new file: 01 dm_tf_external_sales.sql
new file: 02 dm_td_external_keycompatitor.sql new file: 03 dm_td_external_brand_market.sql new file: 04 dm_td_external_calendar.sql new file: 05 dm_td_external_exchangerate.sql new file: 07 dm_td_external_packinfo.sql new file: 08 dm_td_external_corp.sql new file: 09 dm_td_external_geo_type.sql new file: 11 DM_TD_EXTERNAL_MARKET_NEW.sql new file: 12 dm_td_external_org.sql new file: 13 external auth.sql new file: 14 dm_tf_external_retail_special_bkp.sql new file: AIA/01 dm_aia_pack_property.sql new file: AIA/02 dm_ext_aia_sales.sql new file: AIA/02 dm_ext_aia_sales_bakup_20230327.sql new file: AIA/03 dm_aia_flag.sql new file: AIA/04 dm_aia_provided_flag.sql new file: AIA/06 DM_TD_EXT_AIA_PACK_PROPERTY.sql new file: AIA/07 DM_TD_EXT_AIA_PACK2MARKET.sql new file: AIA/08 DM_TD_EXT_AIA_MARKET_RATIO.sql new file: AIA/09 DM_TD_EXT_AIA_MARKET_BRAND_RATIO.sql new file: AIA/10 DM_TD_EXT_AIA_MARKET_PACK_MAPPING.sql new file: AIA/11 DM_TD_EXT_AIA_TARGET_INST.sql new file: AIA/11 dm_aia_targethp_flag.sql new file: "AIA/z_03 dm_aia_flag_\345\244\207\344\273\275\347\224\250.sql" new file: CHC/01 dm_chc_pack_property.sql new file: CHC/02 DM_TF_EXT_CHC_SALES.sql new file: CHC/03 DM_TD_EXT_CHC_PACK_PROPERTY.sql new file: CHC/04 DM_TD_EXT_CHC_PACK2MARKET.sql new file: CHC/05 DM_TD_EXT_CHC_MARKET_RATIO.sql new file: CHC/06 DM_TD_EXT_CHC_MARKET_BRAND_RATIO.sql new file: CHC/07 DM_TD_EXT_CHC_MARKET_PACK_MAPPING.sql new file: CHPA/01 dwd_ims_atc_hierarchy.sql new file: CHPA/01 dwd_ims_nfc_hierarchy.sql new file: CHPA/01 dwd_ims_td_manufacturer_corp.sql new file: CHPA/01 dwd_ims_td_pack_property.sql new file: CHPA/01 dwd_update.sql new file: CHPA/01_FB_BLOB_TO_DWD.sql new file: CHPA/02 DWS_IMS_TD_GEO.sql new file: CHPA/02 dws_ims_td_atc_cn.sql new file: CHPA/02 dws_ims_td_corp_cn.sql new file: CHPA/02 dws_ims_td_date.sql new file: CHPA/02 dws_ims_td_manu_cn.sql new file: CHPA/02 dws_ims_td_market.sql new file: CHPA/02 dws_ims_td_market_ta.sql new file: CHPA/02 dws_ims_td_nfc_cn.sql new file: CHPA/02 dws_ims_td_prod_cn.sql new file: CHPA/02 tmp_ims_td_prod_tmp.sql new file: CHPA/02 tmp_ims_tf_fact_sales.sql new file: CHPA/03 DM_TD_EXT_CHPA_MARKET_BRAND_RATIO.sql new file: CHPA/03 DM_TD_EXT_CHPA_MARKET_PACK_MAPPING.sql new file: CHPA/03 DM_TD_EXT_CHPA_MARKET_RATIO.sql new file: CHPA/03 DM_TD_EXT_CHPA_PACK2MARKET.sql new file: CHPA/03 DM_TD_EXT_CHPA_PACK_PROPERTY.sql new file: CHPA/03 DM_TF_EXT_CHPA_SALES.sql new file: CHPA/03 dm_ims_td_calendar.sql new file: CHPA/03 dm_ims_td_geo.sql new file: CHPA/03 dm_ims_td_market_property.sql new file: CHPA/03 dm_ims_td_org.sql new file: CHPA/03 dm_ims_td_org_hvh.sql new file: CHPA/03 dm_ims_td_pack_property.sql new file: CHPA/03 dm_ims_tf_sales.sql new file: CHPA/03 dm_td_chpa_market_definition.sql new file: CHPA/03 dm_td_ims_city_mapping.sql new file: EC/03 ec_load_data.sql new file: EC/04 DM_TD_EXT_EC_PACK_PROPERTY.sql new file: EC/05 DM_TF_EXT_EC_SALES.sql new file: EC/06 DM_TD_EXT_EC_PACK2MARKET.sql new file: EC/07 DM_TD_EXT_EC_MARKET_RATIO.sql new file: EC/08 DM_TD_EXT_EC_MARKET_BRAND_RATIO.sql new file: EC/09 DM_TD_EXT_EC_MARKET_PACK_MAPPING.sql new file: EC/1 (ec)blob_to_dwd.sql new file: EC/2 dwd_inc_gnd_ext_ec_nationnal_pack_union_all.py new file: Merged_Data/Merged_Data_Config_table_bkp.sql new file: Merged_Data/Merged_Data_Config_table_bymonth.sql new file: Merged_Data/dm_tf_exteranl_sales_merged_data_dtp_others_bkp.sql new file: Merged_Data/dm_tf_exteranl_sales_merged_data_dtp_others_bymonth_bkp.sql new file: ORG/DM_TD_EXT_AIA_ORG.sql new file: ORG/DM_TD_EXT_CHC_ORG.sql new file: ORG/DM_TD_EXT_CHPA_ORG.sql new file: ORG/DM_TD_EXT_COUNTY_ORG.sql new file: ORG/DM_TD_EXT_EC_ORG.sql new file: ORG/DM_TD_EXT_RETAIL_ORG.sql new file: ORG/DM_TD_EXT_THC_ORG.sql new file: ORG/DM_TD_EXT_XIEHE_ORG.sql new file: OTHERS/01 dm_td_report_url.sql new file: OTHERS/02 dws_ext_email_warning.sql new file: OTHERS/external_triggered_email.py new file: Retail/01 load_tmp_data.py new file: Retail/02 split_brand_data.py new file: Retail/03 split_pack_data.py new file: Retail/04 map_to_dws_table.py new file: Retail/05 load_dtp_temp_data.py new file: Retail/06 split_dtp_brand_data.py new file: Retail/07 split_dtp_pack_data.py new file: Retail/08 map_to_dtp_dws_table.py new file: Retail/09 dwd_inc_gnd_ext_retail_nataional.py new file: Retail/10 map_to_retail_dm_table.py new file: Retail/11 map_to_overview_dm_table.py new file: Retail/12 dws_tf_external_retail_dtp_special.sql new file: Retail/13 DM_TF_EXT_RETAIL_SALES.sql new file: Retail/14 DM_TF_EXT_RETAIL_DTP_SALES.sql new file: Retail/15 DM_TD_EXT_RETAIL_PACK_PROPERTY.sql new file: Retail/16 DM_TD_EXT_RETAIL_DTP_PACK_PROPERTY.sql new file: Retail/17 DM_TD_EXT_DTP_PACK2MARKET.sql new file: Retail/17 DM_TD_EXT_RETAIL_PACK2MARKET.sql new file: Retail/18 DM_TD_EXT_DTP_MARKET_RATIO.sql new file: Retail/18 DM_TD_EXT_RETAIL_MARKET_RATIO.sql new file: Retail/19 DM_TD_EXT_DTP_MARKET_BRAND_RATIO.sql new file: Retail/19 DM_TD_EXT_RETAIL_MARKET_BRAND_RATIO.sql new file: Retail/20 DM_TD_EXT_DTP_MARKET_PACK_MAPPING.sql new file: Retail/20 DM_TD_EXT_RETAIL_MARKET_PACK_MAPPING.sql new file: "Retail/z1 dwd_inc_gnd_ext_retail_nataional_\344\275\234\345\272\237.py" new file: "Retail/z2 retail_load_data_\344\275\234\345\272\237.sql" new file: "Retail/z3 retail_overview_data_\344\275\234\345\272\237.sql" new file: THC/01 dm_thc_pack_property.sql new file: THC/02 dm_ext_thc_sales.sql new file: THC/02 dm_ext_thc_sales_bakup_20260327.sql new file: THC/03 DM_TF_EXT_THC_MARKET_SALES_CHT.sql new file: THC/04 dm_tf_external_sales_thc.sql new file: THC/05 DM_TD_EXT_THC_PACK_PROPERTY.sql new file: THC/06 DM_TD_EXT_THC_PACK2MARKET.sql new file: THC/07 DM_TD_EXT_THC_MARKET_RATIO.sql new file: THC/08 DM_TD_EXT_THC_MARKET_BRAND_RATIO.sql new file: THC/09 DM_TD_EXT_THC_MARKET_PACK_MAPPING.sql new file: UNIONALL/DM_TD_EXT_UNIONALL_MARKET_PACK_MAPPING.sql new file: UNIONALL/DM_TD_EXT_UNIONALL_PACKINFO.sql new file: UNIONALL/DM_TF_EXT_UNIONALL_MARKET_SALES.sql new file: UNIONALL/DM_TF_EXT_UNIONALL_MARKET_SALES_v1.0.sql new file: UNIONALL/DM_TF_EXT_UNIONALL_SALES.sql new file: UNIONALL/DM_TF_EXT_UNIONALL_SALES_MAPPING.sql new file: UNIONALL/DM_TF_EXT_UNIONALL_SALES_MAPPING_NIAD.sql new file: XIEHE/01 xiehe_blob_to_dwd.py new file: XIEHE/02 dm_xiehe_pack_property.sql new file: XIEHE/03 dm_ext_xiehe_geo.sql new file: XIEHE/04 dm_ext_xiehe_sales.sql new file: XIEHE/05 dm_td_xiehe_core_dept.sql new file: XIEHE/06 DM_TF_EXT_XIEHE_SALES.sql new file: XIEHE/07 DM_TD_EXT_XIEHE_PACK_PROPERTY.sql new file: XIEHE/08 DM_TD_EXT_XIEHE_PACK2MARKET.sql new file: XIEHE/09 DM_TD_EXT_XIEHE_MARKET_RATIO.sql new file: XIEHE/10 DM_TD_EXT_XIEHE_MARKET_BRAND_RATIO.sql new file: XIEHE/11 DM_TD_EXT_XIEHE_MARKET_PACK_MAPPING.sql new file: XIEHE/bkp_01 xiehe_blob2dwd.py new file: XIEHE/bkp_02 dm_ext_xiehe_sales.sql new file: XIEHE/bkp_03 dm_ext_xiehe_pack_property.sql new file: county/01 tmp_ims_county_fact_sales_sum.sql new file: county/02 tmp_imscounty_Result.sql new file: county/03 dm_ims_td_county_geo.sql new file: county/04 dws_ext_county_tf_sales.sql new file: county/05 dm_ext_county_td_pack_property.sql new file: county/06 dm_td_county_pack_region.sql new file: county/07 dm_ext_county_tf_sales_region.sql new file: county/08 DM_TD_EXT_COUNTY_PACK_PROPERTY.sql new file: county/09 DM_TF_EXT_COUNTY_SALES.sql new file: county/10 DM_TD_EXT_COUNTY_PACK2MARKET.sql new file: county/11 DM_TD_EXT_COUNTY_MARKET_RATIO.sql new file: county/12 DM_TD_EXT_COUNTY_MARKET_BRAND_RATIO.sql new file: county/13 DM_TD_EXT_COUNTY_MARKET_PACK_MAPPING.sql new file: for_AIA_Dashboard/01 dm_td_aia_inst_mkt.sql new file: for_AIA_Dashboard/02 dm_td_aia_auth_sales.sql new file: for_AIA_Dashboard/03 dm_td_aia_original_col.sql new file: for_AIA_Dashboard/04 dm_td_aia_nosales_inst.sql new file: for_AIA_Dashboard/05 dm_td_aia_is_eagle_flag.sql new file: for_AIA_Dashboard/06 dm_td_aia_rank.sql new file: for_AIA_Dashboard/07 dm_ext_aia_data_remove_flag.sql new file: for_AIA_Dashboard/07 dm_td_aia_remove_special_ins_bkp.py new file: for_AIA_Dashboard/08 dm_ext_aia_data_quality_flag.sql new file: z 01 dm_tf_external_sales.sql new file: "z 10 dm_td_external_market_pack_mapping_\344\275\234\345\272\237.sql" new file: "z 11 dm_td_external_market_\344\275\234\345\272\237.sql" new file: "\344\270\212\347\272\277\350\204\232\346\234\254.sql" new file: "\346\225\260\346\215\256\351\252\214\350\257\201.sql"
This commit is contained in:
210
CHPA/01_FB_BLOB_TO_DWD.sql
Normal file
210
CHPA/01_FB_BLOB_TO_DWD.sql
Normal file
@@ -0,0 +1,210 @@
|
||||
-- Databricks notebook source
|
||||
-- MAGIC %md
|
||||
-- MAGIC ### 从blob读取csv文件作为chap法伯的事实表
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- 源表：文件
|
||||
-- 目标表：DWD.DWD_GND_PHARBERS_PROV_FACT
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %run ../../../Common/config
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC from datetime import datetime, timedelta
|
||||
-- MAGIC import pandas as pd
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC if ENVIRONMENT == PRD_ENVIRONMENT_VALUE:
|
||||
-- MAGIC factsales_file_path_template = "abfss://master@azcdatalakeprd.dfs.core.chinacloudapi.cn/ODS/GND/UserUpload/"
|
||||
-- MAGIC elif ENVIRONMENT == TEST_ENVIRONMENT_VALUE:
|
||||
-- MAGIC factsales_file_path_template = "abfss://master@retaildlstoragetest.dfs.core.chinacloudapi.cn/ODS/GND/UserUpload/"
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC # 计算时间得到当天的路径
|
||||
-- MAGIC current_date = datetime.utcnow() + timedelta(hours=8)
|
||||
-- MAGIC date_path = current_date.strftime("%Y/%m/%d/")
|
||||
-- MAGIC base_path = factsales_file_path_template + date_path
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC # 路径是否存在
|
||||
-- MAGIC def path_exists(path):
|
||||
-- MAGIC try:
|
||||
-- MAGIC dbutils.fs.ls(path)
|
||||
-- MAGIC return True
|
||||
-- MAGIC except Exception as e:
|
||||
-- MAGIC if "java.io.FileNotFoundException" in str(e):
|
||||
-- MAGIC return False
|
||||
-- MAGIC else:
|
||||
-- MAGIC print(f"检查路径 {path} 时出错: {e}")
|
||||
-- MAGIC raise
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC # 列出blob上的文件列表
|
||||
-- MAGIC def list_file_name(path):
|
||||
-- MAGIC first_path_list = [i.path for i in dbutils.fs.ls(path)]
|
||||
-- MAGIC second_path_list = [dbutils.fs.ls(i)[0] for i in first_path_list ]
|
||||
-- MAGIC return second_path_list
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC # 从blob下载文件到local
|
||||
-- MAGIC def download_file(file_path, local_path):
|
||||
-- MAGIC # dbutils.fs.cp(file_path, local_path.replace("/dbfs", ""))
|
||||
-- MAGIC dbutils.fs.cp(file_path, local_path)
|
||||
-- MAGIC print(f"已下载 {file_path} 到 {local_path}")
|
||||
-- MAGIC return local_path
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %md
|
||||
-- MAGIC ### 获取路径下的文件名称
|
||||
-- MAGIC - 并挑出符合条件的文件路径
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC try:
|
||||
-- MAGIC if path_exists(base_path):
|
||||
-- MAGIC all_file_list = list_file_name(base_path)
|
||||
-- MAGIC # 生成df来筛选内容
|
||||
-- MAGIC files_df = pd.DataFrame([{
|
||||
-- MAGIC 'path':f.path,
|
||||
-- MAGIC 'modificationtime': f.modificationTime,
|
||||
-- MAGIC 'name': f.name
|
||||
-- MAGIC } for f in all_file_list])
|
||||
-- MAGIC print(f"{base_path} 路径存在")
|
||||
-- MAGIC else:
|
||||
-- MAGIC print(f"{base_path} 路径不存在")
|
||||
-- MAGIC except Exception as e:
|
||||
-- MAGIC print(e)
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC try:
|
||||
-- MAGIC files_df = files_df.sort_values('modificationtime', ascending=False).drop_duplicates('name').sort_index()
|
||||
-- MAGIC files_df = files_df[files_df['name'].str.match(r'^Pharbers_PROV_Fact.*\.csv$')]
|
||||
-- MAGIC files_df
|
||||
-- MAGIC except Exception as e:
|
||||
-- MAGIC print(e)
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %md
|
||||
-- MAGIC ### 读取文件内容
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC import os
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC # 下载数据到local
|
||||
-- MAGIC try:
|
||||
-- MAGIC if files_df['path'].tolist():
|
||||
-- MAGIC # 如果列表不为空
|
||||
-- MAGIC df_all = []
|
||||
-- MAGIC for file in files_df['path'].tolist():
|
||||
-- MAGIC local_path = download_file(file, f"/Volumes/{NGBI_CATALOG}/tmp/volume_tmp/tmp/{os.path.basename(file)}")
|
||||
-- MAGIC file_df = (spark.read.option("header", "true").option("quote", '"').option("escape", '"').option("multiLine", "true").option("mode", "PERMISSIVE").csv(local_path))
|
||||
-- MAGIC file_df = file_df.drop("TA",'Market')
|
||||
-- MAGIC print(f'已读取{local_path}')
|
||||
-- MAGIC df_all.append(file_df)
|
||||
-- MAGIC # df_combine = pd.concat(df_all)
|
||||
-- MAGIC df_ifexists = True
|
||||
-- MAGIC else:
|
||||
-- MAGIC print('没有符合条件的文件')
|
||||
-- MAGIC df_ifexists= False
|
||||
-- MAGIC except Exception as e:
|
||||
-- MAGIC print(e)
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %python
|
||||
-- MAGIC try:
|
||||
-- MAGIC if df_ifexists: # 如果不为空
|
||||
-- MAGIC spark.sql(f"TRUNCATE table tmp.tmp_chpa_raw_data")
|
||||
-- MAGIC num = 1
|
||||
-- MAGIC for i in df_all:
|
||||
-- MAGIC i.withColumnRenamed('IMS.药品ID','IMS_DRUG_ID') \
|
||||
-- MAGIC .withColumnRenamed('是否法伯编码','IS_HOSP_CODE') \
|
||||
-- MAGIC .withColumnRenamed('规格','SPEC') \
|
||||
-- MAGIC .withColumnRenamed('转换比','CONVERSION_RATIO') \
|
||||
-- MAGIC .withColumnRenamed('剂型','DOSAGE_FORM') \
|
||||
-- MAGIC .withColumnRenamed('价格','PRICE')
|
||||
-- MAGIC # i.columns = ['area','city','yyyyqq','yyyymm','reimburse','prescription_source','prescription_dept_lv1','prescription_dept_lv2','prescription_dept_lv3','ATC','new_code','common_name','product_name','pack_des','PackSize','PackageType','nfc','manu_des','drug_delivery_route','prescription','sales_vol','sales_value'] #'h_level',
|
||||
-- MAGIC # sdf = spark.createDataFrame(i)
|
||||
-- MAGIC i.createOrReplaceTempView('fact_sales')
|
||||
-- MAGIC spark.sql(f"INSERT into tmp.tmp_chpa_raw_data SELECT * FROM fact_sales")
|
||||
-- MAGIC print(f'第{num}个')
|
||||
-- MAGIC num +=1
|
||||
-- MAGIC except Exception as e:
|
||||
-- MAGIC print(e)
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- MAGIC %md
|
||||
-- MAGIC ### 将读取到的dataframe写入表中
|
||||
|
||||
-- COMMAND ----------
|
||||
|
||||
-- 全量覆盖
|
||||
insert overwrite dwd.dwd_gnd_pharbers_prov_fact
|
||||
select
|
||||
year,
|
||||
ym,
|
||||
province_c,
|
||||
ims_drug_id,
|
||||
is_hosp_code,
|
||||
prod_corp,
|
||||
prod_cod,
|
||||
pack_cod,
|
||||
phcd,
|
||||
prod_des,
|
||||
cmps_des,
|
||||
corp_des,
|
||||
mnfl_cod,
|
||||
prod_des_c,
|
||||
cmps_c,
|
||||
corp_des_c,
|
||||
pack_des,
|
||||
spec,
|
||||
conversion_ratio,
|
||||
dosage_form,
|
||||
atc4_cod,
|
||||
app1_cod,
|
||||
app1_des,
|
||||
app1_des_c,
|
||||
app2_cod,
|
||||
app2_des,
|
||||
app2_des_c,
|
||||
app3_cod,
|
||||
app3_des,
|
||||
app3_des_c,
|
||||
vbp_batch,
|
||||
vbp,
|
||||
value,
|
||||
totalunit,
|
||||
countingunit,
|
||||
price,
|
||||
manu_des,
|
||||
manu_des_c,
|
||||
NULL AS source_file_path,
|
||||
NULL AS source_file_name,
|
||||
from_utc_timestamp(current_timestamp(),'UTC+8') AS etl_insert_dt
|
||||
from tmp.tmp_chpa_raw_data
|
||||
Reference in New Issue
Block a user