Nico Nico Shunga Archive Project

Status

A full thumbnail archive of Nico Nico Shunga is done by yzqzss at saveweb.org.
The full image archive is also finished, but yet to be published. It should be less than 100GB and will be published to the Internet Archive just like the thumbnail. Saveweb is also saving all the webpages for Nico Nico Shunga.

RainSlide (me) has done a small 例のアレ-related tag archive for Nico Nico Shunga.
Other than images, all the metadata (title, author, description, datetime, tags, comments) are kept.

例のアレ-related tags:
"クッキー☆", "BB素材", "例のアレ", "真夏の夜の淫夢", "野獣先輩", "ガチムチ", "必須アモト酸"

Touhou Project related tags are skipped since:

1. there are 18000+ images, just too much for the time being
2. webpages will be saved by saveweb.org. Most data can be extracted from their WARC archive. (not very simple though).

Shunga thumbnails are all archived

Nico Nico Shunga - Thumbnails (WARC format, 9.1G) archived by saveweb.org:
https://archive.org/details/CHUNGA-THUMB-20250121180740887-crawling

WARC Viewer: https://replayweb.page/ (GitHub repo: https://github.com/webrecorder/replayweb.page)

Cookie☆Shunga are downloaded & uploaded to https://gigafile.nu (password-protected zip)

クッキー☆春画 (2551p) archived by ホラー淫夢WIKI:
https://45.gigafile.nu/0124-l6270cae2b818eac13722e52da8260727
see: https://x.com/horrorinm_WIKI/status/1880909639187050879

Message him on Twitter to get the password. ホラー淫夢WIKI is working on other archives and he might be busy. Since there was over 1000 password request, if you know some friend who has the password, you might want to ask your friend instead.

Infomation

1. Nico Nico Shunga Service end date and time: Wednesday, 2025-01-29 14:00
2. Nico Nico Shunga is just the R-15+ part of Nico Nico Seiga. Nico Nico Seiga currently only allows access from Japan.
   Proxy with a Japanese IP is required for non-Japanese user.
3. Nico Nico Shunga requires login for full access.
   Nico Nico account (or its browser cookie) is required.

• Notice of termination of “Nico Nico Shunga” service (Japanese)
  https://blog.nicovideo.jp/niconews/239363.html
• Niconico - Archiveteam Wiki
  https://wiki.archiveteam.org/index.php/Niconico

Tools

gallery-dl has a working support on Nico Nico Seiga:

• repo: https://github.com/mikf/gallery-dl
• code: https://github.com/mikf/gallery-dl/blob/master/gallery_dl/extractor/seiga.py
• can save:
  • image id & original image
  • datetime, title, description, user id, user name

config Nico Nico account login in gallery-dl.conf:

{
    "extractor": {
        "seiga": {
            "username": "username",
            "password": "password",
            "cookies" : {
	            
            }
        }
    }
}

Then download image(s) with:

gallery-dl --proxy <japanese proxy> --write-metadata --sleep 4.1 -f {user['id']!s:R?target=shunga//}-im{image_id}.{extension} -I links.txt

Older code from other project (might not work):

• Nico Nico Seiga code in CeJS online novels / comics downloader
  https://github.com/chiyasakura/work_crawler/blob/master/comic.ja-JP/nico_seiga.js
• 迫真保存部・webscrapingの裏技！｜左飴
  https://note.com/leftame/n/nf5aa45e90df3

Programming infomation

To get a more stable image list, uninterrupted by new image posting, use the post date old to new sorting (投稿の古い順) or URL with sort=image_created_a. Sadly deleting image will still interrupt the list, so be prepared for this.

Following source code are JavaScript. As of now, all the URL are accessible via HTTP GET.

all shunga listing page

• all shunga listing URL:
  https://seiga.nicovideo.jp/shunga/list?page=2&sort=image_created_a
• 25 images per page
• just get a list of URL:

  [...document.querySelectorAll(".illust_pict_all > .illust_list_img")]
  	.map(a => a.href).join("\n");
  

• get all the image ID + image webpages’ URL + image data string in a listing page:

  [...document.querySelectorAll(".illust_pict_all > .illust_list_img")].map(div => {
  	const link = div.querySelector(":scope .center_img_inner");
  	if (!link) {
  		console.error("Element .center_img_inner not found", location.href, div);
  		return null;
  	}
  	const url = link.href;
  	const image = {
  		id: url.split("/").at(-1),
  		url,
  	};
  
  	const info = div.querySelector(":scope .counter_info");
  	if (!info) {
  		console.warn("Element .counter_info not found", location.href, div);
  		image.views_comments_favs = null;
  	} else {
  		image.views_comments_favs = info.textContent;
  	}
  	return image;
  });
  

shunga with specfic tag listing page

• example shunga with specfic tag listing page URL: https://seiga.nicovideo.jp/tag/東方?target=shunga&sort=image_created_a
• 40 image per page
• just get a list of URL to archive them:

  [...document.querySelectorAll(".item_list > .list_item > a:only-child")]
  	.map(a => a.href).join("\n");
  

• get all the image ID + image webpages’ URL + image data HTML string in a listing page:

  [...document.querySelectorAll(".item_list > .list_item > a:only-child")].map(a => {
  	const url = a.href;
  	const image = {
  		id: url.split("/").at(-1),
  		url,
  	};
  
  	const info = a.querySelector(":scope > .illust_count");
  	if (!info) {
  		console.warn("Element .counter_info not found", location.href, div);
  		image.views_comments_favs_html = null;
  	} else {
  		image.views_comments_favs_html = info.innerHTML.replace(/\s{2,}/g, "");
  	}
  	return image;
  });
  

search listing page

• example search listing page URL for shunga: https://seiga.nicovideo.jp/search/東方?target=shunga&sort=image_created_a
• 20 images per page
• just get a list of URL to archive them:

  [...document.querySelectorAll(".illust_pict_all > .illust_list_img")]
  	.map(a => a.href).join("\n");
  

• get all the image ID + image webpages’ URL + image data string in a listing page:

  [...document.querySelectorAll(".illust_pict_all > .illust_list_img")].map(div => {
  	const link = div.querySelector(":scope .center_img_inner");
  	if (!link) {
  		console.error("Element .center_img_inner not found", location.href, div);
  		return null;
  	}
  	const url = link.href.replace(/\?track=seiga_illust_keyword$/, "");
  	const image = {
  		id: url.split("/").at(-1),
  		url,
  	};
  
  	const info = div.querySelector(":scope .counter_info");
  	if (!info) {
  		console.warn("Element .counter_info not found", location.href, div);
  		image.views_comments_favs = null;
  	} else {
  		image.views_comments_favs = info.textContent;
  	}
  	return image;
  });
  

image webpage

• Notations:
  image id: imXXXX / XXXX; base 16 data (0~f): ffffffff; number (0~9): 00000000

• image webpage URL: https://seiga.nicovideo.jp/seiga/imXXXX

• Original image URL: https://seiga.nicovideo.jp/image/source/11552118
  will be HTTP 302 redirected to a real URL https://lohas.nicoseiga.jp/o/ffffffffffffffffffffffffffffffffffffffff/0000000000/XXXX
  can be accessed by: document.getElementById("illust_link").href

• get Original image URL and metadata in image webpage:

  // everything other than ID can be undefined!
  // datetime is skipped since gallery-dl has a good one,
  // and it can be problematic to parse here.
  const data = {
  	id: location.pathname.split("/").at(-1),
  	title:       document.querySelector(".im_head_bar .title")?.textContent,
  	description: document.querySelector(".im_head_bar .discription")?.innerText,
  	author:      document.getElementById("ko_watchlist_header")?.dataset.id,
  	view: Number.parseInt(
  		document.querySelector(".illust_count > .view .count_value")?.textContent
  	),
  	comment: Number.parseInt(
  		document.querySelector(".illust_count > .comment .count_value")?.textContent
  	),
  	clip: Number.parseInt(
  		document.querySelector(".illust_count > .clip .count_value")?.textContent
  	),
  };
  

• get author link if needed:

  const authorLink =
  	data.author.length > 0
  	? `https://seiga.nicovideo.jp/user/illust/${data.author}`
  	: null;
  

• other metadata can be sourced from following links:

Heatmap image

• Some popular shunga has a “heatmap” image.
• Heatmap image URL: https://lohas.nicoseiga.jp/heatmap?id=XXXX&${ Date.now() }
  will return HTTP 404 if there is no heatmap for a shunga
• Heatmap max size: 600px

Image tags

• image tags JSON URL: https://seiga.nicovideo.jp/ajax/illust/tag/list?id=XXXX&_=${ Date.now() }
• should return all the tags without paging
• example tags JSON:

  {
  	"result": "true",
  	"target_id": "728449",
  	"tag_list": [
  	{ "id": "number as string...",
  		"name": "Comment text, where non-ascii chars are encoded as \uffff",
  		"is_locked":true,"is_lockable":true,"is_category":false},
  	{"id":"36","name":"\u6771\u65b9","is_locked":true,"is_lockable":true,"is_category":true}
  	],
  	"is_owner": false,
  	"category_tag_id": "0",
  	"vote_event": [],
  	"link_target": "shunga"
  }
  

image comments

• image comments JSON URL: https://seiga.nicovideo.jp/ajax/illust/comment/list?id=XXXX&mode=all&_=${ Date.now() }
• should return all the comments without paging
• example comments JSON:

  {
  	"result": "true",
  	"comment_list": [
  		{
  			"id": "000000",
  			"text": "Comment text, where non-ascii chars are encoded as \uffff",
  			"date": "yy\/MM\/dd hh:mm",
  			"user": "an anonymous user id for blocking user",
  			"is_owner": false,
  			"is_visible": true,
  			"is_new": false
  		},
  		...
  	]
  }