diff options
Diffstat (limited to 'app/src/main/kotlin/com/pitchedapps/frost/web/HeadlessHtmlExtractor.kt')
-rw-r--r-- | app/src/main/kotlin/com/pitchedapps/frost/web/HeadlessHtmlExtractor.kt | 88 |
1 files changed, 88 insertions, 0 deletions
diff --git a/app/src/main/kotlin/com/pitchedapps/frost/web/HeadlessHtmlExtractor.kt b/app/src/main/kotlin/com/pitchedapps/frost/web/HeadlessHtmlExtractor.kt new file mode 100644 index 00000000..50f2f6bc --- /dev/null +++ b/app/src/main/kotlin/com/pitchedapps/frost/web/HeadlessHtmlExtractor.kt @@ -0,0 +1,88 @@ +package com.pitchedapps.frost.web + +import android.annotation.SuppressLint +import android.content.Context +import android.webkit.JavascriptInterface +import android.webkit.WebView +import ca.allanwang.kau.utils.gone +import com.pitchedapps.frost.R +import com.pitchedapps.frost.facebook.USER_AGENT_BASIC +import com.pitchedapps.frost.injectors.InjectorContract +import com.pitchedapps.frost.utils.L +import io.reactivex.Single +import io.reactivex.SingleEmitter +import io.reactivex.android.schedulers.AndroidSchedulers +import io.reactivex.schedulers.Schedulers +import org.jetbrains.anko.runOnUiThread +import java.util.concurrent.TimeUnit + +/** + * Created by Allan Wang on 2017-08-12. + * + * Launches a headless html request and returns a result pair + * When successful, the pair will contain the html content and -1 + * When unsuccessful, the pair will contain an empty string and a StringRes for the given error + * + * All errors are rerouted to success calls, so no exceptions should occur. + * The headless extractor will also destroy itself on cancellation or when the request is finished + */ +fun Context.launchHeadlessHtmlExtractor(url: String, injector: InjectorContract, action: (Single<Pair<String, Int>>) -> Unit) { + val single = Single.create<Pair<String, Int>> { e: SingleEmitter<Pair<String, Int>> -> + val extractor = HeadlessHtmlExtractor(this, url, injector, e) + e.setCancellable { + runOnUiThread { extractor.destroy() } + e.onSuccess("" to R.string.html_extraction_cancelled) + } + }.subscribeOn(AndroidSchedulers.mainThread()) + .timeout(20, TimeUnit.SECONDS, Schedulers.io(), { it.onSuccess("" to R.string.html_extraction_timeout) }) + .onErrorReturn { "" to R.string.html_extraction_error } + action(single) +} + +/** + * Given a link and some javascript, will load the link and load the JS on completion + * The JS is expected to call [HeadlessHtmlExtractor.HtmlJSI.handleHtml], which will be sent + * to the [emitter] + */ +@SuppressLint("ViewConstructor") +private class HeadlessHtmlExtractor( + context: Context, url: String, val injector: InjectorContract, val emitter: SingleEmitter<Pair<String, Int>> +) : WebView(context) { + + val startTime = System.currentTimeMillis() + + init { + L.v("Created HeadlessHtmlExtractor for $url") + gone() + setupWebview(url) + } + + @SuppressLint("SetJavaScriptEnabled") + private fun setupWebview(url: String) { + settings.javaScriptEnabled = true + settings.userAgentString = USER_AGENT_BASIC + webViewClient = HeadlessWebViewClient(url, injector) // basic client that loads our JS once the page has loaded + webChromeClient = QuietChromeClient() // basic client that disables logging + addJavascriptInterface(HtmlJSI(), "Frost") + loadUrl(url) + } + + inner class HtmlJSI { + @JavascriptInterface + fun handleHtml(html: String?) { + val time = System.currentTimeMillis() - startTime + emitter.onSuccess((html ?: "") to -1) + post { + L.d("HeadlessHtmlExtractor fetched $url in $time ms") + settings.javaScriptEnabled = false + settings.blockNetworkLoads = true + destroy() + } + } + } + + override fun destroy() { + super.destroy() + L.d("HeadlessHtmlExtractor destroyed") + } +}
\ No newline at end of file |