From 88954aa2e4c857edc6ecb7e0ffdc7ba09682ffe9 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Mike=20F=C3=A4hrmann?= <mike_faehrmann@web.de>
Date: Fri, 4 Nov 2022 12:11:43 +0100
Subject: [PATCH] [gelbooru_v02] implement 'notes' extraction

same code as for 'moebooru' works here as well
---
 gallery_dl/extractor/gelbooru_v02.py | 102 +++++++++++++++++++++++----
 1 file changed, 90 insertions(+), 12 deletions(-)

diff --git a/gallery_dl/extractor/gelbooru_v02.py b/gallery_dl/extractor/gelbooru_v02.py
index c3b04459..47df0f27 100644
--- a/gallery_dl/extractor/gelbooru_v02.py
+++ b/gallery_dl/extractor/gelbooru_v02.py
@@ -110,6 +110,23 @@ class GelbooruV02Extractor(booru.BooruExtractor):
             for key, value in tags.items():
                 post["tags_" + key] = " ".join(value)
 
+    def _notes(self, post, page):
+        note_container = text.extract(page, 'id="note-container"', "<img ")[0]
+        if not note_container:
+            return
+
+        post["notes"] = notes = []
+        for note in note_container.split('class="note-box"')[1:]:
+            extr = text.extract_from(note)
+            notes.append({
+                "width" : int(extr("width:", "p")),
+                "height": int(extr("height:", "p")),
+                "y"     : int(extr("top:", "p")),
+                "x"     : int(extr("left:", "p")),
+                "id"    : int(extr('id="note-body-', '"')),
+                "body"  : text.unescape(text.remove_html(extr(">", "</div>"))),
+            })
+
 
 INSTANCES = {
     "realbooru": {
@@ -285,15 +302,81 @@ class GelbooruV02PostExtractor(GelbooruV02Extractor):
     archive_fmt = "{id}"
     pattern = BASE_PATTERN + r"/index\.php\?page=post&s=view&id=(\d+)"
     test = (
-        ("https://rule34.xxx/index.php?page=post&s=view&id=1995545", {
-            "content": "97e4bbf86c3860be18de384d02d544251afe1d45",
-            "options": (("tags", True),),
+        ("https://rule34.xxx/index.php?page=post&s=view&id=863", {
+            "pattern": r"https://api-cdn\.rule34\.xxx/images"
+                       r"/1/6aafbdb3e22f3f3b412ea2cf53321317a37063f3\.jpg",
+            "content": ("a43f418aa350039af0d11cae501396a33bbe2201",
+                        "67b516295950867e1c1ab6bc13b35d3b762ed2a3"),
+            "options": (("tags", True), ("notes", True)),
             "keyword": {
-                "tags_artist": "danraku",
-                "tags_character": "kashima_(kantai_collection)",
-                "tags_copyright": "kantai_collection",
+                "tags_artist": "reverse_noise yamu_(reverse_noise)",
+                "tags_character": "hong_meiling",
+                "tags_copyright": "touhou",
                 "tags_general": str,
-                "tags_metadata": str,
+                "tags_metadata": "censored translated",
+                "notes": [
+                    {
+                        "body": "It feels angry, I'm losing myself... "
+                                "It won't calm down!",
+                        "height": 65,
+                        "id": 93586,
+                        "width": 116,
+                        "x": 22,
+                        "y": 333,
+                    },
+                    {
+                        "body": "REPUTATION OF RAGE",
+                        "height": 272,
+                        "id": 93587,
+                        "width": 199,
+                        "x": 78,
+                        "y": 442,
+                    },
+                ],
+
+            },
+        }),
+        ("https://hypnohub.net/index.php?page=post&s=view&id=1439", {
+            "pattern": r"https://hypnohub\.net/images"
+                       r"/90/24/90245c3c5250c2a8173255d3923a010b\.jpg",
+            "content": "5987c5d2354f22e5fa9b7ee7ce4a6f7beb8b2b71",
+            "options": (("tags", True), ("notes", True)),
+            "keyword": {
+                "tags_artist": "brokenteapot",
+                "tags_character": "hsien-ko",
+                "tags_copyright": "capcom darkstalkers",
+                "tags_general": str,
+                "tags_metadata": "dialogue text translated",
+                "notes": [
+                    {
+                        "body": "Master Master Master "
+                                "Master Master Master",
+                        "height": 83,
+                        "id": 10577,
+                        "width": 129,
+                        "x": 259,
+                        "y": 20,
+                    },
+                    {
+                        "body": "Response Response Response "
+                                "Response Response Response",
+                        "height": 86,
+                        "id": 10578,
+                        "width": 125,
+                        "x": 126,
+                        "y": 20,
+                    },
+                    {
+                        "body": "Obedience Obedience Obedience "
+                                "Obedience Obedience Obedience",
+                        "height": 80,
+                        "id": 10579,
+                        "width": 98,
+                        "x": 20,
+                        "y": 20,
+                    },
+                ],
+
             },
         }),
         ("https://safebooru.org/index.php?page=post&s=view&id=1169132", {
@@ -316,11 +399,6 @@ class GelbooruV02PostExtractor(GelbooruV02Extractor):
             "url": "5a6ebe07bfff8e6d27f7c30b5480f27abcb577d2",
             "content": "1c3831b6fbaa4686e3c79035b5d98460b1c85c43",
         }),
-        ("https://hypnohub.net/index.php?page=post&s=view&id=73964", {
-            "pattern": r"https://hypnohub\.net/images/7a/37"
-                       r"/7a37c0ba372f35767fb10c904a398831\.png",
-            "content": "02d5f5a8396b621a6efc04c5f8ef1b7225dfc6ee",
-        }),
     )
 
     def __init__(self, match):