最近參與了一個信托行業的BI項目,由於信托業務系統設計的問題,很多都是用戶手工錄入的數據,也有一些是需要分析的但是用戶沒有錄入的數據,針對這樣的數據質量,我們就要在ETL抽取的過程中來對數據流進行校驗,今天我們就說一下如何利用ETL開源工具kettle來完成對數據的基礎性校驗
1:非空校驗
if(INTRUST_FLAG1==null){intrust_flag1_isnull=1;}
2:唯一校驗
var uniquesql_item_id = "SELECT count(*) FROM titem where item_id="+ITEM_ID; if(fireToDB(strConn,uniquesql_item_id)[0[0]==1){item_id_isunique=1;}
3:標准化校驗
var normalsql_intrust_type= "select count(*) from trust_tdictparam where type_id=1104 and type_value="+INTRUST_TYPE;
if(fireToDB(strConn, normalsql_intrust_type)[0][0]>0){intrust_type_isnormal=1;}
完整的JS處理邏輯展示如下

var check_status=0;//如果檢測到有任何一種校驗錯誤,則check_status=1 var check_table="T_ITEM_TITEM";//校驗有關的表名字,可能存在多個 var source_table="titem";//數據源表名稱,如果數據來自多個數據表,則需要聲明多個 var source_table_id="item_id";//數據源表主鍵,如果多個表聯合主鍵,則需要聲明多個主鍵 var check_detail="";//校驗到的錯誤詳細情況 var check_name="";//主鍵名稱 var check_type="";//校驗到的錯誤類型 var check_date=new Date();//校驗時間 var strConn = "bjitic_og"; //////////////////////////////////定義所有校驗枚舉值//////////////////// //1:唯一性枚舉值 var item_id_isunique=0; var item_code_isunique=0; var item_name_isunique=0; //2:非空枚舉值 var intrust_flag1_isnull=0; var intrust_type_isnull=0; var intrust_type_subitem_isnull=0; var intrust_type1_isnull=0; var intrust_type2_isnull=0; var item_num_isnull=0; var item_money_isnull=0; var exp_rate1_isnull=0; var exp_rate2_isnull=0; var intrust_flag3_isnull=0; var intrust_flag4_isnull=0; var entity_type_isnull=0; var deal_type_isnull=0; var managertype_isnull=0; var natrust_type_isnull=0; var custodian_bank_isnull=0; var with_bank_flag_isnull=0; var with_security_flag_isnull=0; var with_private_flag_isnull=0; var with_gov_flag_isnull=0; var ben_period_isnull=0; //3:標准化枚舉值 var intrust_type_isnormal=0; var intrust_type_subitem_isnormal=0; var intrust_type1_isnormal=0; var intrust_type2_isnormal=0; var entity_type_isnormal=0; var deal_type_isnormal=0; var natrust_type_isnormal=0; var custodian_bank_isnormal=0; var innovat_type_isnormal=0; var bank_id_isnormal=0; var gov_regional_isnormal=0; //4:類型校驗枚舉值 var ben_period_isnumber=0; ////////////////////////////////唯一性校驗枚舉值賦值 var uniquesql_item_id = "SELECT count(1) FROM titem where item_id="+ITEM_ID; var uniquesql_item_code = "SELECT count(1) FROM titem where item_code="+ITEM_CODE; var uniquesql_item_name = "SELECT count(1) FROM titem where item_name="+"'"+ITEM_NAME+"'"; if(fireToDB(strConn,uniquesql_item_id)[0][0]==1){item_id_isunique=1;} if(fireToDB(strConn,uniquesql_item_code)[0][0]==1){item_code_isunique=1;} if(fireToDB(strConn,uniquesql_item_name)[0][0]==1){item_name_isunique=1;} ////////////////////////////////非空校驗枚舉值賦值 if(INTRUST_FLAG1==null){intrust_flag1_isnull=1;} if(INTRUST_TYPE==null){intrust_type_isnull=1;} if(INTRUST_TYPE_SUBITEM==null){intrust_type_subitem_isnull=1;} if(INTRUST_TYPE1==null){intrust_type1_isnull=1;} if(INTRUST_TYPE2==null){intrust_type2_isnull=1;} if(ITEM_NUM==null){item_num_isnull=1;} if(ITEM_MONEY==null){item_money_isnull=1;} if(EXP_RATE1==null){exp_rate1_isnull=1;} if(EXP_RATE2==null){exp_rate2_isnull=1;} if(INTRUST_FLAG3==null){intrust_flag3_isnull=1;} if(INTRUST_FLAG4==null){intrust_flag4_isnull=1;} if(ENTITY_TYPE==null){entity_type_isnull=1;} if(DEAL_TYPE==null){deal_type_isnull=1;} if(MANAGERTYPE==null){managertype_isnull=1;} if(NATRUST_TYPE==null){natrust_type_isnull=1;} if(CUSTODIAN_BANK==null){custodian_bank_isnull=1;} if(WITH_BANK_FLAG==null){with_bank_flag_isnull=1;} if(WITH_SECURITY_FLAG==null){with_security_flag_isnull=1;} if(WITH_PRIVATE_FLAG==null){with_private_flag_isnull=1;} if(WITH_GOV_FLAG==null){with_gov_flag_isnull=1;} if(BEN_PERIOD==null){ben_period_isnull=1;} ////////////////////////////////標准化校驗枚舉值賦值 var normalsql_intrust_type= "select count(1) from trust_tdictparam where type_id=1104 and type_value="+INTRUST_TYPE; var normalsql_intrust_type_subitem= "select count(1) from trust_tdictparam where type_id=1151 and type_value="+INTRUST_TYPE_SUBITEM; var normalsql_intrust_type1= "select count(1) from trust_tdictparam where type_id=1138 and type_value="+INTRUST_TYPE1; var normalsql_intrust_type2= "select count(1) from trust_tdictparam where type_id=1139 and type_value="+INTRUST_TYPE2; var normalsql_entity_type= "select count(1) from trust_tdictparam where type_id=1140 and type_value="+ENTITY_TYPE; var normalsql_deal_type= "select count(1) from trust_tdictparam where type_id=1141 and type_value="+DEAL_TYPE; var normalsql_natrust_type= "select count(1) from trust_tdictparam where type_id=1143 and type_value="+NATRUST_TYPE; var normalsql_custodian_bank= "select count(1) from trust_tdictparam where type_id=1103 and type_value="+"'"+CUSTODIAN_BANK+"'";; var normalsql_innovat_type= "select count(1) from trust_tdictparam where type_id=1104 and type_value="+INNOVAT_TYPE; var normalsql_gov_regional= "select count(1) from trust_tdictparam where type_id=9999 and type_value="+GOV_PROV_REGIONAL; if(fireToDB(strConn, normalsql_intrust_type)[0][0]>0){intrust_type_isnormal=1;} if(fireToDB(strConn, normalsql_intrust_type_subitem)[0][0]>0){intrust_type_subitem_isnormal=1;} if(fireToDB(strConn, normalsql_intrust_type1)[0][0]>0){intrust_type1_isnormal=1;} if(fireToDB(strConn, normalsql_intrust_type2)[0][0]>0){intrust_type2_isnormal=1;} if(fireToDB(strConn, normalsql_entity_type)[0][0]>0){entity_type_isnormal=1;} if(fireToDB(strConn, normalsql_deal_type)[0][0]>0){deal_type_isnormal=1;} if(fireToDB(strConn, normalsql_natrust_type)[0][0]>0){natrust_type_isnormal=1;} if(fireToDB(strConn, normalsql_custodian_bank)[0][0]>0){custodian_bank_isnormal=1;} if(fireToDB(strConn, normalsql_innovat_type)[0][0]>0){innovat_type_isnormal=1;} if(fireToDB(strConn, normalsql_gov_regional)[0][0]>0){gov_regional_isnormal=1;} ////////////////////////////////數據類型校驗枚舉值賦值 if(isNum(BEN_PERIOD)) { ben_period_isnumber=1; } //校驗所有表需要校驗的字段,如果有一個校驗失敗,則校驗狀態為1 if ( item_id_isunique==0 || item_code_isunique==0 ||item_name_isunique==0|| //is unique intrust_flag1_isnull==1 || intrust_type_isnull==1 || intrust_type_subitem_isnull==1 ||//is null intrust_type1_isnull==1 || intrust_type2_isnull==1 || item_num_isnull==1 || item_money_isnull==1 || exp_rate1_isnull==1 || exp_rate2_isnull==1 || intrust_flag3_isnull==1 || intrust_flag4_isnull==1 || entity_type_isnull==1 || deal_type_isnull==1 || managertype_isnull==1 || natrust_type_isnull==1 || custodian_bank_isnull==1 || with_bank_flag_isnull==1 || with_security_flag_isnull==1 || with_private_flag_isnull==1 || with_gov_flag_isnull==1 || ben_period_isnull==1 || intrust_type_isnormal==0 || intrust_type_subitem_isnormal==0 || intrust_type1_isnormal==0 ||//is normal intrust_type2_isnormal==0 || entity_type_isnormal==0 || deal_type_isnormal==0 || natrust_type_isnormal==0 || custodian_bank_isnormal==0 || innovat_type_isnormal==0 || bank_id_isnormal==0 || gov_regional_isnormal==0 || ben_period_isnumber==0//is number ) { check_status=1; } //check is unique? return not unique column var isunique_column=""; if(item_id_isunique==0){isunique_column="item_id";} if(item_code_isunique==0){if(isunique_column==""){isunique_column="item_code";}else{isunique_column=isunique_column+"、"+"item_code";}} if(item_name_isunique==0){if(isunique_column==""){isunique_column="item_name";}else{isunique_column=isunique_column+"、"+"item_name";}} //check is null? return null column var isnull_column=""; if(intrust_flag1_isnull==1){isnull_column="intrust_flag1";} if(intrust_type_isnull==1){if(isnull_column==""){isnull_column="intrust_type";}else{isnull_column=isnull_column+"、"+"intrust_type";}} if(intrust_type_subitem_isnull==1){if(isnull_column==""){isnull_column="intrust_type_subitem";}else{isnull_column=isnull_column+"、"+"intrust_type_subitem";}} if(intrust_type1_isnull==1){if(isnull_column==""){isnull_column="intrust_type1";}else{isnull_column=isnull_column+"、"+"intrust_type1";}} if(intrust_type2_isnull==1){if(isnull_column==""){isnull_column="intrust_type2";}else{isnull_column=isnull_column+"、"+"intrust_type2";}} if(item_num_isnull==1){if(isnull_column==""){isnull_column="item_num";}else{isnull_column=isnull_column+"、"+"item_num";}} if(item_money_isnull==1){if(isnull_column==""){isnull_column="item_money";}else{isnull_column=isnull_column+"、"+"item_money";}} if(exp_rate1_isnull==1){if(isnull_column==""){isnull_column="exp_rate1";}else{isnull_column=isnull_column+"、"+"exp_rate1";}} if(exp_rate2_isnull==1){if(isnull_column==""){isnull_column="exp_rate2";}else{isnull_column=isnull_column+"、"+"exp_rate2";}} if(intrust_flag3_isnull==1){if(isnull_column==""){isnull_column="intrust_flag3";}else{isnull_column=isnull_column+"、"+"intrust_flag3";}} if(intrust_flag4_isnull==1){if(isnull_column==""){isnull_column="intrust_flag4";}else{isnull_column=isnull_column+"、"+"intrust_flag4";}} if(entity_type_isnull==1){if(isnull_column==""){isnull_column="entity_type";}else{isnull_column=isnull_column+"、"+"entity_type";}} if(deal_type_isnull==1){if(isnull_column==""){isnull_column="deal_type";}else{isnull_column=isnull_column+"、"+"deal_type";}} if(managertype_isnull==1){if(isnull_column==""){isnull_column="managertype";}else{isnull_column=isnull_column+"、"+"managertype";}} if(natrust_type_isnull==1){if(isnull_column==""){isnull_column="natrust_type";}else{isnull_column=isnull_column+"、"+"natrust_type";}} if(custodian_bank_isnull==1){if(isnull_column==""){isnull_column="custodian_bank";}else{isnull_column=isnull_column+"、"+"custodian_bank";}} if(with_bank_flag_isnull==1){if(isnull_column==""){isnull_column="with_bank_flag";}else{isnull_column=isnull_column+"、"+"with_bank_flag";}} if(with_security_flag_isnull==1){if(isnull_column==""){isnull_column="with_security_flag";}else{isnull_column=isnull_column+"、"+"with_security_flag";}} if(with_private_flag_isnull==1){if(isnull_column==""){isnull_column="with_private_flag";}else{isnull_column=isnull_column+"、"+"with_private_flag";}} if(with_gov_flag_isnull==1){if(isnull_column==""){isnull_column="with_gov_flag";}else{isnull_column=isnull_column+"、"+"with_gov_flag";}} if(ben_period_isnull==1){if(isnull_column==""){isnull_column="ben_period";}else{isnull_column=isnull_column+"、"+"ben_period";}} //check is normal? return not normal column var isnormal_column=""; if(intrust_type_isnormal==0){isnormal_column="intrust_type";} if(intrust_type_subitem_isnormal==0){if(isnormal_column==""){isnormal_column="intrust_type_subitem";}else{isnormal_column=isnormal_column+"、"+"intrust_type_subitem";}} if(intrust_type1_isnormal==0){if(isnormal_column==""){isnormal_column="intrust_type1";}else{isnormal_column=isnormal_column+"、"+"intrust_type1";}} if(intrust_type2_isnormal==0){if(isnormal_column==""){isnormal_column="intrust_type2";}else{isnormal_column=isnormal_column+"、"+"intrust_type2";}} if(entity_type_isnormal==0){if(isnormal_column==""){isnormal_column="entity_type";}else{isnormal_column=isnormal_column+"、"+"entity_type";}} if(deal_type_isnormal==0){if(isnormal_column==""){isnormal_column="deal_type";}else{isnormal_column=isnormal_column+"、"+"deal_type";}} if(natrust_type_isnormal==0){if(isnormal_column==""){isnormal_column="natrust_type";}else{isnormal_column=isnormal_column+"、"+"natrust_type";}} if(custodian_bank_isnormal==0){if(isnormal_column==""){isnormal_column="custodian_bank";}else{isnormal_column=isnormal_column+"、"+"custodian_bank";}} if(innovat_type_isnormal==0){if(isnormal_column==""){isnormal_column="iinnovat_type";}else{isnormal_column=isnormal_column+"、"+"innovat_type";}} if(bank_id_isnormal==0){if(isnormal_column==""){isnormal_column="bank_id";}else{isnormal_column=isnormal_column+"、"+"bank_id";}} if(gov_regional_isnormal==0){if(isnormal_column==""){isnormal_column="gov_regional";}else{isnormal_column=isnormal_column+"、"+"gov_regional";}} //check is number? return not number column var isnumber_column=""; if(ben_period_isnumber==0) { isnumber_column="ben_period"; } //最終輸出的錯誤詳細情況 if(item_id_isunique==0 || item_code_isunique==0 ||item_name_isunique==0) { check_type="違反唯一規則"; check_detail="表"+source_table+"中,字段"+isunique_column+"違反了唯一規則"; } if ( intrust_flag1_isnull==1 || intrust_type_isnull==1 || intrust_type_subitem_isnull==1 || intrust_type1_isnull==1 || intrust_type2_isnull==1 || item_num_isnull==1 || item_money_isnull==1 || exp_rate1_isnull==1 || exp_rate2_isnull==1|| intrust_flag3_isnull==1 || intrust_flag4_isnull==1 || entity_type_isnull==1 || deal_type_isnull==1 || managertype_isnull==1 || natrust_type_isnull==1 || custodian_bank_isnull==1 || with_bank_flag_isnull==1 || with_security_flag_isnull==1 || with_private_flag_isnull==1 || with_gov_flag_isnull==1 || ben_period_isnull==1 ) { if(check_detail=="") { check_type="違反非空規則"; check_detail="字段"+isnull_column+"違反了非空規則"; } else { check_type=check_type+","+"違反非空規則"; check_detail=check_detail+",字段"+isnull_column+"違反了非空規則"; } } if ( intrust_type_isnormal==0 || intrust_type_subitem_isnormal==0 || intrust_type1_isnormal==0 || intrust_type2_isnormal==0 || entity_type_isnormal==0 || deal_type_isnormal==0 || natrust_type_isnormal==0 || custodian_bank_isnormal==0 || innovat_type_isnormal==0 || bank_id_isnormal==0 || gov_regional_isnormal==0 ) { if(check_detail=="") { check_type="違反標准化規則"; check_detail="字段"+isnormal_column+"違反了標准化規則"; } else { check_type=check_type+","+"違反標准化規則"; check_detail=check_detail+",字段"+isnormal_column+"違反了標准化規則"; } } if(ben_period_isnumber==0) { if(check_detail=="") { check_type="違反數據類型為數字規則"; check_detail="字段"+isnumber_column+"違反了數據類型為數字規則"; } else { check_type=check_type+","+"違反數據類型為數字規則"; check_detail=check_detail+",字段"+isnumber_column+"違反了數據類型為數字規則"; } } if(check_detail!="") { check_detail=check_detail+","+source_table_id+"="+ITEM_ID; } check_name=ITEM_NAME;
校驗日志表結果:
輸出關鍵性指標-表,那些字段違反了那些規則,時間