21st Century C

Part II. The Language

This is the part where we reconsider everything about the C language.

There are two parts to the process: working out what bits of the language not to use, and then finding out about the new things. Some of the new things are syntactic features, such as being able to initialize a list of struct elements by name; some of the new things are functions that have been written for us and are now common, such as the functions that will allow us to write to strings without quite as much pain.

The chapters cover the material as follows:

Chapter 6 provides a guide for those perplexed (or perhaps made a bit uneasy) by pointers.

Chapter 7 is where we start building by tearing down. We’ll go over a survey of concepts covered by the typical textbooks that I believe should be downplayed or considered deprecated.

Chapter 8 addresses C concepts that are too useful to throw out, but that have a number of subtle awkwardnesses.

In Chapter 9, we pay special attention to strings and work out how to handle them without memory allocation or character-counting madness. malloc will be lonely, because you’ll never call it.

Chapter 10 presents newer syntax, which will let us write function calls in ISO-standard C with inputs such as lists of arbitrary length (e.g., sum(1, 2.2, [...] 39, 40)) or named, optional elements (e.g., new_person(.name="Joe", .age=32, .sex='M')). Like rock and roll, these syntactic features saved my life. If I hadn’t known about them, I would have abandoned C a long time ...