So I'm about 700 kanji into RTK and think I'd like to switch over to RTKLite. I'm very confused how to do so. Can someone help?

Originally I thought I would just download a "lite" deck from the Anki shared decks page. But the two listed there don't have the stroke diagrams and stories that my "real" RtK deck has.

I read what seems to be the "authoritative" thread on rtk lite here: http://forum.koohii.com/showthread.php?tid=993. But many of the links there don't work (e.g. http://userscripts.org/ does not load).

To be honest, if there were an easy way to just delete the non-JLPT N2 cards from my deck I would be quite happy to do so. Does anyone know if a way to do this in an automated fashion?

Thanks.

Just as a short side note. If you have the time, I would do the full RTK 1. If you want to read native books, in my opinion you really need it. If you are just aiming for a N2 test, then you can get away with the light version. But I would think the goal would be native material. Sorry I didn't answer your question.

PotbellyPig Wrote:Just as a short side note. If you have the time, I would do the full RTK 1. If you want to read native books, in my opinion you really need it. If you are just aiming for a N2 test, then you can get away with the light version. But I would think the goal would be native material. Sorry I didn't answer your question.

I've never used RTK at all, but I can read native books.

EDIT: Let me be a little less facetious. This is the idea behind RTK Lite. While some people blast through RTK and have a great time with it, other people get bogged down and either never finish or take a very long time. RTK Lite is a way to more quickly enable you to move on to study of real Japanese. At that point, if you want, you can go back and do the other kanji you skipped in RTK, or you can learn new kanji by other methods. But you still get some benefit from the system.

But every time this gets brought up, half the people responding seem unable to tell the difference between "I only want to learn 1000 kanji from RTK" and "I think I will only ever need 1000 kanji in my life."

[ファブリス]

You may not need the "alter sequence" script anymore, though it used to make things a little smoother.

The script was mainly for navigatin the Study pages and using the flashcards on RevTK.

For flashcards on RevTK, you can add a "custom selection". I think this selection from woelpad should work, but you have to add them only as you study them so you can review the new cards each day.

For stories, well you could try print that out, and then use marker to know how far you are, and use the number to browse. A bit cumbersome, I'm afraid there is no "custom sequences" support on the site as of yet.

A request similar to yours popped up a few days ago: http://forum.koohii.com/showthread.php?p...#pid213537 I mention there that I made, and have been using, a "Lite" list based on A Frequency Dictionary of Japanese by Yukio Tono, Makoto Yamazaki, and Kikuo Maekawa (2013), containing the top 5000 words, with accompanying sentences, in the most up-to-date Japanese corpus, and compiled into an Anki deck by Nayr called Core5000. This list has 1503 kanji. I'm slowly working on creating an Anki deck based on this, and will post here (and there) when it's done.

Can you post a link to the full RTK deck you're using right now, with the stroke order you mention? It should be trivial to add stroke order to a deck that lacks these, but I'd like to add everything people find useful in other decks to my Core5000-based RTK Lite deck.

ariariari Wrote:an easy way to just delete the non-JLPT N2 cards from my deck

Here's how I "converted" my current deck to my RTK Lite deck. Open the Anki Browser, select my deck, search for "deck:current" (no quotes), select all cards, and click "Suspend". Then, I used the regular expression search to search for the kanji I wanted to study and unsuspend those: i.e., search for "deck:current (一 or 二 or 三)" to see cards containing those kanji, select them all, and unsuspend them.

Here's the regular expression for selecting kanji per http://forum.koohii.com/showthread.php?p...8#pid80718 in case your regexp-fu is rusty :

deck:current (一 or 二 or 三 or 四 or 五 or 六 or 七 or 八 or 九 or 十 or 口 or 日 or 月 or 田 or 目 or 古 or 吾 or 冒 or 明 or 品 or 呂 or 早 or 世 or 胃 or 旦 or 旧 or 自 or 白 or 百 or 中 or 千 or 舌 or 升 or 昇 or 丸 or 寸 or 専 or 博 or 占 or 上 or 下 or 朝 or 貝 or 員 or 見 or 児 or 元 or 頁 or 凡 or 負 or 万 or 句 or 肌 or 勺 or 的 or 首 or 乙 or 乱 or 直 or 具 or 真 or 工 or 左 or 右 or 有 or 刀 or 刃 or 切 or 召 or 昭 or 則 or 副 or 別 or 丁 or 町 or 可 or 頂 or 子 or 了 or 女 or 好 or 母 or 貫 or 兄 or 小 or 少 or 大 or 多 or 夕 or 外 or 名 or 石 or 肖 or 砂 or 光 or 太 or 器 or 省 or 厚 or 奇 or 川 or 州 or 順 or 水 or 氷 or 永 or 泉 or 原 or 願 or 泳 or 活 or 消 or 況 or 河 or 泊 or 湖 or 測 or 土 or 圧 or 圭 or 封 or 寺 or 時 or 均 or 火 or 炎 or 灯 or 畑 or 灰 or 点 or 照 or 魚 or 漁 or 里 or 黒 or 量 or 埋 or 同 or 向 or 字 or 守 or 完 or 安 or 寄 or 富 or 貯 or 木 or 林 or 森 or 植 or 枯 or 村 or 相 or 机 or 本 or 札 or 案 or 燥 or 未 or 末 or 味 or 妹 or 若 or 草 or 苦 or 薄 or 葉 or 暮 or 苗 or 兆 or 犬 or 状 or 然 or 猫 or 牛 or 特 or 告 or 先 or 洗 or 介 or 界 or 茶 or 合 or 塔 or 王 or 玉 or 宝 or 現 or 呈 or 全 or 理 or 主 or 注 or 柱 or 金 or 銅 or 針 or 道 or 導 or 造 or 逃 or 辺 or 車 or 連 or 輸 or 前 or 各 or 格 or 略 or 客 or 額 or 夏 or 処 or 条 or 落 or 冗 or 軍 or 運 or 夢 or 高 or 亭 or 京 or 涼 or 景 or 舎 or 周 or 週 or 士 or 吉 or 売 or 学 or 覚 or 栄 or 書 or 敗 or 枚 or 故 or 敬 or 言 or 警 or 計 or 訓 or 詰 or 話 or 語 or 読 or 調 or 談 or 式 or 試 or 域 or 茂 or 成 or 城 or 減 or 浅 or 止 or 歩 or 肯 or 歴 or 武 or 正 or 政 or 定 or 走 or 超 or 越 or 是 or 題 or 建 or 延 or 衣 or 装 or 裏 or 遠 or 初 or 布 or 幅 or 帽 or 市 or 姉 or 帯 or 刺 or 制 or 製 or 転 or 芸 or 雨 or 雲 or 曇 or 冬 or 天 or 橋 or 立 or 泣 or 章 or 競 or 童 or 商 or 適 or 滴 or 匕 or 北 or 背 or 比 or 昆 or 皆 or 混 or 旨 or 脂 or 毎 or 海 or 乞 or 乾 or 腹 or 複 or 欠 or 吹 or 歌 or 軟 or 次 or 資 or 音 or 暗 or 識 or 境 or 亡 or 荒 or 望 or 方 or 坊 or 訪 or 放 or 説 or 鋭 or 曽 or 増 or 贈 or 東 or 凍 or 廷 or 燃 or 歳 or 県 or 地 or 池 or 虫 or 独 or 風 or 己 or 起 or 改 or 記 or 包 or 電 or 家 or 場 or 湯 or 羊 or 美 or 洋 or 達 or 差 or 着 or 集 or 進 or 雑 or 準 or 確 or 午 or 許 or 権 or 観 or 羽 or 習 or 翌 or 曜 or 濯 or 困 or 固 or 国 or 団 or 因 or 園 or 回 or 店 or 庫 or 庭 or 庁 or 床 or 磨 or 心 or 忘 or 忍 or 認 or 志 or 誌 or 思 or 応 or 意 or 想 or 息 or 恵 or 恐 or 感 or 憂 or 忙 or 怖 or 憎 or 慣 or 必 or 手 or 看 or 我 or 義 or 議 or 抱 or 批 or 招 or 打 or 捨 or 指 or 持 or 損 or 拾 or 担 or 操 or 接 or 研 or 戒 or 械 or 鼻 or 刑 or 型 or 才 or 財 or 材 or 存 or 在 or 乃 or 及 or 吸 or 丈 or 史 or 吏 or 更 or 硬 or 又 or 双 or 奴 or 怒 or 友 or 抜 or 投 or 設 or 支 or 技 or 枝 or 軽 or 反 or 坂 or 板 or 返 or 販 or 爪 or 乳 or 浮 or 将 or 採 or 菜 or 受 or 授 or 愛 or 払 or 広 or 拡 or 鉱 or 台 or 治 or 始 or 窓 or 去 or 法 or 会 or 至 or 室 or 到 or 互 or 育 or 流 or 出 or 山 or 岩 or 炭 or 入 or 込 or 分 or 貧 or 公 or 谷 or 浴 or 容 or 溶 or 欲 or 賞 or 党 or 堂 or 常 or 皮 or 波 or 破 or 被 or 残 or 列 or 死 or 耳 or 取 or 最 or 恥 or 職 or 買 or 置 or 環 or 夫 or 規 or 替 or 賛 or 失 or 鉄 or 臣 or 蔵 or 臓 or 賢 or 巨 or 力 or 男 or 労 or 募 or 努 or 加 or 協 or 行 or 律 or 復 or 得 or 徒 or 待 or 彼 or 役 or 程 or 税 or 和 or 移 or 秒 or 秋 or 私 or 利 or 香 or 季 or 委 or 米 or 粉 or 粒 or 迷 or 奥 or 数 or 類 or 様 or 求 or 球 or 救 or 竹 or 笑 or 箱 or 筆 or 筒 or 等 or 算 or 答 or 築 or 人 or 住 or 位 or 仲 or 体 or 件 or 仕 or 他 or 伝 or 仏 or 休 or 仮 or 信 or 依 or 例 or 個 or 健 or 側 or 停 or 値 or 倒 or 億 or 使 or 便 or 倍 or 優 or 宿 or 保 or 付 or 符 or 府 or 任 or 賃 or 代 or 袋 or 貸 or 化 or 花 or 貨 or 傾 or 何 or 荷 or 久 or 内 or 丙 or 肉 or 座 or 卒 or 以 or 似 or 営 or 善 or 年 or 夜 or 液 or 換 or 遊 or 旅 or 勿 or 物 or 易 or 尼 or 泥 or 屋 or 屈 or 掘 or 居 or 層 or 局 or 遅 or 刷 or 尺 or 昼 or 戸 or 肩 or 戻 or 涙 or 雇 or 示 or 礼 or 祝 or 福 or 社 or 禁 or 祭 or 察 or 由 or 油 or 宙 or 届 or 押 or 申 or 伸 or 神 or 捜 or 果 or 菓 or 課 or 斤 or 所 or 祈 or 近 or 折 or 断 or 質 or 昨 or 作 or 雪 or 録 or 急 or 寝 or 婦 or 掃 or 当 or 争 or 事 or 康 or 君 or 群 or 両 or 満 or 画 or 歯 or 曲 or 斗 or 料 or 科 or 図 or 用 or 備 or 昔 or 借 or 散 or 廿 or 席 or 度 or 渡 or 焼 or 半 or 判 or 券 or 巻 or 勝 or 片 or 版 or 不 or 否 or 杯 or 矢 or 族 or 知 or 矛 or 柔 or 務 or 帰 or 弓 or 引 or 強 or 弱 or 沸 or 費 or 第 or 弟 or 号 or 朽 or 汚 or 与 or 写 or 身 or 老 or 考 or 孝 or 教 or 者 or 著 or 署 or 暑 or 諸 or 狭 or 挟 or 追 or 師 or 官 or 管 or 父 or 交 or 効 or 較 or 校 or 足 or 路 or 骨 or 過 or 際 or 陽 or 防 or 院 or 降 or 階 or 穴 or 空 or 突 or 究 or 探 or 深 or 丘 or 兵 or 糸 or 線 or 練 or 緒 or 続 or 絵 or 給 or 絡 or 結 or 終 or 級 or 紅 or 紹 or 経 or 約 or 細 or 総 or 綿 or 緑 or 幼 or 後 or 幾 or 機 or 玄 or 畜 or 系 or 係 or 孫 or 卸 or 御 or 服 or 命 or 令 or 零 or 齢 or 冷 or 領 or 勇 or 通 or 踊 or 疑 or 犯 or 厄 or 危 or 宛 or 腕 or 卵 or 留 or 貿 or 印 or 酉 or 酒 or 配 or 尊 or 豆 or 頭 or 短 or 豊 or 喜 or 皿 or 血 or 盗 or 温 or 塩 or 銀 or 根 or 即 or 節 or 退 or 限 or 良 or 娘 or 食 or 飯 or 飲 or 館 or 平 or 呼 or 評 or 希 or 凶 or 胸 or 殺 or 純 or 鈍 or 辛 or 辞 or 壁 or 新 or 親 or 幸 or 報 or 叫 or 収 or 陸 or 勢 or 熱 or 亥 or 刻 or 述 or 術 or 寒 or 毒 or 麦 or 青 or 精 or 情 or 晴 or 清 or 静 or 責 or 績 or 積 or 表 or 契 or 喫 or 害 or 割 or 生 or 星 or 姓 or 性 or 産 or 拝 or 春 or 実 or 奉 or 棒 or 勤 or 漢 or 難 or 垂 or 乗 or 今 or 含 or 念 or 予 or 預 or 野 or 西 or 価 or 要 or 腰 or 票 or 標 or 煙 or 南 or 門 or 問 or 間 or 簡 or 開 or 閉 or 聞 or 非 or 悲 or 罪 or 候 or 決 or 快 or 偉 or 違 or 干 or 刊 or 汗 or 軒 or 岸 or 宇 or 余 or 除 or 途 or 塗 or 束 or 頼 or 速 or 整 or 険 or 検 or 重 or 動 or 働 or 種 or 病 or 疲 or 痛 or 医 or 匹 or 区 or 欧 or 迎 or 登 or 発 or 療 or 形 or 彦 or 顔 or 参 or 修 or 珍 or 文 or 対 or 斉 or 済 or 楽 or 薬 or 率 or 央 or 英 or 映 or 赤 or 変 or 恋 or 湾 or 黄 or 横 or 色 or 絶 or 甘 or 期 or 基 or 舞 or 無 or 組 or 祖 or 査 or 助 or 畳 or 並 or 普 or 湿 or 業 or 共 or 供 or 異 or 洪 or 港 or 暴 or 爆 or 選 or 殿 or 井 or 囲 or 耕 or 亜 or 悪 or 円 or 角 or 触 or 解 or 再 or 講 or 構 or 論 or 輪 or 編 or 冊 or 氏 or 紙 or 婚 or 低 or 底 or 民 or 眠 or 捕 or 補 or 郊 or 部 or 都 or 郵 or 段 or 司 or 伺 or 詞 or 舟 or 航 or 般 or 船 or 来 or 気 or 飛 or 沈 or 妻 or 面 or 革 or 靴 or 声 or 呉 or 誤 or 蒸 or 承 or 極 or 番 or 毛 or 宅 or 長 or 張 or 髪 or 展 or 単 or 戦 or 脳 or 悩 or 鳥 or 鳴 or 島 or 暖 or 偶 or 隅 or 逆 or 免 or 晩 or 勉 or 象 or 像 or 馬 or 験 or 駐 or 駅 or 膚 or 劇 or 能 or 演 or 震 or 農 or 濃 or 送 or 関 or 咲)

You might need to split this big search query into two or three smaller chunks if Anki chokes on the full-sized one (I doubt Python's regular expression module is optimized to handle these absurdly-long queries ). This should get you quickly to where you want to go, while keeping your review logs intact?

aldebrn Wrote:A request similar to yours popped up a few days ago: http://forum.koohii.com/showthread.php?p...#pid213537 I mention there that I made, and have been using, a "Lite" list based on A Frequency Dictionary of Japanese by Yukio Tono, Makoto Yamazaki, and Kikuo Maekawa (2013), containing the top 5000 words, with accompanying sentences, in the most up-to-date Japanese corpus, and compiled into an Anki deck by Nayr called Core5000. This list has 1503 kanji. I'm slowly working on creating an Anki deck based on this, and will post here (and there) when it's done.

Can you post a link to the full RTK deck you're using right now, with the stroke order you mention? It should be trivial to add stroke order to a deck that lacks these, but I'd like to add everything people find useful in other decks to my Core5000-based RTK Lite deck.

Thank you very much for your reply. The deck is titled "RTK 6th Edition [Stories, Stroke Diagrams, Readings]". However, Damien just informed me that this deck was been removed from ankiweb due to copyright issues (https://anki.tenderapp.com/discussions/a...large-deck).

aldebrn Wrote:

ariariari Wrote:an easy way to just delete the non-JLPT N2 cards from my deck

Here's how I "converted" my current deck to my RTK Lite deck. Open the Anki Browser, select my deck, search for "deck:current" (no quotes), select all cards, and click "Suspend". Then, I used the regular expression search to search for the kanji I wanted to study and unsuspend those: i.e., search for "deck:current (一 or 二 or 三)" to see cards containing those kanji, select them all, and unsuspend them.
...

Thank you! This is getting closer to I want. But when I type:

deck:current (一 or 二 or 三)

It returns more than 3 items. My deck has multiple fields per each card. The actual kanji is only written in the field "kanji". Is there a way to modify the above query so that it only searches in the field called "kanji"?

I am familiar with SQL, so I kept of trying "kanji IN (一, 二, 三)" but it didn't work. I'm not sure what the right syntax here is, or whether anki even supports it at all.

ariariari Wrote:I am familiar with SQL, so I kept of trying "kanji IN (一, 二, 三)" but it didn't work. I'm not sure what the right syntax here is, or whether anki even supports it at all.

You don't need to know SQL to issue a right query - you need to RTFM.

You want (kanji:一 or kanji:二 or...)

Fields are stored in a string delimited by \x1f. fieldname:value searches for fields that contain value and nothing else.

kanji一 or 二 or ...) seems to work.

anotherjohn Wrote:kanji一 or 二 or ...) seems to work.

That seems to get parsed into (kanji:一) or 二 or ..., so no.

Vempele Wrote:anotherjohn wrote:

kanji一 or 二 or ...) seems to work.

That seems to get parsed into (kanji:一) or 二 or ..., so no.

You're right. It worked on my kanji deck because there's nothing in the other fields. How lame.

For those interested, here is a PHP script I wrote to take a file that contains a kanji list separated by spaces, and returns a valid Anki query to select those kanji. I used this to split my RTK anki deck into 5 separate decks which correspond to the JLPT kanji lists posted elsewhere on the site.

<?

# input: "一 二 三 四"
# first convert to: "kanji:一 or kanji:二"
# final format needs to be something like:
# deck:current (id:2 or id:3)
$in = file_get_contents("n2-kanji.txt");

function str_split_unicode($str, $l = 0) {
if ($l > 0) {
$ret = array();
$len = mb_strlen($str, "UTF-8");
for ($i = 0; $i < $len; $i += $l) {
$ret[] = mb_substr($str, $i, $l, "UTF-8");
}
return $ret;
}
return preg_split("//u", $str, -1, PREG_SPLIT_NO_EMPTY);
}

$in_array = str_split_unicode($in);

$out = array();
foreach ($in_array as $char)
{
if ($char != " ")
{
$out[] = "kanji:" . $char . " ";
}
}
$out = implode(" or ", $out);


$out = "deck:current (" . $out . ")";
print "\n" . $out . "\n";

Here's a question I've always had: why split RTK along JLPT lines? Are you going to take the five tests? Do these partitions match the grade school learning order, or is there something about them that makes them really attractive?

I ask because I've read that knowing too much kanji can impair one's performance on JLPT because words that are commonly written in kanji are written in the exams as kana, and require you to produce kana. Also I seem to be under the impression that that JLPT levels don't match well to literary genres (i.e., you can find books targeting middle school readers, or elementary school readers, but not to JLPT N3 readers). I'm quite happy using a frequency dictionary to govern my RTK Lite spirals: since there are 1503 kanji in the top 5000 Japanese "words", I've split these up into three spirals for RTK Lite: top 500 first, then 1000, then all 1503 kanji, followed by all RTK1 kanji, then RTK3,... I ask about JLPT to seek information to evaluate this plan.

ariariari Wrote:It returns more than 3 items. My deck has multiple fields per each card. The actual kanji is only written in the field "kanji". Is there a way to modify the above query so that it only searches in the field called "kanji"?

Although I don't know the answer to your question, this shouldn't be that much of a problem right? Unless a large percentage of cards make reference to other kanji, unsuspending cards containing the 1115 Lite kanji anywhere in them should result is not much more than 1115 kanji unsuspended right? Or are you a perfectionist

ariariari Wrote:I am familiar with SQL, so I kept of trying "kanji IN (一, 二, 三)" but it didn't work. I'm not sure what the right syntax here is, or whether anki even supports it at all.

I think this is a real crummy aspect of Anki's design. If you load your collection.anki2 file into a SQLite table browser, you'll see that the "notes" table contains various columns, but all the note fields (the fields you use to construct your cards like your "kanji") are all stored as a JSON string inside one column! So the specific note fields like "kanji" aren't available through SQL queries unless you regexp inside entries. I guess I can see why this was done, Anki being fact-agnostic, but it does make its databases much denser to third-party queries...

ariariari Wrote:I used this to split my RTK anki deck into 5 separate decks which correspond to the JLPT kanji lists posted elsewhere on the site.

If it's not too much trouble, could you throw up the five resulting files containing kanji at each JLPT level on gist.github.com or something?

ariariari Wrote:For those interested, here is a PHP script I wrote to take a file that contains a kanji list separated by spaces, and returns a valid Anki query to select those kanji.

Alternatively, if you have a list separated by spaces, open it in a text editor and find/replace ' ' with ' or kanji:'

I decided to switch to RTK-lite in the middle of my study too. I simply got the list of RTK-lite to be sure not to skip any radical, and went about my deck suspending the kanjis I didn't want to learn yet. It is manual, but don't take that much time and I get to chose some more kanjis to keep. You might spend more time fighting against regex and scripts than it would be to manually suspend a bit ahead of your study.