From 3c596100879b321cd11423e8cbef45091b59ef07 Mon Sep 17 00:00:00 2001 From: rulego-team Date: Mon, 16 Jun 2025 20:12:31 +0800 Subject: [PATCH] chore:add readme --- README.md | 146 --------------------------------------------------- README_ZH.md | 33 ------------ 2 files changed, 179 deletions(-) diff --git a/README.md b/README.md index 51a6aa3..6eaa202 100644 --- a/README.md +++ b/README.md @@ -237,152 +237,6 @@ func main() { } ``` -### 🔍 Pattern Matching Notes - -**Note**: StreamSQL currently does not support standard SQL `LIKE` syntax, but provides more powerful alternatives: - -- **Prefix matching**: `STARTSWITH(field, 'prefix')` replaces `field LIKE 'prefix%'` -- **Suffix matching**: `ENDSWITH(field, 'suffix')` replaces `field LIKE '%suffix'` -- **Contains matching**: `INDEXOF(field, 'substring') >= 0` replaces `field LIKE '%substring%'` -- **Regex matching**: `REGEXP_MATCHES(field, '^pattern$')` supports complex pattern matching - -**Examples**： -```sql --- Replace: deviceId LIKE 'sensor%' -WHERE STARTSWITH(deviceId, 'sensor') - --- Replace: message LIKE '%error%' -WHERE INDEXOF(message, 'error') >= 0 - --- Replace complex pattern: deviceId LIKE 'sensor[0-9]+' -WHERE REGEXP_MATCHES(deviceId, '^sensor[0-9]+$') -``` - -### Non-Aggregation Scenarios - -StreamSQL supports real-time data transformation and filtering without aggregation operations. This mode provides immediate processing and output for each input record, making it ideal for data cleaning, enrichment, and real-time filtering. - -```go -// Real-time data transformation and filtering example -package main - -import ( - "fmt" - "time" - "github.com/rulego/streamsql" -) - -func main() { - ssql := streamsql.New() - defer ssql.Stop() - - // Non-aggregation SQL - immediate data transformation - rsql := `SELECT deviceId, - upper(deviceType) as device_type, - temperature * 1.8 + 32 as temp_fahrenheit, - concat(location, '-', deviceId) as full_location, - now() as processed_time - FROM stream - WHERE temperature > 0 AND deviceId LIKE 'sensor%'` - - err := ssql.Execute(rsql) - if err != nil { - panic(err) - } - - // Handle real-time transformation results - ssql.Stream().AddSink(func(result interface{}) { - fmt.Printf("Transformed data: %+v\n", result) - }) - - // Input raw data - rawData := []map[string]interface{}{ - { - "deviceId": "sensor001", - "deviceType": "temperature", - "temperature": 25.0, - "humidity": 60, - "location": "warehouse-A", - }, - { - "deviceId": "sensor002", - "deviceType": "humidity", - "temperature": 22.5, - "humidity": 55, - "location": "warehouse-B", - }, - { - "deviceId": "pump001", // Will be filtered out - "deviceType": "actuator", - "temperature": 30.0, - "location": "factory", - }, - } - - // Each data record is processed immediately - for _, data := range rawData { - ssql.Stream().AddData(data) - time.Sleep(100 * time.Millisecond) // Simulate real-time arrival - } - - time.Sleep(500 * time.Millisecond) // Wait for processing -} -``` - -#### Use Cases for Non-Aggregation Scenarios - -**1. Real-time Data Cleaning and Validation** -```sql --- Filter invalid records and normalize data formats -SELECT deviceId, - CAST(temperature AS FLOAT) as temperature, - LOWER(status) as status, - COALESCE(location, 'unknown') as location -FROM stream -WHERE temperature IS NOT NULL AND deviceId != '' -``` - -**2. Data Enrichment and Transformation** -```sql --- Add calculated fields and enrichment -SELECT *, - temperature * 1.8 + 32 as temp_fahrenheit, - CASE WHEN temperature > 30 THEN 'hot' - WHEN temperature < 10 THEN 'cold' - ELSE 'normal' END as temp_category, - FORMAT(humidity, 2) as formatted_humidity -FROM stream -``` - -**3. Real-time Alerting and Monitoring** -```sql --- Filter critical events for immediate alerting -SELECT deviceId, temperature, humidity, now() as alert_time -FROM stream -WHERE temperature > 50 OR humidity < 10 -``` - -**4. Data Format Conversion** -```sql --- Convert data format for downstream systems -SELECT TO_JSON(MAP( - 'id', deviceId, - 'metrics', MAP('temp', temperature, 'hum', humidity), - 'meta', MAP('location', location, 'type', deviceType) -)) as json_output -FROM stream -``` - -**5. Real-time Data Routing** -```sql --- Route data based on conditions -SELECT *, - CASE WHEN deviceType = 'sensor' THEN 'sensor_topic' - WHEN deviceType = 'actuator' THEN 'actuator_topic' - ELSE 'default_topic' END as routing_key -FROM stream -``` - ### Nested Field Access StreamSQL supports querying nested structured data using dot notation (`.`) syntax to access nested fields: diff --git a/README_ZH.md b/README_ZH.md index c4c3bd9..6984a45 100644 --- a/README_ZH.md +++ b/README_ZH.md @@ -253,39 +253,6 @@ func main() { } ``` -### 🔍 模式匹配功能 - -StreamSQL 支持标准 SQL 的 `LIKE` 语法进行模式匹配： - -- **前缀匹配**: `field LIKE 'prefix%'` - 匹配以指定前缀开头的字符串 -- **后缀匹配**: `field LIKE '%suffix'` - 匹配以指定后缀结尾的字符串 -- **包含匹配**: `field LIKE '%substring%'` - 匹配包含指定子字符串的字符串 -- **单字符通配符**: `field LIKE 'patte_n'` - `_` 匹配任意单个字符 -- **复杂模式**: `field LIKE 'prefix%suffix'` - 组合前缀和后缀匹配 - -**示例**： -```sql --- 前缀匹配：查找以'sensor'开头的设备ID -WHERE deviceId LIKE 'sensor%' - --- 后缀匹配：查找以'error'结尾的消息 -WHERE message LIKE '%error' - --- 包含匹配：查找包含'alert'的日志 -WHERE logMessage LIKE '%alert%' - --- 单字符通配符：匹配三位数错误代码如E01, E02等 -WHERE errorCode LIKE 'E_0' - --- 复杂模式：匹配log_开头.log结尾的文件 -WHERE filename LIKE 'log_%.log' -``` - -**兼容的字符串函数**： -- `STARTSWITH(field, 'prefix')` - 等价于 `field LIKE 'prefix%'` -- `ENDSWITH(field, 'suffix')` - 等价于 `field LIKE '%suffix'` -- `REGEXP_MATCHES(field, '^pattern$')` - 支持更复杂的正则表达式匹配 - ### 嵌套字段访问 StreamSQL 还支持对嵌套结构数据进行查询，可以使用点号（`.`）语法访问嵌套字段：