Re: [whatwg/encoding] Make TextEncoder and TextDecoder be transform streams (#127)

@hsivonen Consider the following transform stream:

```javascript
const rechunker = new TransformStream({
  transform(chunk, controller) {
    for (let start = 0; start < chunk.length; start += 1024) {
      const end = Math.min(start + 1024, chunk.length);
      controller.enqueue(chunk.substring(start, end));
    }
  }
});
```
This takes a stream of strings and splits up any chunk that is longer than 1024 code units.

Here are the questions I ask myself:
1. _Is this a useful and sensible thing to want to do?_ I believe so. If the stream you're processing may have arbitrarily large chunks, you're forwarding it to some expensive operation, and you want to avoid jank, I think the correct approach is to include something like this in your pipe.
2. _It is reasonable for a developer to expect this to work and not corrupt the input?_ I think so. 
   1. Corruption wouldn't even happen unless you streamed it to `TextEncoder`, and that could be several steps away down the pipe.
   2. It just applies the Javascript built-in `substring()` function, and nothing on 
 https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/String/substring says anything about data corruption.
   3. It will be a long time before any problems are noticed.
3. _Is this implementation correct?_ Maybe? Maybe not? I don't know.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/whatwg/encoding/pull/127#issuecomment-359694445