i wanna read a plain text file in zig
#reading a file in zig
old kettle
vale perch
You'll need to be more specific. You can suck an entire file into one buffer, read byte by byte, read by lines... Do you have to handle text encodings?
wintry hollow
you probably "do not" want to read byte by byte. in general, unless you are dealing with large files, just read the whole thing, then process it.
wintry hollow
here's a general example:
fn readFile(allocator: Allocator, path: []const u8) ![]const u8 {
var file = try std.fs.cwd().openFile(path, .{});
defer file.close();
var fsz = (try file.stat()).size;
var br = std.io.bufferedReader(file.reader());
var reader = br.reader();
return try reader.readAllAlloc(allocator, fsz);
}
old kettle
so take a file descriptor, get its size, make a reader and tell it to read by allocating on the heap?
frank stone
or you can use std.fs.cwd().readFileAlloc (https://ziglang.org/documentation/master/std/#A;std:fs.Dir)
old kettle
or you can use `std.fs.cwd().readFileAlloc` (https://ziglang.org/documentation/m...
oh this one returns a slice?
regal sky
i wanna read a plain text file in zig
frank stone
oh this one returns a slice?
It returns an Error Union Type because the operation can fail. But yes basically it returns a slice
old kettle
It returns an [Error Union Type](https://ziglang.org/documentation/master/#Error...
right ![]u8
so i just do !void on main and use it with try
frank stone
yes, you can also use catch if you want
frank stone
👍
wintry hollow
or you can use `std.fs.cwd().readFileAlloc` (https://ziglang.org/documentation/m...
you'd probably still want to grab the size ahead of time here (although there are still cases this could cause FileTooBig anyway)
frank stone
you'd probably still want to grab the size ahead of time here (although there ar...
yeah you are right, while the function does use readToEndAllocOptions internally it just passes null so the default allocation in 1kb.
no sorry @wintry hollow the function already gets the file size to allocate before reading
pearl quail
i think i wanna suck it into a stack buffer?
if you want to suck it into a stack buffer instead you can do that too
no need for an allocator
vale perch
Stack memory is more constrained than heap memory, so one must be cognizant of file sizes.
wintry hollow
no sorry <@133933289957228544> the function already gets the file size to alloca...
err, where? there's a 0 size which ultimately uses 1024 for a buffer size. the only other thing is a check to make sure the max size fits in to usize.
old kettle
if you want to suck it into a stack buffer instead you can do that too
yes please
would probably be cleanest
oh wait i got an idea
i am going to use comptime to read it at compile time with an allocator
and store the value in a stack variable
frank stone
i am going to use comptime to read it at compile time with an allocator
you can use @embedFile for this https://ziglang.org/documentation/master/#embedFile
old kettle
yeah solved then