UXStrItem

UXStrItem is one string, boxed so it can go in a collection.

#use <UXKit>            // or #import "UXText.xc"

Overview

class UXStrItem : Object {
    u8* s;      // the string, NUL-terminated
}

The class has one field.

Why it exists

Array holds Objects, and a bare u8* is not one, so a list of strings needs a wrapper. UXStrItem is that wrapper.

UXPathComp and UXKVEntry exist for the same reason. Boxing is explicit in xtc, not automatic, so the wrapper appears in signatures:

static Array<UXStrItem>* split(u8* s, u8 delim)

You mostly do not touch it

UXText.partAt unwraps for you, so the common loop never names the type:

Array<UXStrItem>* parts = UXText.split(line, (u8)',');
for (i32 i = 0; i < (i32)parts.count(); i = i + 1) {
    Stdio.printf("  [%s]\n", UXText.partAt(parts, i));
}

UXText.join takes the whole array back, so a split → filter → join never builds one by hand.

Building one yourself

Two cases need it: making a list to hand to join, and holding strings that did not come from a split.

Array<UXStrItem>* items = new Array();
UXStrItem* it = new UXStrItem();
it.s = UXStr.dup(name);          // copy if `name` is borrowed
items.add(it);

s is a plain pointer: assignment does not copy

Setting s stores the pointer you give it. It is not managed, retained or duplicated.

Everything in UXText gives the box bytes it allocated itself, so a split result owns its strings for as long as the array lives. If you assign a string that belongs to something else, such as a scratch buffer or a value from a driver, copy it first with UXStr.dup.

init sets s to "", not null, so a new item is safe to print before it is filled in. join relies on this because it walks every element.

Fields

s

u8* s

The string. Never null after init.

Conforms to

• Inherits Object

See also

• UXText: split, join, and everything else that produces or consumes these
• UXStr: dup, for when the string is borrowed
• UXPathComp: the same boxing for a path component