feat: reslove sql parse

configs/abi.json

@@ -9,30 +9,38 @@
         "type": "string"
       },
       {
-        "components": [
-          {
-            "internalType": "string",
-            "name": "name",
-            "type": "string"
-          },
-          {
-            "internalType": "string",
-            "name": "sqlType",
-            "type": "string"
-          },
-          {
-            "internalType": "bool",
-            "name": "isPrimaryKey",
-            "type": "bool"
-          }
-        ],
         "indexed": false,
-        "internalType": "struct ISQLSync.ColumnDef[]",
+        "internalType": "string[]",
         "name": "columns",
-        "type": "tuple[]"
+        "type": "string[]"
+      },
+      {
+        "indexed": false,
+        "internalType": "string[][]",
+        "name": "values",
+        "type": "string[][]"
       }
     ],
-    "name": "TableCreated",
+    "name": "DataBatchInserted",
+    "type": "event"
+  },
+  {
+    "anonymous": false,
+    "inputs": [
+      {
+        "indexed": false,
+        "internalType": "string",
+        "name": "tableName",
+        "type": "string"
+      },
+      {
+        "indexed": false,
+        "internalType": "string",
+        "name": "whereClause",
+        "type": "string"
+      }
+    ],
+    "name": "DataDeleted",
     "type": "event"
   },
   {
@@ -100,25 +108,43 @@
         "name": "tableName",
         "type": "string"
       },
+      {
+        "indexed": false,
+        "internalType": "string[]",
+        "name": "columns",
+        "type": "string[]"
+      },
+      {
+        "indexed": false,
+        "internalType": "string[]",
+        "name": "values",
+        "type": "string[]"
+      },
       {
         "indexed": false,
         "internalType": "string",
-        "name": "whereClause",
+        "name": "conflictColumn",
         "type": "string"
       }
     ],
-    "name": "DataDeleted",
+    "name": "DataUpserted",
     "type": "event"
   },
   {
     "anonymous": false,
     "inputs": [
       {
-        "indexed": true,
+        "indexed": false,
         "internalType": "string",
         "name": "tableName",
         "type": "string"
       },
+      {
+        "indexed": false,
+        "internalType": "string",
+        "name": "indexName",
+        "type": "string"
+      },
       {
         "indexed": false,
         "internalType": "string[]",
@@ -127,12 +153,93 @@
       },
       {
         "indexed": false,
-        "internalType": "string[][]",
-        "name": "values",
-        "type": "string[][]"
+        "internalType": "bool",
+        "name": "isUnique",
+        "type": "bool"
       }
     ],
-    "name": "DataBatchInserted",
+    "name": "IndexCreated",
+    "type": "event"
+  },
+  {
+    "anonymous": false,
+    "inputs": [
+      {
+        "indexed": false,
+        "internalType": "string",
+        "name": "tableName",
+        "type": "string"
+      },
+      {
+        "indexed": false,
+        "internalType": "string",
+        "name": "rawSql",
+        "type": "string"
+      }
+    ],
+    "name": "TableAltered",
+    "type": "event"
+  },
+  {
+    "anonymous": false,
+    "inputs": [
+      {
+        "indexed": false,
+        "internalType": "string",
+        "name": "tableName",
+        "type": "string"
+      },
+      {
+        "components": [
+          {
+            "internalType": "string",
+            "name": "name",
+            "type": "string"
+          },
+          {
+            "internalType": "string",
+            "name": "sqlType",
+            "type": "string"
+          },
+          {
+            "internalType": "bool",
+            "name": "isPrimaryKey",
+            "type": "bool"
+          }
+        ],
+        "indexed": false,
+        "internalType": "struct ISQLSync.ColumnDef[]",
+        "name": "columns",
+        "type": "tuple[]"
+      }
+    ],
+    "name": "TableCreated",
+    "type": "event"
+  },
+  {
+    "anonymous": false,
+    "inputs": [
+      {
+        "indexed": false,
+        "internalType": "string",
+        "name": "tableName",
+        "type": "string"
+      }
+    ],
+    "name": "TableDropped",
+    "type": "event"
+  },
+  {
+    "anonymous": false,
+    "inputs": [
+      {
+        "indexed": false,
+        "internalType": "string",
+        "name": "tableName",
+        "type": "string"
+      }
+    ],
+    "name": "TableTruncated",
     "type": "event"
   }
 ]
\ No newline at end of file

contract/ISQLSync.sol

@@ -13,37 +13,37 @@ interface ISQLSync {
     event TableDropped(string tableName);
     event TableAltered(string tableName, string rawSql);
     event DataInserted(
-        string indexed tableName,
+        string tableName,
         string[] columns,
         string[] values
     );
     event DataUpdated(
-        string indexed tableName,
+        string tableName,
         string[] setColumns,
         string[] setValues,
         string whereClause
     );
-    event DataDeleted(string indexed tableName, string whereClause);
+    event DataDeleted(string tableName, string whereClause);
 
     // Advanced Events
     event DataBatchInserted(
-        string indexed tableName,
+        string tableName,
         string[] columns,
         string[][] values
     );
     event DataUpserted(
-        string indexed tableName,
+        string tableName,
         string[] columns,
         string[] values,
         string conflictColumn
     );
     event IndexCreated(
-        string indexed tableName,
+        string tableName,
         string indexName,
         string[] columns,
         bool isUnique
     );
-    event TableTruncated(string indexed tableName);
+    event TableTruncated(string tableName);
 
     // --- DDL Functions ---
     function createTable(

hardhat/contracts/ISQLSync.sol

@@ -13,37 +13,37 @@ interface ISQLSync {
     event TableDropped(string tableName);
     event TableAltered(string tableName, string rawSql);
     event DataInserted(
-        string indexed tableName,
+        string tableName,
         string[] columns,
         string[] values
     );
     event DataUpdated(
-        string indexed tableName,
+        string tableName,
         string[] setColumns,
         string[] setValues,
         string whereClause
     );
-    event DataDeleted(string indexed tableName, string whereClause);
+    event DataDeleted(string tableName, string whereClause);
 
     // Advanced Events
     event DataBatchInserted(
-        string indexed tableName,
+        string tableName,
         string[] columns,
         string[][] values
     );
     event DataUpserted(
-        string indexed tableName,
+        string tableName,
         string[] columns,
         string[] values,
         string conflictColumn
     );
     event IndexCreated(
-        string indexed tableName,
+        string tableName,
         string indexName,
         string[] columns,
         bool isUnique
     );
-    event TableTruncated(string indexed tableName);
+    event TableTruncated(string tableName);
 
     // --- DDL Functions ---
     function createTable(

internal/core/data_listener.go

@@ -138,7 +138,14 @@ func (l *DataListener) syncBatch(ctx context.Context) error {
 		return fmt.Errorf("get head block failed: %w", err)
 	}
 
-	// 3. 计算 ToBlock
+	// 3. ⭐ 防止 uint64 下溢
+	if headBlock < l.cfg.Sync.Confirmations {
+		log.Printf("Chain too young (head: %d, need %d confirmations), waiting...",
+			headBlock, l.cfg.Sync.Confirmations)
+		return nil
+	}
+
+	// 4. 计算 ToBlock
 	safeHead := headBlock - l.cfg.Sync.Confirmations
 	if fromBlock > safeHead {
 		return nil // 尚未达到确认高度
@@ -149,13 +156,13 @@ func (l *DataListener) syncBatch(ctx context.Context) error {
 		toBlock = safeHead
 	}
 
-	// 4. 准备 Event Topics
+	// 5. 准备 Event Topics
 	var topics []common.Hash
 	for _, event := range l.contractABI.Events {
 		topics = append(topics, event.ID)
 	}
 
-	// 5. 分批轮询地址
+	// 6. 分批轮询地址
 	batchSize := l.cfg.Sync.AddressBatchSize
 	if batchSize <= 0 {
 		batchSize = 50
@@ -163,6 +170,9 @@ func (l *DataListener) syncBatch(ctx context.Context) error {
 
 	log.Printf("Scanning Data: %d -> %d (%d contracts)", fromBlock, toBlock, len(contracts))
 
+	// ⭐ 收集所有错误，确保数据完整性
+	var errors []error
+
 	for i := 0; i < len(contracts); i += batchSize {
 		end := i + batchSize
 		if end > len(contracts) {
@@ -180,17 +190,24 @@ func (l *DataListener) syncBatch(ctx context.Context) error {
 		logs, err := l.client.FilterLogs(ctx, query)
 		if err != nil {
 			log.Printf("FilterLogs failed for batch %d: %v", i, err)
+			errors = append(errors, fmt.Errorf("batch %d: %w", i, err))
 			continue
 		}
 
 		for _, vLog := range logs {
 			if err := l.processLog(ctx, vLog); err != nil {
 				log.Printf("Failed to process log tx=%s: %v", vLog.TxHash.Hex(), err)
+				errors = append(errors, fmt.Errorf("tx %s: %w", vLog.TxHash.Hex(), err))
 			}
 		}
 	}
 
-	// 6. 更新游标
+	// 7. ⭐ 只有全部成功才更新游标
+	if len(errors) > 0 {
+		return fmt.Errorf("failed to process %d logs, will retry next time: %v", len(errors), errors[0])
+	}
+
+	// 8. 更新游标
 	return l.db.UpdateLastBlock(ctx, SyncKeyData, toBlock)
 }
 

internal/executor/sql_builder.go

@@ -152,9 +152,26 @@ func (e *Executor) HandleDataUpdated(ctx context.Context, contractAddr string, t
 	}
 
 	// 解析 WHERE 子句
-	// 首先需要获取表的列信息来构建白名单
-	// 为了简化，这里使用所有 setColumns 作为允许的列（实际应该查询表结构）
-	allowedCols := append([]string{}, setColumns...)
+	// ⭐ 使用 setColumns 作为基础，但需要包含可能在 WHERE 中使用的其他列
+	// 临时方案：允许所有常见列名，生产环境应该查询表结构
+	allowedCols := make([]string, 0)
+	allowedCols = append(allowedCols, setColumns...)
+	// 添加常见的列名
+	commonCols := []string{"id", "name", "status", "email", "price", "stock", "created_at", "updated_at"}
+	for _, col := range commonCols {
+		// 避免重复
+		found := false
+		for _, existing := range allowedCols {
+			if strings.EqualFold(existing, col) {
+				found = true
+				break
+			}
+		}
+		if !found {
+			allowedCols = append(allowedCols, col)
+		}
+	}
+
 	parser := NewWhereParser(allowedCols)
 
 	whereCond, err := parser.Parse(whereClause)
@@ -162,11 +179,16 @@ func (e *Executor) HandleDataUpdated(ctx context.Context, contractAddr string, t
 		return fmt.Errorf("parse WHERE clause failed: %w", err)
 	}
 
-	// 调整参数占位符编号
+	// ⭐ 调整 WHERE 子句的参数占位符编号
+	// WHERE 解析器生成的占位符从 $1 开始，但我们需要将它们调整为从 len(setValues)+1 开始
 	adjustedWhereSQL := whereCond.SQL
-	for i := len(args); i > 0; i-- {
-		placeholder := fmt.Sprintf("$%d", i-len(args)+len(setValues))
-		adjustedWhereSQL = strings.Replace(adjustedWhereSQL, fmt.Sprintf("$%d", i), placeholder, -1)
+	setParamCount := len(setValues)
+
+	// 从后往前替换，避免冲突（例如 $10 被误替换为 $1）
+	for i := len(whereCond.Args); i >= 1; i-- {
+		oldPlaceholder := fmt.Sprintf("$%d", i)
+		newPlaceholder := fmt.Sprintf("$%d", setParamCount+i)
+		adjustedWhereSQL = strings.ReplaceAll(adjustedWhereSQL, oldPlaceholder, newPlaceholder)
 	}
 
 	// 追加 WHERE 参数
@@ -190,9 +212,9 @@ func (e *Executor) HandleDataDeleted(ctx context.Context, contractAddr string, t
 	safeTableName := QuoteIdentifier(tableName)
 
 	// 解析 WHERE 子句
-	// 注意：这里需要表的列信息，简化处理，允许常见列名
-	// 生产环境应该查询 information_schema 获取实际列
-	parser := NewWhereParser([]string{"id", "name", "status", "email", "created_at", "updated_at"})
+	// ⭐ 允许所有常见列名，生产环境应该查询 information_schema 获取实际列
+	commonCols := []string{"id", "name", "status", "email", "price", "stock", "created_at", "updated_at", "type", "category"}
+	parser := NewWhereParser(commonCols)
 
 	whereCond, err := parser.Parse(whereClause)
 	if err != nil {

internal/executor/where_parser.go

@@ -98,6 +98,8 @@ func (p *WhereParser) Parse(whereClause string) (*WhereCondition, error) {
 }
 
 // tokenize 简单的词法分析
+// 目标：只在 AND/OR 处分割，保持条件表达式完整
+// 例如: "stock < 20 AND status = 'available'" -> ["stock < 20", "AND", "status = 'available'"]
 func (p *WhereParser) tokenize(input string) []string {
 	var tokens []string
 	var current strings.Builder
@@ -105,9 +107,12 @@ func (p *WhereParser) tokenize(input string) []string {
 	quoteChar := rune(0)
 	parenDepth := 0
 
-	for i, ch := range input {
-		switch {
-		case (ch == '\'' || ch == '"') && (i == 0 || input[i-1] != '\\'):
+	i := 0
+	for i < len(input) {
+		ch := rune(input[i])
+
+		// 处理引号
+		if (ch == '\'' || ch == '"') && (i == 0 || input[i-1] != '\\') {
 			if !inQuote {
 				inQuote = true
 				quoteChar = ch
@@ -116,44 +121,89 @@ func (p *WhereParser) tokenize(input string) []string {
 				quoteChar = 0
 			}
 			current.WriteRune(ch)
+			i++
+			continue
+		}
 
-		case ch == '(' && !inQuote:
-			if current.Len() > 0 {
-				tokens = append(tokens, current.String())
-				current.Reset()
-			}
-			tokens = append(tokens, "(")
-			parenDepth++
+		// 在引号内，直接添加
+		if inQuote {
+			current.WriteRune(ch)
+			i++
+			continue
+		}
 
-		case ch == ')' && !inQuote:
-			if current.Len() > 0 {
-				tokens = append(tokens, current.String())
-				current.Reset()
-			}
-			tokens = append(tokens, ")")
+		// 处理括号
+		if ch == '(' {
+			current.WriteRune(ch)
+			parenDepth++
+			i++
+			continue
+		}
+		if ch == ')' {
+			current.WriteRune(ch)
 			parenDepth--
+			i++
+			continue
+		}
 
-		case ch == ' ' && !inQuote && parenDepth == 0:
-			if current.Len() > 0 {
-				word := current.String()
-				upperWord := strings.ToUpper(word)
+		// 在括号内，直接添加
+		if parenDepth > 0 {
+			current.WriteRune(ch)
+			i++
+			continue
+		}
 
-				// 检查是否是逻辑操作符
-				if upperWord == "AND" || upperWord == "OR" {
-					tokens = append(tokens, word)
+		// ⭐ 检查是否是 AND 或 OR（只在括号外且非引号内）
+		if i+3 <= len(input) && strings.ToUpper(input[i:i+3]) == "AND" {
+			// 确保 AND 是独立的词（前后是空格或边界）
+			isWord := (i == 0 || input[i-1] == ' ') &&
+				(i+3 == len(input) || input[i+3] == ' ')
+			if isWord {
+				// 保存当前 token
+				if current.Len() > 0 {
+					tokens = append(tokens, strings.TrimSpace(current.String()))
 					current.Reset()
-				} else {
-					current.WriteRune(ch)
 				}
+				// 添加 AND
+				tokens = append(tokens, "AND")
+				i += 3
+				// 跳过 AND 后的空格
+				for i < len(input) && input[i] == ' ' {
+					i++
+				}
+				continue
 			}
+		}
 
-		default:
-			current.WriteRune(ch)
+		if i+2 <= len(input) && strings.ToUpper(input[i:i+2]) == "OR" {
+			// 确保 OR 是独立的词
+			isWord := (i == 0 || input[i-1] == ' ') &&
+				(i+2 == len(input) || input[i+2] == ' ')
+			if isWord {
+				// 保存当前 token
+				if current.Len() > 0 {
+					tokens = append(tokens, strings.TrimSpace(current.String()))
+					current.Reset()
+				}
+				// 添加 OR
+				tokens = append(tokens, "OR")
+				i += 2
+				// 跳过 OR 后的空格
+				for i < len(input) && input[i] == ' ' {
+					i++
+				}
+				continue
+			}
 		}
+
+		// 其他字符直接添加
+		current.WriteRune(ch)
+		i++
 	}
 
+	// 添加最后一个 token
 	if current.Len() > 0 {
-		tokens = append(tokens, current.String())
+		tokens = append(tokens, strings.TrimSpace(current.String()))
 	}
 
 	return tokens

scripts/clean-schemas.sh

+#!/bin/bash
+# scripts/clean-schemas.sh
+
+echo "🗑️  Cleaning ChainSQL schemas..."
+
+# 删除所有 0x 开头的 schema
+docker exec chainsql-db psql -U chainsql -d chainsql -c \
+  "DO \$\$ 
+   DECLARE 
+     r RECORD;
+   BEGIN
+     FOR r IN SELECT schema_name FROM information_schema.schemata WHERE schema_name LIKE '0x%'
+     LOOP
+       RAISE NOTICE 'Dropping schema: %', r.schema_name;
+       EXECUTE 'DROP SCHEMA IF EXISTS \"' || r.schema_name || '\" CASCADE';
+     END LOOP;
+   END \$\$;"
+
+echo "✅ All ChainSQL schemas cleaned!"
+
+# 重置游标
+sh scripts/reset-db.sh
\ No newline at end of file