On Monday, 6 January 2025 at 19:30:27 UTC, Renato wrote:
My use case is to parse a file format where a line is expected to have 4 items. I am currently using `takeExactly(4)` but that doesn't let me index on it.
I suspect you might unknowingly be the victim of autodecoding. If your line is a string, ergo an array of immutable `char`s (or a `wstring`), Phobos functions will treat that as a range of `dchar`s. Since reading four Unicode code points (`dchar`s) might potentially consume more than four UTF-8 code points (`char`s) from the line, the result of `takeExactly` can't be random access.
You could use [`.byCodeUnit`](https://dlang.org/phobos/std_utf.html#byCodeUnit) on your line before giving it to `takeExactly`. I think `takeExactly` is random access if it's source range is. However, be aware that this will split any non-ASCII Unicode symbol in the line to multiple characters, all invalid by themselves. It depends on your use case whether this matters or not. If it does, the `staticArray` solution earlier in this thread is probably what you want instead.
