fix ul.direlist

scripts/patch_ninemanga_niadd.py

@@ -17,7 +17,10 @@ parser_file = (
 )
 
 if not parser_file.is_file():
-    print(f"ERROR: parser file not found: {parser_file}", file=sys.stderr)
+    print(
+        f"ERROR: parser file not found: {parser_file}",
+        file=sys.stderr,
+    )
     sys.exit(1)
 
 
@@ -44,6 +47,40 @@ def replace_exact(name: str, old: str, new: str) -> None:
     print(f"[OK] {name}")
 
 
+#
+# Niadd search pagination:
+#
+# NineManga:
+#   ?page=2
+#
+# Niadd:
+#   ?page=2.html
+#
+replace_exact(
+    "Niadd search pagination",
+    '''                append("/search/")
+                append("?page=")
+                append(page.toString())
+''',
+    '''                append("/search/")
+                append("?page=")
+                append(page.toString())
+                if (domain.endsWith("niadd.com")) {
+                    append(".html")
+                }
+''',
+)
+
+
+#
+# Niadd category pagination:
+#
+# Page 1:
+#   /category/
+#
+# Page 2+:
+#   /category/index_2.html
+#
 replace_exact(
     "Niadd category pagination",
     '''            } else {
@@ -67,6 +104,142 @@ replace_exact(
 )
 
 
+#
+# NineManga used:
+#
+#   ul.direlist > li
+#
+# Niadd no longer has ul.direlist.
+#
+# Instead, discover manga cards through their canonical /manga/ links.
+#
+# The same manga can have several links:
+#   - cover
+#   - title
+#   - description
+#
+# Therefore links are grouped by absolute manga URL.
+#
+replace_exact(
+    "Niadd manga list parser",
+    '''        val doc = webClient.httpGet(url).parseHtml()
+        val root = doc.body().selectFirstOrThrow("ul.direlist")
+        val baseHost = root.baseUri().toHttpUrl().host
+        return root.select("li").map { node ->
+            val href = node.selectFirstOrThrow("a").attrAsAbsoluteUrl("href")
+            val relUrl = href.toRelativeUrl(baseHost)
+            val dd = node.selectFirst("dd")
+            Manga(
+                id = generateUid(relUrl),
+                url = relUrl,
+                publicUrl = href,
+                title = dd?.selectFirst("a.bookname")?.text()?.toCamelCase().orEmpty(),
+                altTitles = emptySet(),
+                coverUrl = node.selectFirst("img")?.src(),
+                rating = RATING_UNKNOWN,
+                authors = emptySet(),
+                contentRating = null,
+                tags = emptySet(),
+                state = null,
+                source = source,
+                description = dd?.selectFirst("p")?.html(),
+            )
+        }
+''',
+    '''        val doc = webClient.httpGet(url).parseHtml()
+
+        if (domain.endsWith("niadd.com")) {
+            val mangaLinks = doc.select("a[href]").filter { link ->
+                val href = link.attr("href")
+                href.startsWith("/manga/") ||
+                    href.startsWith("https://${domain}/manga/")
+            }
+
+            val linksByUrl = mangaLinks.groupBy { link ->
+                link.attrAsAbsoluteUrl("href")
+            }
+
+            return linksByUrl.mapNotNull { (href, links) ->
+                val relUrl = href.toRelativeUrl(domain)
+
+                val texts = links
+                    .map { it.text().trim() }
+                    .filter { it.isNotEmpty() }
+                    .distinct()
+
+                val rawTitle = texts.minByOrNull { it.length }
+                    ?: return@mapNotNull null
+
+                val description = texts
+                    .asSequence()
+                    .filter { it != rawTitle }
+                    .filter { it.length > rawTitle.length }
+                    .maxByOrNull { it.length }
+                    ?.removePrefix(rawTitle)
+                    ?.trim()
+                    ?.takeIf { it.isNotEmpty() }
+
+                val coverUrl = links
+                    .asSequence()
+                    .mapNotNull { link ->
+                        link.selectFirst("img")?.src()
+                            ?: link.parent()?.selectFirst("img")?.src()
+                            ?: link.parent()?.parent()?.selectFirst("img")?.src()
+                    }
+                    .firstOrNull()
+
+                Manga(
+                    id = generateUid(relUrl),
+                    url = relUrl,
+                    publicUrl = href,
+                    title = rawTitle.toCamelCase(),
+                    altTitles = emptySet(),
+                    coverUrl = coverUrl,
+                    rating = RATING_UNKNOWN,
+                    authors = emptySet(),
+                    contentRating = null,
+                    tags = emptySet(),
+                    state = null,
+                    source = source,
+                    description = description,
+                )
+            }
+        }
+
+        val root = doc.body().selectFirstOrThrow("ul.direlist")
+        val baseHost = root.baseUri().toHttpUrl().host
+
+        return root.select("li").map { node ->
+            val href = node.selectFirstOrThrow("a").attrAsAbsoluteUrl("href")
+            val relUrl = href.toRelativeUrl(baseHost)
+            val dd = node.selectFirst("dd")
+
+            Manga(
+                id = generateUid(relUrl),
+                url = relUrl,
+                publicUrl = href,
+                title = dd?.selectFirst("a.bookname")?.text()?.toCamelCase().orEmpty(),
+                altTitles = emptySet(),
+                coverUrl = node.selectFirst("img")?.src(),
+                rating = RATING_UNKNOWN,
+                authors = emptySet(),
+                contentRating = null,
+                tags = emptySet(),
+                state = null,
+                source = source,
+                description = dd?.selectFirst("p")?.html(),
+            )
+        }
+''',
+)
+
+
+#
+# Russian statuses on Niadd are not consistently cased:
+#
+#   постоянный
+#   Завершенный
+#
 replace_exact(
     "Russian status parsing",
     '''        //ru
@@ -80,6 +253,12 @@ replace_exact(
 )
 
 
+#
+# Activate Russian source and migrate it to Niadd.
+#
+# Keep NINEMANGA_RU intentionally unchanged so existing Futon
+# database/source references remain compatible.
+#
 replace_exact(
     "NineManga RU to Niadd RU",
     '''    @Broken
@@ -100,7 +279,18 @@ replace_exact(
 )
 
 
-parser_file.write_text(text, encoding="utf-8")
+parser_file.write_text(
+    text,
+    encoding="utf-8",
+)
+
 
 print()
 print(f"[OK] Patched: {parser_file}")
+print()
+print("[OK] Niadd domain")
+print("[OK] Niadd category pagination")
+print("[OK] Niadd search pagination")
+print("[OK] Niadd manga list parser")
+print("[OK] Russian statuses")
+print("[OK] Russian source enabled")