[🤗] - Read code files line by line rather than all at once

[obfuscatedgenerated]

Strtok strips repeated instances of the separator, so we can't accurately output the line number of an error. Currently we output the "instruction index", but if we wrote a custom strtok implementation, this could be improved.

[obfuscatedgenerated]

It might make more sense to instead read the file line-by-line (with something like fgets). This could also reduce memory usage since we don't allocate a buffer for the whole file at once. lex() would have to accept a file pointer instead, but it makes a lot more sense.