Talk:Text processing/Max licenses in use

Assumptions

We might be better able to compare code if we document our assumptions.

J

The J solution makes assumptions to allow the interesting code to stand out (rather than the drudgery). This obviously comes at the cost of some robustness (and, incidentally, optimizes for speed).

• Assumes the format of the file is fixed (there will always be spaces in columns 7 11 13 33 37 41). This can be changed easily to find columns of all spaces and cut there; doing so would actually make the program shorter and simpler, but obviously slower.

• Assumes columns 3 and 6 are Date and Job ID respectively. Could easily be checked; pattern matching would permit the code to identify the columns heuristically.

• Assumes a license will be checked out before it's checked in, and checked in before it's checked out, and that initially all licenses are checked in.

• Does *not* assume the file exists or, if it exists, that it contains any data (i.e. it may be zero bytes long). For example, if the file is empty, you get this:

Maximum simultaneous license use is 0 at the following times:

Note that I didn't have to do any special checks to get this behavior, it just fell naturally out of the code. (This happens a lot in J.)