK * 10, 100),
           limit: topK,
-          ...(filter || namespace ? { filter: { ...(filter || {}), ...(namespace ? { namespace } : {}) } } : {}),
+          ...(Object.keys(publicFilter).length > 0 || namespace
+            ? { filter: { ...publicFilter, ...(namespace ? { namespace } : {}) } }
+            : {}),
         },
       },
       {