Solution-A:

This solution is along what you tried to do.

items = dom.findall('Items/Item')
# use a '@' before any attribute in the list of target columns. This will help 
# us treat the attributes separately from the regular tags.
target_cols = ['PartNumber', 'BrandAAIAID', 'BrandLabel', 'ItemLevelGTIN', '@GTINQualifier']

for item in items:
    instance_dict = dict()
    for col in target_cols:
        label = col.replace("@","")
        if col.startswith("@"):
            instance_dict.update({col: item.find('.//*[@{}]'.format(label)).attrib[label]})
        else:
            instance_dict.update({col: item.find(label).text})

    val = tuple(instance_dict[col_name] for col_name in target_cols)
    print(instance_dict)
    # **write to db here**

Solution-B:

Here is a solution for reading the xml data in as a dataframe/dict. And then you could grab the fields that you need to write to the db.

Let's make some data

I introduced a dummy second <Item></Item> tag to check if this processes multiple tags properly.

xml_string = """
<PARTS>
    <Header>
        <Version>6.5</Version>
    </Header>
    <Items>
        <Item MaintenanceType="A">
            <HazardousMaterialCode>N</HazardousMaterialCode>
            <ItemLevelGTIN GTINQualifier="UP">00651860733074</ItemLevelGTIN>
            <PartNumber>14-230615</PartNumber>
            <BrandAAIAID>BBGL</BrandAAIAID>
            <BrandLabel>Bilstein</BrandLabel>
            <ACESApplications>Y</ACESApplications>
            <ItemQuantitySize UOM="EA">1.0</ItemQuantitySize>
            <ContainerType>BX</ContainerType>
            <QuantityPerApplication Qualifier="MAX" UOM="EA">1</QuantityPerApplication>
        </Item>
        <Item MaintenanceType="B">
            <HazardousMaterialCode>N</HazardousMaterialCode>
            <ItemLevelGTIN GTINQualifier="UP">00651860733084</ItemLevelGTIN>
            <PartNumber>14-230620</PartNumber>
            <BrandAAIAID>BBGL</BrandAAIAID>
            <BrandLabel>BilsteinZ</BrandLabel>
            <ACESApplications>Y</ACESApplications>
            <ItemQuantitySize UOM="EA">1.0</ItemQuantitySize>
            <ContainerType>BX</ContainerType>
            <QuantityPerApplication Qualifier="MAX" UOM="EA">1</QuantityPerApplication>
        </Item>
    </Items>
</PARTS>
"""

Solution

Optionally use pandas to see the entire xml as a dataframe.

Import Libraries

import pandas as pd
from xml.etree import ElementTree as ET
import mysql.connector

Now recursively read each tag and write to the db.

target_cols = ['PartNumber', 'BrandAAIAID', 'BrandLabel', 'ItemLevelGTIN', 'GTINQualifier']

sql = "INSERT INTO scoreitem (B15_PartNumber, B20_BrandAAIAID, B25_BrandLabel, B10_ItemLevelGTIN, " \
      "B11_GTINQualifier) VALUES (%s, %s, %s, %s, %s)"

dict_cols = [None] # (either a list with a None) or (is equal to target_cols)

write_to_db = False # Set this to true when you write to db

file_name = 'bsn.xml'

# Let us test with the xml_string first
# set 'xml_source_is_file = True' 
# when working with a file.
xml_source_is_file = False
if xml_source_is_file:
    dom = ET.parse(file_name)
else:
    dom = ET.fromstring(xml_string)

if write_to_db:
    mydb = mysql.connector.connect(user='frank', password='xxxxxx', 
                                   host='127.0.0.1',         
                                   database='scoresre', 
                                   auth_plugin='mysql_native_password')
    mycursor = mydb.cursor()

consolidated_dict = dict()

for xx in list(dom):
    if xx.tag == 'Items':
        #print(list(xx))
        for i, item in enumerate(xx):
            instance_dict = dict()
            #print(list(item))
            for ee in list(item):
                #print(ee)
                kee = ee.tag.replace('@','')
                if (kee in dict_cols) or (dict_cols[0] is None):
                    instance_dict.update({kee: ee.text})
                if isinstance(ee, dict):                    
                    for e in list(ee):
                        #print(e)
                        ke = e.tag.replace('@','')
                        if (ke in dict_cols) or (dict_cols[0] is None):
                            instance_dict.update({ke: e.text})
                        temp_dict = e.attrib
                        if len(temp_dict) > 0:
                            for jj in temp_dict.keys():
                                kjj = jj.replace('@','')
                                if (kjj in dict_cols) or (dict_cols[0] is None):
                                    instance_dict.update({kjj: temp_dict.get(jj)})
                temp_dict = ee.attrib
                if len(temp_dict) > 0:
                    for jj in temp_dict.keys():
                        kjj = jj.replace('@','')
                        if (kjj in dict_cols) or (dict_cols[0] is None):
                            instance_dict.update({kjj: temp_dict.get(jj)})                
            #print(instance_dict)
            consolidated_dict.update({i: instance_dict})
            val = tuple(instance_dict[col_name] for col_name in target_cols)
            print(val)
            # Write to db here
            if write_to_db:
                mycursor.execute(sql, val)
                mydb.commit()

df = pd.DataFrame(consolidated_dict).T
df

PS: Note that there are two switches, write_to_db and xml_source_is_file. In your case, you would need to set both of them as True to write to the database and read data in from the xml file.