The solutions are written for Python 3.6

From the root; they can be run from the command line as follows:

$ python -m tests.postcode_validator_test

The specification:

• Validate postcodes in the data file
• Report on the row_id where validation fails
• Untar the file if required (using gzip)
• At end of running import; produce file named 'failed_validation.csv' with same columns as above

Given the import csv; taken from: Google drive and placed in the data directory; you could run the import:

$ python -m src.postcode_importer data\import_data.csv.gz

Summary

• Creates validator with compiled regular expression
• Opens output immediately to avoid keeping large data set in memory while processing
• Uses GZip to read file
• Skips first line to avoid header
• Prints (reports) rowId and postcode when failed validation to the console

Specification:

• Modifyto produce success file
• Order the output

Using the same import data from Part 2 -- you could run the import:

$ python -m src.postcode_importer_part3 data\import_data.csv.gz

Summary

• In memory dictionary is used to process the data; this uses more memory but allows us to sort the entire list at the end before we write it out -- to avoid use of memory we could read and write a single line at a line and insert the line into the output at the appropriate location - however this would be extremely slow so wasn't even attempted since a memory constraint wasn't given
• Dictionary keys are sorted during iteration and items are looked up based on sorted Key; An OrderedDictionary wasn't used since we don't care about insertion order here

• 08-Feb-2017 Moved analysis to seperate ANALYSIS.md file
• 07-Feb-2017 Initial implementation