Challenge question: Histogram of word lengths

Write a function plot$\_$word$\_$length$\_$histogran$($filenane$)$ that reads a file and plots a histogram of the distribution of word lengths in that file. For example, a file named test toxt containing the following:

Here are some words of varying lengths, one of which is ostentatiously lengthy.

should produce the following histogram:

Once you have stored the whole contents of the file $($as a string$)$ in the variable contents, you should extract the words using:

words $-$ re$.$findall$([2-zA-z]+{?}^{\text{'}},$ contents$)$

Requirements:

If your function receives an invalid filename, it should catch the exception and print the message "The file $\{$filename$\}$ was not found." where $($filename$\}$ is the name of the file given to the function.

The title should be "Word lengths in $\{$filename$]",$ where $[$filename$\}$ is the name of the file given to the function.

The $x-$label and $y-$label should be "Word length" and "Number of words" respectively.

You should set the $x-$ticks and their labels explicitly from $0$ to $20($inclusive$)$ in steps of $2.$

The bins should each be $1$ wide, starting at $1$ and finishing at $20($inclusive$).$ You must set these bins explicitly.

It is not a density plot $-$ the vertical axis displays a count of how many words are in each bin.

Your plot must have grid lines.

Feel free to write your own test files. If you would like a large text file to test your program with, here is Mary Shelley's Frankenstein.