Your question cannot really be reasonably answered. You're kind of asking the wrong question here.
There is no constant number of characters per line. A line, in the general, technical sense, can be an arbitrarily long sequence of characters that ends with a newline character \n. If that character just doesn't occur, the line can become extremely long. It really depends on the contents of what you write. It's not like in a word processor (or on this web site), where line breaks occur automatically, and the possible line length depends on font size, or page orientation.
In source code, this is also coupled with the difficulty of determining what exactly a line of code is. Do source code comments, that aren't actually code, i.e. processing instructions, count? Do empty lines (lines with only whitespace) count? For source code, there's a metric called Source Lines of Code, or SLOC. This article has some more information about it.
In a linked article, the H states:
It's worth noting that these figures do include the comments, blank lines, documentation, scripts and userland tools included with the kernel (find . -type f -not -regex '\./\.git.*' | xargs cat | wc -l).
So it's really the number of newline characters, including comments, empty lines, etc. of all files excluding the revision control metadata (git) files. This measures the physical lines (counting the \n characters), and really depends on things like source code formatting.