Advent of Code 2020 - Day 4

This is a running series of posts on the Advent of Code 2020. The introductory post has the details, including the accompanying GitHub repo, in case you reached this directly.

Another one for parsing and validating.

Part 1

The parsing part is a little more complex than what we have seen so far, as we can no longer assume that a record corresponds to a line. Other than that, it's just colon-separated key, value pairs. So instead of building a record in one go, we start with an empty one, adding key, value pairs as they come until reaching a newline, which marks a record switch.

Two potential pitfalls to look out for. First, remember to add the record still being built at the end of the loop. Second, both at this step, and whenever closing a record and opening a new one, check whether the record has any contents; two consecutive blank lines or a blank line at the end of the file would introduce an empty record otherwise.

The validation in this part just needs to check if a record contains all required fields; once again generator expressions make this condition sound almost as natural language:

all(field in record for field in fields)

Then just go through all the records and count those that are valid.

Part 2

This part makes the validation a little bit more involved by specifying requirements for the values in each field. Just validate each.

Some are straightforward, especially after converting the value into a number. Others require a little more work. Some points of interest:

• The sets defined at the top of the module allow a quick check of some fields. In particular, the pid field needs to be checked like this to ensure that it has leading zeroes to pad up to 9 digits.
• Height needs to be treated with care, as the units are not always present, and assuming so may leave an empty string for the numeric part; therefore we require at least four characters, or return there as invalid. After that, it's easy.