chenwu/MarketAnalysis-ETL
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add EXTERNAL SQL scripts

Browse Source
This commit is contained in:
chenwu
2026-04-20 14:55:25 +08:00
parent c05ba7ec7e
commit b4979eed82
344 changed files with 61619 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

326
EXTERNAL/THC/06 DM_TD_EXT_THC_PACK2MARKET.sql Normal file
View File

@@ -0,0 +1,326 @@
-- Databricks notebook source
-- CREATE OR REPLACE TABLE DM.DM_TD_EXT_THC_PACK2MARKET (
-- MARKET STRING,
-- PACK_CODE STRING,
-- PACK_DESC STRING,
-- STGH_DESC STRING,
-- PACK_LCH STRING,
-- PROD_CODE STRING,
-- CMPS_CODE STRING,
-- CMPS_DESC STRING,
-- ATC1_CODE STRING,
-- ATC2_CODE STRING,
-- ATC3_CODE STRING,
-- ATC4_CODE STRING,
-- APP1_CODE STRING,
-- APP2_CODE STRING,
-- APP3_CODE STRING,
-- BIO_DESC STRING,
-- GENE_ORIG_DESC STRING,
-- ETH_OTC_DESC STRING,
-- NRDL_DESC STRING,
-- NRDL_ENTRY_DATE STRING,
-- EDL_DESC STRING,
-- TCM_DESC STRING,
-- PAED_DESC STRING,
-- GQCE_DESC STRING,
-- VBP_DESC STRING,
-- MANU_CODE STRING,
-- MANU_DESC STRING,
-- MNFL_CODE STRING,
-- MNFL_DESC STRING,
-- CORP_CODE STRING,
-- CORP_DESC STRING,
-- BRANDTYPE STRING,
-- BU STRING,
-- STARTTIME STRING,
-- ENDTIME STRING,
-- MARKET_RATIO STRING,
-- ETL_INSERT_DT TIMESTAMP,
-- ETL_UPDATE_DT TIMESTAMP)
-- USING delta
-- -- LOCATION 'abfss://master@azcdatalakeprd.dfs.core.chinacloudapi.cn/DM/dm_td_ext_thc_pack2market';
-- -- 上面是生产环境location，下面是测试环境location
-- LOCATION 'abfss://master@retaildlstoragetest.dfs.core.chinacloudapi.cn/DM/dm_td_ext_thc_pack2market';
-- COMMAND ----------
-- CREATE OR REPLACE TABLE DM.DM_TD_EXT_thc_PACK2MARKET_KEYCOMPETITOR (
-- KEY_COMPETITOR STRING,
-- MARKET STRING,
-- PACK_CODE STRING)
-- USING delta
-- -- LOCATION 'abfss://master@azcdatalakeprd.dfs.core.chinacloudapi.cn/DM/dm_td_ext_thc_pack2market_keycompetitor';
-- -- 上面是生产环境location，下面是测试环境location
-- LOCATION 'abfss://master@retaildlstoragetest.dfs.core.chinacloudapi.cn/DM/dm_td_ext_thc_pack2market_keycompetitor';
-- COMMAND ----------
-- CREATE or REPLACE TABLE DM.DM_TD_EXT_thc_PACK2MARKET_CLASS (
-- MARKET STRING,
-- PACK_CODE STRING,
-- CLASS STRING)
-- USING delta
-- -- LOCATION 'abfss://master@azcdatalakeprd.dfs.core.chinacloudapi.cn/DM/dm_td_ext_thc_pack2market_class';
-- -- 上面是生产环境location，下面是测试环境location
-- LOCATION 'abfss://master@retaildlstoragetest.dfs.core.chinacloudapi.cn/DM/dm_td_ext_thc_pack2market_class';
-- COMMAND ----------
-- 使用 EXCEPT（集合差操作）简化逻辑
INSERT OVERWRITE TABLE DM.DM_TD_EXT_THC_PACK2MARKET
-- 包含规则
SELECT DISTINCT
T2.MARKET,
T1.PACK_CODE,
T1.PACK_DESC,
T1.STGH_DESC,
T1.PACK_LCH,
T1.PROD_CODE,
T1.CMPS_CODE,
T1.CMPS_DESC,
T1.ATC1_CODE,
T1.ATC2_CODE,
T1.ATC3_CODE,
T1.ATC4_CODE,
T1.APP1_CODE,
T1.APP2_CODE,
T1.APP3_CODE,
T1.BIO_DESC,
T1.GENE_ORIG_DESC AS GENE_ORIG,
T1.ETH_OTC_DESC AS ETH_OTC_D,
T1.NRDL_DESC,
T1.NRDL_ENTRY_DATE AS NRDL_ENTR,
T1.EDL_DESC,
T1.TCM_DESC,
T1.PAED_DESC,
T1.GQCE_DESC,
T1.VBP_DESC,
T1.MANU_CODE,
T1.MANU_DESC,
T1.MNFL_CODE,
T1.MNFL_DESC,
T1.CORP_CODE,
T1.CORP_DESC,
T1.BRANDTYPE,
T2.BU,
T2.STARTTIME,
T2.ENDTIME,
COALESCE(T2.EXTEND_MARKET_RATIO, '1') AS MARKET_RATIO,
FROM_UTC_TIMESTAMP(CURRENT_TIMESTAMP(),'UTC+8') AS ETL_INSERT_DT,
FROM_UTC_TIMESTAMP(CURRENT_TIMESTAMP(),'UTC+8') AS ETL_UPDATE_DT
FROM DM.DM_TD_EXT_THC_PACK_PROPERTY AS T1
LEFT JOIN (
SELECT
MARKET,
BU,
STARTTIME,
ENDTIME,
EXTEND_MARKET_RATIO,
ATC1_CODE,
ATC2_CODE,
ATC3_CODE,
ATC4_CODE,
NFC1_CODE,
NFC2_CODE,
NFC3_CODE,
CORPORATION_CODE,
MANUFACTURER_CODE,
PRODUCT_CODE,
PACK_CODE,
STRENGTH,
MOLECULE_CODE
FROM DWD.DWD_GND_IMS_TBLMARKET_BYMONTH
WHERE EXTEND_MARKET IS NULL AND (NOT_IN_FLAG IS NULL OR NOT_IN_FLAG = '1')
) AS T2
ON (T2.ATC1_CODE IS NULL OR T1.ATC1_CODE = T2.ATC1_CODE)
AND (T2.ATC2_CODE IS NULL OR T1.ATC2_CODE = T2.ATC2_CODE)
AND (T2.ATC3_CODE IS NULL OR T1.ATC3_CODE = T2.ATC3_CODE)
AND (T2.ATC4_CODE IS NULL OR T1.ATC4_CODE = T2.ATC4_CODE)
AND (T2.NFC1_CODE IS NULL OR T1.APP1_CODE = T2.NFC1_CODE)
AND (T2.NFC2_CODE IS NULL OR T1.APP2_CODE = T2.NFC2_CODE)
AND (T2.NFC3_CODE IS NULL OR T1.APP3_CODE = T2.NFC3_CODE)
AND (T2.CORPORATION_CODE IS NULL OR T1.CORP_CODE = T2.CORPORATION_CODE)
AND (T2.MANUFACTURER_CODE IS NULL OR T1.MANU_CODE = T2.MANUFACTURER_CODE)
AND (T2.PRODUCT_CODE IS NULL OR T1.PROD_CODE = T2.PRODUCT_CODE)
AND (T2.PACK_CODE IS NULL OR T1.PACK_CODE = T2.PACK_CODE)
AND (T2.STRENGTH IS NULL OR T1.STGH_DESC = T2.STRENGTH)
AND (T2.MOLECULE_CODE IS NULL OR T1.CMPS_CODE = T2.MOLECULE_CODE)
WHERE T2.MARKET IS NOT NULL
EXCEPT
-- 排除规则
SELECT DISTINCT
T2.MARKET,
T1.PACK_CODE,
T1.PACK_DESC,
T1.STGH_DESC,
T1.PACK_LCH,
T1.PROD_CODE,
T1.CMPS_CODE,
T1.CMPS_DESC,
T1.ATC1_CODE,
T1.ATC2_CODE,
T1.ATC3_CODE,
T1.ATC4_CODE,
T1.APP1_CODE,
T1.APP2_CODE,
T1.APP3_CODE,
T1.BIO_DESC,
T1.GENE_ORIG_DESC AS GENE_ORIG,
T1.ETH_OTC_DESC AS ETH_OTC_D,
T1.NRDL_DESC,
T1.NRDL_ENTRY_DATE AS NRDL_ENTR,
T1.EDL_DESC,
T1.TCM_DESC,
T1.PAED_DESC,
T1.GQCE_DESC,
T1.VBP_DESC,
T1.MANU_CODE,
T1.MANU_DESC,
T1.MNFL_CODE,
T1.MNFL_DESC,
T1.CORP_CODE,
T1.CORP_DESC,
T1.BRANDTYPE,
T2.BU,
T2.STARTTIME,
T2.ENDTIME,
COALESCE(T2.EXTEND_MARKET_RATIO, '1') AS MARKET_RATIO,
FROM_UTC_TIMESTAMP(CURRENT_TIMESTAMP(),'UTC+8') AS ETL_INSERT_DT,
FROM_UTC_TIMESTAMP(CURRENT_TIMESTAMP(),'UTC+8') AS ETL_UPDATE_DT
FROM DM.DM_TD_EXT_THC_PACK_PROPERTY AS T1
LEFT JOIN (
SELECT
MARKET,
BU,
STARTTIME,
ENDTIME,
EXTEND_MARKET_RATIO,
ATC1_CODE,
ATC2_CODE,
ATC3_CODE,
ATC4_CODE,
NFC1_CODE,
NFC2_CODE,
NFC3_CODE,
CORPORATION_CODE,
MANUFACTURER_CODE,
PRODUCT_CODE,
PACK_CODE,
STRENGTH,
MOLECULE_CODE
FROM DWD.DWD_GND_IMS_TBLMARKET_BYMONTH
WHERE EXTEND_MARKET IS NULL AND NOT_IN_FLAG = '0'
) AS T2
ON (T2.ATC1_CODE IS NULL OR T1.ATC1_CODE = T2.ATC1_CODE)
AND (T2.ATC2_CODE IS NULL OR T1.ATC2_CODE = T2.ATC2_CODE)
AND (T2.ATC3_CODE IS NULL OR T1.ATC3_CODE = T2.ATC3_CODE)
AND (T2.ATC4_CODE IS NULL OR T1.ATC4_CODE = T2.ATC4_CODE)
AND (T2.NFC1_CODE IS NULL OR T1.APP1_CODE = T2.NFC1_CODE)
AND (T2.NFC2_CODE IS NULL OR T1.APP2_CODE = T2.NFC2_CODE)
AND (T2.NFC3_CODE IS NULL OR T1.APP3_CODE = T2.NFC3_CODE)
AND (T2.CORPORATION_CODE IS NULL OR T1.CORP_CODE = T2.CORPORATION_CODE)
AND (T2.MANUFACTURER_CODE IS NULL OR T1.MANU_CODE = T2.MANUFACTURER_CODE)
AND (T2.PRODUCT_CODE IS NULL OR T1.PROD_CODE = T2.PRODUCT_CODE)
AND (T2.PACK_CODE IS NULL OR T1.PACK_CODE = T2.PACK_CODE)
AND (T2.STRENGTH IS NULL OR T1.STGH_DESC = T2.STRENGTH)
AND (T2.MOLECULE_CODE IS NULL OR T1.CMPS_CODE = T2.MOLECULE_CODE)
WHERE T2.MARKET IS NOT NULL;
-- COMMAND ----------
INSERT OVERWRITE TABLE DM.DM_TD_EXT_thc_PACK2MARKET_KEYCOMPETITOR
WITH kc AS (
SELECT
keycompetitor, no,
ROW_NUMBER() OVER (ORDER BY
CASE
WHEN ATC1_Code IS NOT NULL THEN 1
WHEN ATC2_Code IS NOT NULL THEN 2
WHEN ATC3_Code IS NOT NULL THEN 3
WHEN ATC4_Code IS NOT NULL THEN 4
WHEN Molecule_Code IS NOT NULL THEN 5
WHEN Product_Code IS NOT NULL THEN 6
WHEN Pack_Code IS NOT NULL THEN 7
ELSE 999
END,
CASE WHEN NFC1_CODE IS NULL THEN 0 ELSE 1 END,
CASE WHEN NFC2_CODE IS NULL THEN 0 ELSE 1 END,
CASE WHEN NFC3_CODE IS NULL THEN 0 ELSE 1 END
) AS no1,
ATC1_Code, ATC2_Code, ATC3_Code, ATC4_Code,
NFC1_Code, NFC2_Code, NFC3_Code,
corporation_code, Manufacturer_Code,
Product_Code, Pack_Code, Strength, Molecule_Code, market
FROM dwd.dwd_gnd_ims_tblkeycompetitor
),
joined AS (
SELECT
/*+ BROADCAST(kc) */
COALESCE(kc.keycompetitor, 'OTHERS') AS Key_Competitor, -- ✅ UPDATE 逻辑内联
t1.market, t1.PACK_CODE
/*bug 调试区
,
kc.no,
kc.no1,
t1.PROD_COD, t1.CMPS_COD, t1.CMPS_DES,
t1.ATC1_COD, t1.ATC2_COD, t1.ATC3_COD, t1.ATC4_COD,
t1.APP1_COD, t1.APP2_COD, t1.APP3_COD,
t1.BIO_DESC, t1.GENE_ORIG_DESC, t1.ETH_OTC_DESC,
t1.NRDL_DESC, t1.NRDL_Entry_Date, t1.EDL_DESC, t1.TCM_DESC,
t1.PAED_DESC, t1.GQCE_DESC, t1.VBP_DESC,
t1.MANU_COD, t1.MANU_DES, t1.MNFL_COD, t1.MNFL_DES,
t1.CORP_COD, t1.CORP_DES, t1.BrandType, t1.bu, t1.Market_Ratio
*/
,ROW_NUMBER() OVER (
PARTITION BY t1.market, t1.PACK_CODE, t1.PROD_CODE
ORDER BY kc.no1 DESC, kc.no DESC
) AS _rn
FROM DM.DM_TD_EXT_thc_PACK2MARKET t1
LEFT JOIN kc
ON (kc.ATC1_Code IS NULL OR t1.ATC1_CODE = kc.ATC1_Code)
AND (kc.ATC2_Code IS NULL OR t1.ATC2_CODE = kc.ATC2_Code)
AND (kc.ATC3_Code IS NULL OR t1.ATC3_CODE = kc.ATC3_Code)
AND (kc.ATC4_Code IS NULL OR t1.ATC4_CODE = kc.ATC4_Code)
AND (kc.NFC1_Code IS NULL OR t1.APP1_CODE = kc.NFC1_Code)
AND (kc.NFC2_Code IS NULL OR t1.APP2_CODE = kc.NFC2_Code)
AND (kc.NFC3_Code IS NULL OR t1.APP3_CODE = kc.NFC3_Code)
AND (kc.corporation_code IS NULL OR t1.CORP_CODE = kc.corporation_code)
AND (kc.Manufacturer_Code IS NULL OR t1.MANU_CODE = kc.Manufacturer_Code)
AND (kc.Product_Code IS NULL OR t1.PROD_CODE = kc.Product_Code)
AND (kc.Pack_Code IS NULL OR t1.PACK_CODE = kc.Pack_Code)
AND (kc.Strength IS NULL OR t1.STGH_DESC = kc.Strength)
AND (kc.Molecule_Code IS NULL OR t1.CMPS_CODE = kc.Molecule_Code)
AND (kc.market IS NULL OR t1.market = kc.market)
)
SELECT distinct Key_Competitor,market,pack_code
FROM joined
WHERE _rn = 1;
-- COMMAND ----------
INSERT OVERWRITE TABLE DM.DM_TD_EXT_thc_PACK2MARKET_CLASS
SELECT
/*+ BROADCAST(T2) */
DISTINCT
T1.MARKET,
T1.PACK_CODE,
COALESCE( T2.CLASS, 'OTHERS') AS CLASS
FROM DM.DM_TD_EXT_thc_PACK2MARKET T1
LEFT JOIN DWD.DWD_GND_TBLCLASS T2
ON (T2.MARKET IS NULL OR T1.MARKET = T2.MARKET)
AND (T2.PACK_CODE IS NULL OR T1.PACK_CODE = T2.PACK_CODE)
AND (T2.PRODUCT_CODE IS NULL OR T1.PROD_CODE = T2.PRODUCT_CODE)
AND (T2.MOLECULE_CODE IS NULL OR T1.CMPS_CODE = T2.MOLECULE_CODE)
AND (T2.CORPORATION_CODE IS NULL OR T1.CORP_CODE = T2.CORPORATION_CODE)
AND (T2.MANUFACTURER_CODE IS NULL OR T1.MANU_CODE = T2.MANUFACTURER_CODE)
AND (T2.ATC1_CODE IS NULL OR T1.ATC1_CODE = T2.ATC1_CODE)
AND (T2.ATC2_CODE IS NULL OR T1.ATC2_CODE = T2.ATC2_CODE)
AND (T2.ATC3_CODE IS NULL OR T1.ATC3_CODE = T2.ATC3_CODE)
AND (T2.ATC4_CODE IS NULL OR T1.ATC4_CODE = T2.ATC4_CODE)
AND (T2.NFC1_CODE IS NULL OR T1.APP1_CODE = T2.NFC1_CODE)
AND (T2.NFC2_CODE IS NULL OR T1.APP2_CODE = T2.NFC2_CODE)
AND (T2.NFC3_CODE IS NULL OR T1.APP3_CODE = T2.NFC3_CODE)
AND (T2.STRENGTH IS NULL OR T1.STGH_DESC = T2.STRENGTH)