经过年复一年的开发积累，企业的代码仓逐渐变得臃肿，甚至变成屎山代码。这些屎山代码，往往经过N个程序员之手，他们水平参差不起，风格不一。如何对这些屎山代码进行统一的管理，让它们可以被监控、评价和批量改造？建立“代码管理系统”的第一个难点在于，如何在庞大的代码仓中，快速的查找出具有某些特征的代码段。由于我们需要查找的是代码段，而不是代码行，用传统的正则表达式难以实现，需要通过语法解析器进行自定义语法配置，然后进行代码查找。 以小实例说明 ： ### 实例1： 找出JAVA代码中，入参数量超过4个的函数：# 配置查找规则(Code_manage.syn)如下所示：__DEF_CASE_SENSITIVE__ Y __DEF_FUZZY__ Y __DEF_DEBUG__ N __DEF_LINE_COMMENT__ // __DEF_LINES_COMMENT__ /* */ __DEF_STR__ __NAME__ <1,200> [1,1]ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz_$?? [0,199]ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789_?? [NO] import if else for while break continue class return try except finally final static public private __DEF_PATH__ __FUNCTION_DEF__ 0101 : x1 @ | public : x2 @ + private 0 0 : x3 @ CAN_SKIP | static 1 1 : x4 @ | __NAME__ : x5 @ | __NAME__ : x6 @ | ( 1111 : p1 @ CAN_SKIP | final : p11 @ | __NAME__ : p111 @ | __NAME__ : p2 @ | , : p22 @ CAN_SKIP | final : p222 @ | __NAME__ : p2222 @ | __NAME__ : p3 @ | , : p33 @ CAN_SKIP | final : p333 @ | __NAME__ : p3333 @ | __NAME__ NNNN : p4 @ | , : p44 @ CAN_SKIP | final : x444 @ | __NAME__ : x4444 @ | __NAME__ 1111 : xx @ | )# 假设java代码(MyCode.java) 如下所示：private int alreadyBufferedSize = 0; // The index in the byte[] found at buffers.getLast() to be written next private int index = 0; // Is the stream closed? private boolean closed = false; public FastByteArrayOutputStream(int initialBlockSize) { Assert.isTrue(initialBlockSize > 0, "Initial block size must be greater than 0"); this.initialBlockSize = initialBlockSize; this.nextBlockSize = initialBlockSize; } @Override public void applyBeanPropertyValues(Object existingBean, String beanName, int autowireMode, boolean dependencyCheck, int initSize) throws BeansException { markBeanAsCreated(beanName); BeanDefinition bd = getMergedBeanDefinition(beanName); BeanWrapper bw = new BeanWrapperImpl(existingBean); initBeanWrapper(bw); applyPropertyValues(beanName, bd, bw, bd.getPropertyValues()); } @Override public Object initializeBean(Object existingBean, String beanName) { return initializeBean(beanName, existingBean, null); }根据配置规则，执行查找命令： ZGLanguage -e Code_manage.syn -f MyCode.java可以得到结果：C:\>ZGLanguage -e Code_manage.syn -f MyCode.java Run type : Find Syntax file : Code_manage.syn code file : MyCode.java Output file : out.zgl -------------------------------------------------------------------- ### Found code by : __FUNCTION_DEF__ | Lines : 17 ~ 17 : -------------------------------------------------------------------- public void applyBeanPropertyValues(Object existingBean, String beanName, int autowireMode, boolean dependencyCheck, int initSize)可以看出，查找结果只输出了函数 applyBeanPropertyValues，它的入参数量为5个，其他2个函数的入参均不超过4个，因此被忽略。 ### 实例2： 提取SQL代码中的关联(on)和筛选(where)代码段：# 配置查找规则(Code_manage.syn)如下所示：__DEF_DEBUG__ N __DEF_FUZZY__ Y __DEF_CASE_SENSITIVE__ N __DEF_LINE_COMMENT__ -- __DEF_LINES_COMMENT__ /* */ __DEF_PATH__ __WHERE__ 1 : x1 | where : x2 | __PATH_4_EXPR__ __DEF_PATH__ __ON__ 1 : x1 | __\b__ : x2 + __\t__ : x3 + __\n__ : x4 | on : x5 | __PATH_4_EXPR__ __DEF_SUB_PATH__ __PATH_4_EXPR__ 1 : x1 | __SUB_PATH_EXPR__ : x2 + __ONE_PATH_EXPR__ __DEF_SUB_PATH__ __SUB_PATH_EXPR__ 1 : x1 | ( : x2 | __ONE_PATH_EXPR__ : x3 | ) __DEF_SUB_PATH__ __ONE_PATH_EXPR__ NN : @ | __NAME__ : @ + __INT__ : @ + __FLOAT__ : @ + __CASE_WHEN__ : @ + __STRING__ : @ + __CAST_AS__ : @ + __FUNCTION__ : @ + __SUB_PATH_EXPR__ : @ + = : @ + <> : @ + != : @ + > : @ + >= : @ + < : @ + <= : @ + . : @ + , : @ + + : @ + - : @ + * : @ + / : @ + || : @ + null : @ + between : @ + and : @ + or : @ + like : @ + in : @ STRING + not in : @ STRING + is null : @ STRING + is not null __DEF_STR__ __NAME__ <1,100> [1,1]ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz_?? [0,100]ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789_?? [NO] select inner left join on from where group order by having union all with as table date time __DEF_STR__ __FLOAT__ <1,100> [1,50]0123456789 [1,1]. [1,50]0123456789 __DEF_STR__ __INT__ <1,100> [1,100]0123456789 __DEF_SUB_PATH__ __STRING__ 1 : x1 | ' : x2 | __ANY__ : x3 | ' __DEF_SUB_PATH__ __DECIMAL__ 111 : x1 | decimal 0 : x2 | ( 01 : x3 | __INT__ 00 : x4 | , 00 : x5 | __INT__ 01 : x6 | ) __DEF_SUB_PATH__ __VAR_NAME__ 1 : x1 | $ : x2 | { : x3 | __NAME__ : x4 | } __DEF_SUB_PATH__ __CASE_WHEN__ 1 : x1 @ | case N : x2 @ | when : x3 @ | __PATH_4_EXPR__ : x4 @ | then : x5 @ | __PATH_4_EXPR__ 1 : x6 @ CAN_SKIP | else : x7 @ CAN_SKIP | __PATH_4_EXPR__ : x8 @ | end __DEF_SUB_PATH__ __CAST_AS__ 1 : x1 | cast : x2 | ( 1 : x3 | __PATH_4_EXPR__ : x4 | as : x5 | date : x6 + int : n1 + double : n2 + float : n3 + bigint : x8 + __DECIMAL__ 1 : xx | ) __DEF_SUB_PATH__ __FUNCTION__ 1 : x1 @ | __NAME__ : x2 @ | ( N : x3 @ CAN_SKIP | __PATH_4_EXPR__ e : x4 @ CAN_SKIP | , 1 : x5 @ | )# 假设SQL代码(myproc.sql) 如下所示：CREATE OR REPLACE PROCEDURE PROC_F_CWWS_LOAN ( P_AS_OF_DATE IN DATE, RET_FLG OUT VARCHAR2, RET_MSG OUT VARCHAR2 ) IS -- 声明变量并初始化 V_COUNT NUMBER := 0; V_PROC_NAME VARCHAR2(200) := 'PROC_F_CWWS_LOAN'; V_PROC_DESC VARCHAR2(100) := 'xxxx业务数据ETL处理'; V_P_FREQ VARCHAR2(4) := ''; BEGIN --写入初始日志 INSERT INTO M_RUNLOG VALUES (SYSDATE, V_PROC_NAME, 'it is 1'); COMMIT; --设置会话日期格式 EXECUTE IMMEDIATE ' ALTER SESSION SET NLS_DATE_FORMAT = ''YYYY-MM-DD'''; --查询参数表中，该程序对应的频率值 SELECT P_FREQ INTO V_P_FREQ FROM ETL_PROC_STATUS_DEF WHERE PROC_NAME = V_PROC_NAME; --判断是调度频率 ETL.ETL_ADD_PARTITION('MA_F_LOAN', P_AS_OF_DATE, 'ETL'); --从还款计划表中取每笔账户最近一次小于等于数据日期还款日，作为上次还款日 INSERT INTO ETL.TMP_XD_LAST_PAYDATE (OBJECTNO, LAST_PAYDATE) SELECT OBJECTNO, LAST_PAYDATE FROM (SELECT T.OBJECTNO, MAX(TO_DATE(PAYDATE, 'YYYY-MM-DD')) LAST_PAYDATE FROM NYBDP.O_CWWS_ACCT_PAYMENT_SCHEDULE T WHERE T.AS_OF_DATE = P_AS_OF_DATE AND T.SEQID <> '999' AND TO_DATE(T.PAYDATE, 'YYYY-MM-DD') < P_AS_OF_DATE GROUP BY T.OBJECTNO); INSERT INTO M_RUNLOG VALUES (SYSDATE, V_PROC_NAME, 'it is 3'); COMMIT; MERGE INTO ETL.MA_F_LOAN A USING (SELECT /*+PARALLEL(8)*/ T.ACCOUNT_NUMBER, T.GL_ACCOUNT_ID, T.INT_GL_ACCOUNT_ID FROM ETL.MA_F_LOAN T INNER JOIN ETL.MA_D_GL_SUBJECT T1 ON T.INT_GL_ACCOUNT_ID = T1.SUBJECT_NO3 AND T1.SUBJECT_NAME3 LIKE '%已减值%' AND T1.AS_OF_DATE = P_AS_OF_DATE WHERE T.AS_OF_DATE = P_AS_OF_DATE AND T.ACCOUNT_NUMBER IN (SELECT ACCOUNT_NUMBER FROM (SELECT /*+PARALLEL(8)*/ T2.ACCOUNT_NUMBER, COUNT(1) FROM ETL.MA_F_LOAN T2 WHERE T2.AS_OF_DATE = P_AS_OF_DATE GROUP BY T2.ACCOUNT_NUMBER HAVING COUNT(1) > 1))) B ON (A.ACCOUNT_NUMBER = B.ACCOUNT_NUMBER AND A.AS_OF_DATE = P_AS_OF_DATE AND A.GL_ACCOUNT_ID = B.GL_ACCOUNT_ID AND A.INT_GL_ACCOUNT_ID = B.INT_GL_ACCOUNT_ID) WHEN MATCHED THEN UPDATE SET A.CUR_BOOK_BAL = 0, A.OVERDUE_BAL = 0; COMMIT; RET_FLG := '0'; RET_MSG := '执行成功'; EXCEPTION WHEN OTHERS THEN --写入异常日志 ETL.PROC_ETL_LOG(P_AS_OF_DATE,V_PROC_NAME,V_PROC_DESC,V_COUNT,-1,SQLCODE,SQLERRM); RET_MSG := SQLCODE || ':' || SQLERRM; END; /根据配置规则，执行查找命令： ZGLanguage -e Code_manage.syn -f myproc.sql可以得到结果：C:\>ZGLanguage -e Code_manage.syn -f myproc.sql Run type : Find Syntax file : Code_manage.syn code file : myproc.sql Output file : out.zgl -------------------------------------------------------------------- ### Found code by : __WHERE__ | Lines : 27 ~ 27 : -------------------------------------------------------------------- WHERE PROC_NAME = V_PROC_NAME -------------------------------------------------------------------- ### Found code by : __WHERE__ | Lines : 39 ~ 41 : -------------------------------------------------------------------- WHERE T.AS_OF_DATE = P_AS_OF_DATE AND T.SEQID <> '999' AND TO_DATE(T.PAYDATE, 'YYYY-MM-DD') < P_AS_OF_DATE -------------------------------------------------------------------- ### Found code by : __ON__ | Lines : 52 ~ 54 : -------------------------------------------------------------------- ON T.INT_GL_ACCOUNT_ID = T1.SUBJECT_NO3 AND T1.SUBJECT_NAME3 LIKE '%宸插噺鍊?' AND T1.AS_OF_DATE = P_AS_OF_DATE -------------------------------------------------------------------- ### Found code by : __WHERE__ | Lines : 55 ~ 56 : -------------------------------------------------------------------- WHERE T.AS_OF_DATE = P_AS_OF_DATE AND T.ACCOUNT_NUMBER IN -------------------------------------------------------------------- ### Found code by : __WHERE__ | Lines : 61 ~ 61 : -------------------------------------------------------------------- WHERE T2.AS_OF_DATE = P_AS_OF_DATE -------------------------------------------------------------------- ### Found code by : __ON__ | Lines : 64 ~ 66 : -------------------------------------------------------------------- ON (A.ACCOUNT_NUMBER = B.ACCOUNT_NUMBER AND A.AS_OF_DATE = P_AS_OF_DATE AND A.GL_ACCOUNT_ID = B.GL_ACCOUNT_ID AND A.INT_GL_ACCOUNT_ID = B.INT_GL_ACCOUNT_ID) WHEN MATCHED THEN UPDATE SET A.CUR_BOOK_BAL = 0, A.OVERDUE_BAL = 0可以看出，查找结果将 myproc.sql 代码中的 where 和 on 代码块及其所在行号提取出来。  

import http from '@ohos.net.http';let client = http.createHttp()@Entry@Componentstruct Page { @State httpRequest: string = 'Hello World' private demoUrl = 'https://c4ebde166f0740acbcfcc853f7ecf9d7.apig.cn-north-4.huaweicloudapis.com/API_library_data_12' build() { Row() { Column() { Text(this.httpRequest) .onClick(()=>{ this.onRequestHttp() }) } .width('100%') } } onRequestHttp(){ client.request(this.demoUrl,(error,data:http.HttpResponse)=>{ this.httpRequest=data.result.toString() client.destroy(); }) }}