本页尚未翻译，因此以英文显示。 参与翻译

RxDart vs FxDart › #16 · Upload in batches of 4

Upload in batches of 4

Toss-up

• fx
• chunk
• map

Requirement

Ten files are queued for upload and the API accepts at most four per request. Group the queue into batches of 4 (the last one is short) and print each batch's size and ids. The data is in the code; both versions must print the lines shown under Expected output.

batch of 4: img-01 img-02 img-03 img-04
batch of 4: img-05 img-06 img-07 img-08
batch of 2: img-09 img-10

Side by side

Why they differ

Non-overlapping batching is core vocabulary in both models, and both say it in one word: bufferCount(4) collects four events before emitting a List, chunk(4) pulls four values into a List per demand. Both flush the short trailing batch when the source runs out. The map that formats each batch is then character-for-character identical.

Where the models would start to diverge is just outside this example's frame. A stream buffers because values arrive on their own schedule — bufferCount also has time-based siblings (bufferTime) that a pull pipeline deliberately doesn't offer, because "what arrived in the last second" has no meaning when the consumer controls arrival. A pull chain chunks because the consumer wants four-at-a-time demand — which is also why chunk composes directly with downstream concurrency (send each batch while assembling the next). For a fixed queue of ten, neither advantage is exercised: a tie.