With my previous article about getting into REST applications, I figured it would be a good idea for me to set up a data source so I could practice. Besides, I had reinstalled SQL Server 2019 on my machine, and I needed to import some data so that I could brush up on my SQL skills as well.
Being the baseball nut that I am, of course, I had to import baseball statistics, so I decided to reimport the most recent data from Sean Lahman’s baseball database. The last time I did this exercise, I downloaded a database format. I don’t remember what format I used (the links all say “Access” — which I don’t remember downloading), but the files I used had an .sql extension. This time, I used the comma-delimited version, which downloaded a zip file containing files with a .csv extension.
I wanted to import the files directly into my database and have them create the tables upon doing so, so I opened up my SSMS, created a new Baseball database, and looked into how to do this. After poking around a bit (and a little bit of Googling), I found that flat files could be imported by right-clicking the database of your choice (in the below example, “Baseball”), clicking Tasks, and selecting Import Flat File.
Selecting this opened an Import Flat File wizard. First, it prompted me to select the input file. (Note: if you are importing multiple files, as I did for this little exercise, the wizard is smart enough to remember your last folder when you click Browse.)
When it looks at the flat file, it gives you a preview of the data that you’re importing. Since, for this exercise, I’m importing comma-delimited flat files, it was able to put my data into nice, neat columns.
Clicking “Next” brought me to a screen where I could modify the columns. I like this option a lot, as it gives me an opportunity to set up my data schema the way I want. If you’re a SQL or database newbie, I strongly suggest that you learn about primary keys and data types and take the time to set them up at this point.
In this particular example, I set my yearID to char(4), stint to int (I will likely change this to tinyint), teamID to char(3), lgID to char(2), and pretty much everything after lgID to int. I also set my first five columns as a composite primary key and everything else to be nullable.
I must have set these columns up successfully, because when I ran it, it did so without complaining.
I wish I could say that I imported all of my flat files without a hitch, but I did run into a few that didn’t run successfully the first time. Here are some of the issues that I came across.
- I had opened a file in Excel to check data types, forgotten to close it, and the import complained that it couldn’t work because the file was still open.
- I miscalculated a few field sizes, and came across messages saying that my column sizes were too short (for example, setting nvarchar(10) for a column that included data with 15 characters).
- There were a few cases where I simply had the wrong data type.
- My Pitching table included a column for ERA, which I was surprised to see. Reason: ERA (Earned Run Average, for those of you who are baseball-challenged) is a calculated statistic, like batting average. However, batting average was not included in the Batting table. So, I set the column data type to float. However, when I tried to import it, it failed. When I looked at the data, I found entries under ERA that said “inf” (for “infinity”)*. In this case, I did some data cleansing. I got rid of these entries and saved the flat file. It then imported with no problem
(*Some of you might be wondering, how do you get an ERA of infinity? Answer: you give up runs without getting anyone out! Mathematically, you would get a divide-by-zero error for calculating ERA, but in baseball parlance, it means you give up runs and can’t get anyone out!)
So hopefully at this point, you now have an idea as to how to import flat files into a SQL Server database (and maybe even got a small taste of data types and primary keys). And hopefully, this little utility saves you a lot of grief when trying to import flat files.