difftastic/vendored_parsers/tree-sitter-hare/example/sub.ha

56 lines
1.5 KiB
Plaintext

use encoding::utf8;
export type end = void;
fn utf8_byte_len_bounded(iter: *iterator, end: size) size = {
let pos = 0z;
for (let i = 0z; i < end; i += 1) {
let r: rune = match (strings::next(iter)) {
void => break,
r: rune => r,
};
pos += utf8::runesz(r);
};
return pos;
};
fn utf8_byte_len_unbounded(iter: *iterator) size = {
let pos = 0z;
for (true) {
let r: rune = match (strings::next(iter)) {
void => break,
r: rune => r,
};
pos += utf8::runesz(r);
};
return pos;
};
// Returns a substring in the range [start, end - 1], where each argument is the
// index of the Nth rune. If the end argument is given as [strings::end], the
// end of the substring is the end of the original string. The lifetime of the
// substring is the same as that of the original string.
//
// Note that substringing runewise is not always the correct thing to do, and it
// may cause unexpected linguistic errors to arise. You may need to use
// [unicode::graphsub] instead.
export fn sub(s: str, start: size, end: (size | end)) str = {
let iter = iter(s);
let starti = utf8_byte_len_bounded(&iter, start);
let endi = match (end) {
sz: size => starti + utf8_byte_len_bounded(&iter, sz - start),
end => starti + utf8_byte_len_unbounded(&iter),
};
let bytes = toutf8(s);
return fromutf8_unsafe(bytes[starti..endi]);
};
@test fn sub() void = {
assert(sub("a string", 2, end) == "string");
assert(sub("a string", 0, 1) == "a");
assert(sub("a string", 0, 3) == "a s");
assert(sub("a string", 2, 8) == "string");
};