SQL Order By Within A Count(Distinct) - sql

Question

I have the following tables:
filetype1
F1_ID | F1_ORDR | FILENAME
1 | 1 | file1.txt
2 | 2 | file2.txt
3 | 3 | file3.txt
4 | 2 | file4.txt
5 | 4 | file5.txt
filetype2
F2_ID | F2_ORDR | FILENAME
1 | 1 | file6.txt
2 | 2 | file7.txt
3 | 4 | file8.txt
ordr
OR_ID | OR_VENDOR
1 | 1
2 | 1
3 | 1
4 | 1
vendor
VE_ID | VE_NAME
1 | Company1
My goal is to have a list of vendors and a count of the number of orders where a file is connected for each type. For example, the end result of this data should be:
VENDOR | OR_CT | F1_CT | F2_CT
Company1 | 4 | 4 | 3
Because at least 1 type1 file was attached to 4 distinct orders and at least 1 type2 file was attached to 3 distinct orders. Currently my SQL code looks like this:
SELECT vendor.ve_id, vendor.ve_name,
(SELECT COUNT(or_id)
FROM ordr
WHERE ordr.or_vendor = vendor.ve_id) as OR_COUNT,
(SELECT COUNT(DISTINCT f1_order)
FROM filetype1 INNER JOIN ordr ON filetype1.f1_ordr = ordr.or_id
WHERE ordr.or_vendor = vendor.ve_id) as F1_CT,
(SELECT COUNT(DISTINCT f2_ordr)
FROM filetype2 INNER JOIN ordr ON filetype2.f2_ordr = ordr.or_id
WHERE ordr.or_vendor = vendor.ve_id) as F2_CT
FROM vendor
ORDER BY vendor.ve_name
Unfortunately this yields the following results:
VENDOR | OR_COUNT | F1_COUNT | F2_COUNT
Company1 | 4 | 5 | 3
My only guess is that because I'm using COUNT(DISTINCT) the COUNT is automatically assuming the DISTINCT is ordering by F1_ID instead of by F1_ORDR
If anyone can assist me on how to tell the COUNT(DISTINCT) to order by F1_ORDR that would be most helpful. I have searched the vast internet for a solution but its hard to explain what I want to a search engine, forums, etc. My database uses Microsoft SQL Server. My knowledge of database management is almost completely self taught, so I'm just glad I made it this far on my own. My expertise is in web design. Thank you for your time.

Answer 1

Your SQL yields the result you want for me.
Two pieces of advice
Order is a bad name for a table - it conflicts with a reserved word, and will cause you no end of hassle
You should join your tables like so
FROM filetype1
inner join [order]
on filetype1.f1_order = or_id
rather than using a where clause
Perhaps try this instead
select
vendor.VE_ID, vendor.VE_NAME,
count(distinct or_id),
count(distinct f1_order),
count(distinct f2_order)
from
vendor
left join [order]
on vendor.VE_ID = [order].OR_VENDOR
inner join filetype1
on [order].OR_ID = filetype1.F1_ORDER
left join filetype2
on [order].OR_ID = filetype2.F2_ORDER
group by
vendor.VE_ID, vendor.VE_NAME

Answer 2

Try this:
SELECT
vdr.VE_NAME
,COUNT(DISTINCT OR_ID) AS OR_ID
,COUNT(DISTINCT ft1.F1_ORDER) AS FT1_COUNT
,COUNT(DISTINCT ft2.F2_ORDER) AS FT2_COUNT
FROM
vendor vdr
LEFT OUTER JOIN [order] odr
ON vdr.VE_ID = odr.OR_VENDOR
INNER JOIN filetype1 ft1
ON odr.OR_ID = ft1.F1_ORDER
LEFT OUTER JOIN filetype2 ft2
ON odr.OR_ID = ft2.F2_ORDER
GROUP BY
vdr.VE_ID
,vdr.VE_NAME

Answer 3

I will propose you this:
Merge filetype1 and filetype2 tables in one table(filetype) and add another field named - f_type(for instance) of type INT or TINTYINT to store the filetype (1 or 2). This has the benefits of painlessly adding another filetype later
Now the query will look something like this:
SELECT
vendor.ve_name,
count(DISTINCT filetype.f_order),
filetype.f_type
FROM
filetype
INNER JOIN `order`
ON filetype.f_order = `order`.or_id
INNER JOIN vendor
ON `order`.or_vendor = vendor.ve_id
GROUP BY vendor.ve_id,filetype.f_type
This will give the count of orders for filetype.
For the total orders just add another query:
SELECT count(*) FROM `order`