Anri/mobilismScrap
Archived
1
0
Fork 0
Code Issues Pull requests Projects Releases 6 Wiki Activity

First attempt to add data extraction from the search results

Browse source
This commit is contained in:
Mylloon 2021-09-01 02:02:31 +02:00
parent cec312c93c
commit 2b2e61fbe9
1 changed files with 35 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

37
app/src/main/java/com/mylloon/mobidl/Scraper.kt
View file

@ -5,6 +5,7 @@ import com.github.kittinunf.fuel.core.FuelManager
import com.github.kittinunf.fuel.httpGet
import com.github.kittinunf.fuel.httpPost
import com.mylloon.mobidl.MainActivity.Companion.applicationContext
import java.lang.Exception
import java.lang.System.currentTimeMillis
import com.github.kittinunf.result.Result.Failure as FuelFailure
import com.github.kittinunf.result.Result.Success as FuelSuccess
@ -126,10 +127,42 @@ class Scraper(
private fun parse(htmlPage: String): MutableList<String>? { // Parse HTML response to a clean list.
if (debug) println("Fetching results for $app...")
Toast.makeText(applicationContext(), "Fetching results for $app...", Toast.LENGTH_SHORT)
.show()
// println(htmlPage)
if ("No suitable matches were found." in htmlPage) return mutableListOf("noResults")
val elements: MutableList<String> = htmlPage.split("<tr>\n<td>").toMutableList()
val finalElements = mutableListOf<Map<String, String?>>()
elements.removeFirst()
val lastIndex = elements.toList().lastIndex
elements[lastIndex] = elements[lastIndex].split("</td>\n</tr>")[0]
for (i in elements.indices) {
var title: String?
var author: String?
var link: String?
var data: String?
title = try {
Regex(" ?&amp; ?").replace("", Regex("""class="topictitle">(.*)</a>""").find(elements[i])?.value!!)
} catch (e: Exception) {
null
}
author = try {
val regex = """(<br />|</strong>)\n\n?<i class="icon-user"></i> by <a href="\./memberlist\.php\?mode=viewprofile&amp;u=\d+"( style="color: #.*;" class="username-coloured")?>(.*)</a>"""
Regex(regex).find(elements[i])?.value!!
} catch (e: Exception) {
null
}
link = try {
Regex("""\./viewtopic\.php\?f=(\d*)&amp;t=(\d*)&amp""").find(elements[i])?.value!!
} catch (e: Exception) {
null
}
data = try {
Regex("""</a> <i class="icon-time"></i> <small>(.*)</small>""").find(elements[i])?.value!!
} catch (e: Exception) {
null
}
finalElements.add(i, mapOf("title" to title, "author" to author, "link" to link, "data" to data))
}
println(finalElements)
return null
}