v5.1

scripts/patch_ninemanga_niadd.py

@@ -425,38 +425,6 @@ replace_exact(
 
                 val dateText = anchorDate ?: rowDate
 
-                val metadataText = when {
-                    anchorDate != null -> anchorText
-                    rowDate != null -> rowText
-                    else -> ""
-                }
-
-                val textBeforeDate = if (dateText != null) {
-                    metadataText
-                        .substringBefore(dateText)
-                        .trim()
-                } else {
-                    metadataText
-                }
-
-                /*
-                 * Niadd exposes the page count immediately before
-                 * the chapter date, for example:
-                 *
-                 *   Падение русала 1 - 42 41 Aug 02, 2024
-                 *                           ^^
-                 *
-                 * Store that number in a URL fragment. Fragments
-                 * are local metadata and are not sent to the server.
-                 */
-                val pageCount = Regex(
-                    """\\s+(\\d+)\\s*$"""
-                )
-                    .find(textBeforeDate)
-                    ?.groupValues
-                    ?.getOrNull(1)
-                    ?.toIntOrNull()
-
                 val chapterTitle = if (
                     dateText != null &&
                     rawTitle.contains(dateText)
@@ -472,15 +440,6 @@ replace_exact(
                     rawTitle
                 }
 
-                val chapterUrl = if (
-                    pageCount != null &&
-                    pageCount > 0
-                ) {
-                    "$href#niadd-pages=$pageCount"
-                } else {
-                    href
-                }
-
                 MangaChapter(
                     id = generateUid(href),
                     title = chapterTitle.takeIf {
@@ -488,7 +447,7 @@ replace_exact(
                     },
                     number = i + 1f,
                     volume = 0,
-                    url = chapterUrl,
+                    url = href,
                     uploadDate = parseChapterDateByLang(
                         dateText.orEmpty()
                     ),
@@ -641,42 +600,78 @@ replace_exact(
     '''    override suspend fun getPages(
         chapter: MangaChapter,
     ): List<MangaPage> {
+        val chapterUrl = chapter.url
+            .substringBefore("#")
+
+        val doc = webClient
+            .httpGet(
+                chapterUrl.toAbsoluteUrl(domain)
+            )
+            .parseHtml()
+
         if (domain.endsWith("niadd.com")) {
-            val originalUrl = chapter.url
+            val pageLabelRegex = Regex(
-                .substringBefore("#")
+                """^\\s*(\\d+)\\s*/\\s*(\\d+)\\s*$"""
-                .substringBefore("?")
+            )
 
-            val totalPages = chapter.url
+            val pageOptions = doc
-                .substringAfter(
+                .select("option[value]")
-                    "#niadd-pages=",
+                .mapNotNull { option ->
-                    "",
+                    val match = pageLabelRegex.matchEntire(
-                )
+                        option.text().trim()
-                .substringBefore("&")
+                    ) ?: return@mapNotNull null
-                .toIntOrNull()
-                ?.takeIf { it > 0 }
-                ?: error(
-                    "Niadd page count is missing for ${chapter.url}. " +
-                        "Refresh manga details and try again."
-                )
 
-            val normalizedUrl = originalUrl
+                    val pageNumber = match
-                .removeSuffix("/")
+                        .groupValues
+                        .getOrNull(1)
+                        ?.toIntOrNull()
+                        ?: return@mapNotNull null
 
-            val baseUrl = if (
+                    val totalPages = match
-                normalizedUrl.endsWith(".html")
+                        .groupValues
-            ) {
+                        .getOrNull(2)
-                normalizedUrl.removeSuffix(".html")
+                        ?.toIntOrNull()
-            } else {
+                        ?: return@mapNotNull null
-                normalizedUrl
-            }
 
-            return (1..totalPages).map { pageNumber ->
+                    val url = option.attr("value").trim()
-                val url = if (pageNumber == 1) {
+                    if (url.isEmpty()) {
-                    originalUrl
+                        return@mapNotNull null
-                } else {
+                    }
-                    "$baseUrl-$pageNumber.html"
+
+                    Triple(pageNumber, totalPages, url)
+                }
+                .distinctBy { (pageNumber, _, _) ->
+                    pageNumber
+                }
+                .sortedBy { (pageNumber, _, _) ->
+                    pageNumber
                 }
 
+            val totalPages = pageOptions
+                .firstOrNull()
+                ?.second
+                ?: error(
+                    "Cannot find Niadd page selector for $chapterUrl"
+                )
+
+            check(
+                pageOptions.all { (_, total, _) ->
+                    total == totalPages
+                }
+            ) {
+                "Inconsistent Niadd page count for $chapterUrl"
+            }
+
+            check(
+                pageOptions.size == totalPages &&
+                    pageOptions.map { it.first } ==
+                    (1..totalPages).toList()
+            ) {
+                "Incomplete Niadd page selector for $chapterUrl: " +
+                    "found ${pageOptions.size} of $totalPages pages"
+            }
+
+            return pageOptions.map { (_, _, url) ->
                 MangaPage(
                     id = generateUid(url),
                     url = url,
@@ -686,12 +681,6 @@ replace_exact(
             }
         }
 
-        val doc = webClient
-            .httpGet(
-                chapter.url.toAbsoluteUrl(domain)
-            )
-            .parseHtml()
-
         return doc
             .body()
             .requireElementById("page")