Bulk data import using csv to table

Need a module which import the csv to sqlalchemy table 1. the csv needs to be validated there are some validation checks that needs to be done 2. the error records needs to be separated and create a new csv file which mentions the error with respect to the record 3. The valid records needs to be inserted to the table 4. The code should be scalable upto 10,000 records The validation checks 1. Asset ID should be category short name + random 8 digits - flag error
 2. Asset ID should not be present in the table - flag error
 3. asset ID is not part of CSV file a new ASSET ID is generated and new record is created, this assumes asset category exists - (SHORTNAME + RANDOM 8 DIGITS)
 4. asset ID is not provided and asset category is also not provided - flag error
 5. category does not exists(CATEGORY_MAPPING) - flag error
 6. asset ID is provided and category is not provided - flag error
 7. If the required field value is NULL - flag error The assetId is derived from category 'CATEGORY SHORT NAME MAPPING' + random 8 digits the 1st 2 characters in asset ID and 'CATEGORY SHORT NAME MAPPING' should match CATEGORY SHORT NAME MAPPING - 
{ “CABINET”:”CA”, “ROW”:”RO” “RACK”:AC” } Mapping the headers - { “Name”:”STD1”, “Asset Id”:”STD2”, “Description”:”STD3”, “Tags”:”STD4” } Mapping fields to properties - { “STD1”: { “fieldEnabled”:0, “fieldMandatory”:1, “fieldType”:”SELECT”, “fieldValue”:”CATEGORY_MAPPING” }, “STD2”: { “fieldEnabled”:0, “fieldType”:”ENUM”, “fieldValue”:”statusEnum” “fieldMandatory”:1, “default”:”ACTIVE” } } CATEGORY_MAPPING- 
[ “CABINET”:1, “ROW”:2 “RACK”:3 ] statusEnum - 
[‘active’,’inactive’] This should be blueprint and also should be able to integrate with existing project