PulsarRPA 丰富的 API 使得我们的绝大多数编程场景下，都能够使用一行代码解决“加载-解析-提取”。本文介绍如何使用 Java 风格的异步编程，来解决批量的网页采集问题。
object JvmAsync {
val session = createSession()
fun loadAll() {
fromResource("seeds10.txt").parallelStream()
.map(session::open).map(session::parse).map(FeaturedDocument::guessTitle)
.forEach { println(it) }
}
fun loadAllAsync2() {
val futures = fromResource("seeds10.txt")
.asSequence()
.map { "$it -i 1d" }
.map { session.loadAsync(it) }
.map { it.thenApply { session.parse(it) } }
.map { it.thenApply { it.guessTitle() } }
.map { it.thenAccept { println(it) } }
.toList()
.toTypedArray()
CompletableFuture.allOf(*futures).join()
}
fun loadAllAsync3() {
val futures = session.loadAllAsync(fromResource("seeds10.txt"))
.map { it.thenApply { session.parse(it) } }
.map { it.thenApply { it.guessTitle() } }
.map { it.thenAccept { println(it) } }
.toTypedArray()
CompletableFuture.allOf(*futures).join()
}
fun loadAllAsync4() {
val futures = session.loadAllAsync(fromResource("seeds10.txt"))
.map { it.thenApply { session.parse(it) }.thenApply { it.guessTitle() }.thenAccept { println(it) } }
.toTypedArray()
CompletableFuture.allOf(*futures).join()
}
}