Description

This assignment is based on a class of problem solved in enterprise computing; extraction, transformation, and loading. This is often referred to as ETL. The inputs will be data extracted from a leading aviation industry data and consulting firm, GCR. (See GCR.com for additional data.) The data is in a well known format where each data element is separated from the previous and following data elements by using a comma. It should be noted that this method of data manipulation is extremely common. The explicit order of the data fields and the desired outputs are defined in the “Specifications”.

1          Objectives

The objectives of this assignment are to demonstrate proficiency in file I/O, data structures, and data transformation using C language resources. Specifically, you will read in data from a text file, use that data to populate a data structure, and print that data to STDOUT by accessing the newly populated structure.

1.1     Extraction

The first part of ETL is extraction. The filename of a text file will be passed to your program via the command line. The data contained in that file is to be read into memory (i.e., extracted). Your program will be compiled and run on Eustis using the following commands:

gcc -o etl hw1etl.c ./etl inputFile

It is entirely possible that the input file either does not exist or is not where it is supposed to be. In such an event, your program should print an error message to STDERR that indicates which file is missing, then your program should exit safely. Use the following format for your error message (fileName should display

the actual name of the missing file):

etl ERROR: File “fileName” not found.

The input file is in CSV (comma separated values) format where each line contains the data for one airport and the fields are as printed below. Note that these fields vary in size and content. Some fields may even be empty. Also note that the data for some of the fields are a melange of types. Specifically, the FAA Site Number and both latitude and longitude contain numbers, punctuation, and text.

For this assignment, treat all input data as character data.

Table 1: Airports Data Fields

1.2     Transformation

The second part of ETL is transformation. A list of comma separated values is convenient for text files, but it is far less convenient in memory. Once the data for a single airport has been read into a buffer, you will need to parse the buffer based on the commas between the data fields. The parsed data will then be used to populate a structure of the type struct airPdata (i.e. the data has been transformed from CSV to a data structure). The format of airPdata, shown below, will be defined in airPdata.h. Note that the airPdata structure uses the same names as the input file’s Field Names (See Table 1 on page 2), though not all of the Field Names are used.

typedef struct airPdata{

char *siteNumber; //FAA Site Number char *LocID; //Airport’s ‘‘Short Name’’, \textit{e.g.} MCO char *fieldName; //Airport Name char *city; //Associated City char *state; //State char *latitude; //Latitude char *longitude; //Longitude char controlTower;//Control Tower, this is a single character (Y/N)

} airPdata;

Remember, some of these fields will be of differing lengths for each airport. When you allocate memory structure’s fields, you can assume that no entry will be longer than 50 characters (plus 1 character for the terminating NULL).

1.3     Loading

Finally, the third part of ETL is loading. With the data now in an airPdata structure it can be easily accessed by functions and/or other programs (i.e., loaded). For this assignment, you will use pass the airPdata structure to a function ( PrintData(airPdata airport) ) that will print the data to STDOUT (aka the console). Before calling PrintData for the first time, make sure you print a header line that names each column. Specifically, use the following two lines of code:

printf(“%-12s %-11s %-42s %-34s %-3s %-15s %-16s Tower\n”,

“FAA Site”, “Short Name”, “Airport Name”, “City”, “ST”,

“Latitude”, “Longitude”);

printf(“%-12s %-11s %-42s %-34s %-3s %-15s %-16s =====\n”,

“========”, “==========”, “============”, “====”, “==”,

“========”, “=========”);

The “-” preceding each of the format specifiers left-justifies the printed values, while the numbers indicate the width of the printed field. Your data should be left-justified as well and should use widths that are identical to those in the header line. It is your choice whether you want to populate one airPdata structure and then print it, or to populate an array of airPdata structures and then print each of them. If you choose to populate and print one at a time, be sure to free any allocated memory before reallocating for the same variable. If you choose to read in all of the airports before printing them, you will have an easier time modifying your HW1 code when it comes time for HW3 and will only need to free the memory when you are done. Again, the choice is yours. An example of what the output should look like is shown on the next page.

2          Required Functions

void printData(airPdata airport);

Description: Prints the data for a given airport, using the same format as the provided header line.

Input: A pointer to an airPdata structure.

Special Cases: If a NULL pointer is passed to this function, print an error message to STDERR and return from the function without printing anything to STDOUT.

Returns: Nothing

3          Testing

There are several possible approaches for parsing the input data. Regardless of the approach you use, make sure to test your code on Eustis even if it works perfectly on your machine . If your code does not compile on Eustis you will receive a 0 for the assignment. There will be four (4) files provided for testing your code, they are as follows.

Table 2: Test Files

The expected output for these test cases will also be provided. To compare your output to the expected output you will first need to redirect STDOUT to a text file. Run your code with the following command (substitute the actual name of the input CSV file):

./etl inputFile > output.txt

The run the following command (substitute the actual name of the expected output file):

diff output.txt expectedOutputFile

If there are any differences the relevant lines will be displayed (note that even a single extra space will cause a difference to be detected). If nothing is displayed, then congratulations the outputs match! For each of the four (4) test cases, your code will need to output to STDOUT text that is identical to the corresponding expectedOutputFile. If your code crashes for a particular test case, you will not get credit for that case.