T-SQL trouble with conditional select into - sql-server

Question

I'm running into some problems with my attempts to create a specific table. I have read some similar answers on this site already, but I'm not sure I'm understanding their answers and/or how it applies to my issue.
I have an existing table which contains ingredients for a recipe (table definition below):
Create Table Recipes.dbo.IngredientsByRecipeID
(
RecipeID int,
IngredientName varchar(max),
UnitTypeOfIngredient varchar(max),
QuantityOfIngredient decimal(6,3)
);
I'm trying to create and display the contents of a table which contains all ingredient names used in all recipes, in alphabetical order, and with all duplicate entries removed.
I've tried a few different variations on how to implement it, but I'm not sure now if I'm running into problems with my implementation or problems with how SQL Server works (I'm new to T-SQL and its ways).
My method for sorting the ingredients alphabetically into a new, duplicates-free table is as follows:
/*these three lines represent code which I've been trying to use to control the problems I keep getting with tables existing or not existing, I don't know if they are really helping or not. The first two only drop and create the table, the second is one attempt at a working solution for my needs*/
--drop table AlphabeticallySortedIngredientsList;
--create table AlphabeticallySortedIngredientsList (IngredientName varchar(max));
--select IngredientName into AlphabeticallySortedIngredientsList from IngredientsByRecipeID where ((select count(IngredientName) from AlphabeticallySortedIngredientsList) = 0) order by IngredientName asc;
--here is my latest attempt at getting a working solution for my needs
select IngredientName
into AlphabeticallySortedIngredientsList
from IngredientsByRecipeID
where not exists (select *
from AlphabeticallySortedIngredientsList
where IngredientName = AlphabeticallySortedIngredientsList.IngredientName)
order by IngredientName asc;
With each implementation I initially had some kind of syntax errors, but after (apparently) cleaning up those errors, I'm always left with one last error and while the error changes based on input, the same object is causing the errors. If I leave the 'drop table' line from above uncommented then this is the error I get:
Msg 208, Level 16, State 1, Line 5
Invalid object name 'AlphabeticallySortedIngredientsList'.
If I also uncomment the 'create table AlphabeticallySortedIngredientsList' line, then I get this error:
Msg 2714, Level 16, State 6, Line 9
There is already an object named 'AlphabeticallySortedIngredientsList' in the database.
And if I drop the table and then comment both lines out, I get the same error as above:
Msg 2714, Level 16, State 6, Line 9
There is already an object named 'AlphabeticallySortedIngredientsList' in the database.
I'm perplexed as it seems that my 'select IngredientName into AlphabeticallySortedIngredientList' line cannot make up its mind as to whether or not it is creating the table. If I drop the table before my solution runs, then it says it cannot be found. If I create the table before my solution runs, then it says the table already exists.
I'm guessing the problem is that the 'select' part of the statement is creating AlphabeticallySortedIngredientsList (given it doesn't exist) but then the conditional tries to look for a value in AlphabeticallySortedIngredientsList, which does not exist yet as the operation to create it hasn't occurred yet. Am I on the right track with this assessment, and if so, how can I fix the problem?
I need to preserve the original contents of IngredientsByRecipeID, for the record. AlphabeticallySortedIngredientList only serves the purpose of presenting a clean and orderly list of all ingredients from IngredientsByRecipeID.
I should also mention that if I remove the conditional (as in, I comment out the 'where' clause), then the select into works without any errors. This is with both drop table and create table commented out, and with no table AlphabeticallySortedIngredientList already existing in the database.
Maybe it would be better to just add them all to the table without deleting duplicates, and then go through and delete duplicates? My problem might be with trying to do it all in one step.

Answer 1

The SELECT/INTO syntax creates a new table. You cannot use this and reference the table you are creating in the WHERE clause. Want you want to do is use DISTINCT on the SELECT clause to remove duplicates before you create the table. If you need to use an existing table then you want to use an INSERT statement where the values to insert come from a SELECT clause. Note, if you need to keep the table order by name, it should have the name as the primary key. Since you want unique entries on this, you probably want this anyway.
SELECT DISTINCT IngredientName
INTO AlphabeticallySortedIngredientsList
FROM IngredientsByRecipeID
ADD CONSTRAINT PK_AlphabeticallySortedIngredientsList PRIMARY KEY CLUSTERED (IncredientName)
Generally, I would not suggest keeping this table as you can always get the data from the existing table. Keeping the data in separate tables will force you to periodically update one or the other to keep them in sync. If you do opt to keep the names in a separate table, then you should probably add an Id column, then update the other (original) table with the Id of the name and remove the name column from it. Add a foreign key constraint on the name Id column so that you are required to add the name there and link it to the recipe.

Answer 2

Create table Ingredients
(ingredient varChar(50) primary key not null)
Insert Ingredients(Ingredient)
Select distinct IngredientName
From IngredientsByRecipeID

Answer 3

Yes you are on the right track. The SELECT ... INTO statement will create the table based on the SELECT expression. Trying to access it prematurely in the WHERE NOT EXISTS clause is causing the conflict.
Trying using a SELECT DISTINCT or GROUP BY clause to remove duplicates without relying on WHERE NOT EXISTS.