diff --git a/CHANGELOG.md b/CHANGELOG.md new file mode 100644 --- /dev/null +++ b/CHANGELOG.md @@ -0,0 +1,12 @@ +## 3.1.0 + +- Overall simplification of library and reduction in boilerplate + - `AsKanji` typeclass removed + - `Rank` and `Level` are now just `Level` + - More liberal use of `Map` and `Set` +- `nanq` renamed to `kanji` and moved inside same project + +## 2.0.0 + +- New JSON-only output +- Revamped, modernized backend diff --git a/Data/Kanji.hs b/Data/Kanji.hs deleted file mode 100644 --- a/Data/Kanji.hs +++ /dev/null @@ -1,150 +0,0 @@ -{-# LANGUAGE TupleSections #-} -{-# LANGUAGE RankNTypes #-} - --- | --- Module : Data.Kanji --- Copyright : (c) Colin Woodbury, 2015, 2016 --- License : GPL3 --- Maintainer: Colin Woodbury --- --- A library for analysing the density of Kanji in given texts, --- according to their "Level" classification, as defined by the --- Japan Kanji Aptitude Testing Foundation (日本漢字能力検定協会). - -module Data.Kanji - ( - -- * Kanji - AsKanji(..) - , Kanji(..) - , allKanji - , isKanji - , hasLevel - -- * Levels - , Level(..) - , Rank(..) - , level - , levels - , isKanjiInLevel - , levelFromRank - -- * Analysis - , percentSpread - , levelDist - , averageLevel - , uniques - -- ** Densities - , kanjiDensity - , elementaryDen - , middleDen - , highDen - , adultDen - ) where - -import Control.Arrow hiding (second) -import Data.Bool (bool) -import Data.List (sort, group) -import qualified Data.Map.Lazy as M -import qualified Data.Set as S -import Lens.Micro - -import Data.Kanji.Levels -import Data.Kanji.Types - ---- - --- | All Japanese Kanji, grouped by their Level (級) in ascending order. --- Here, ascending order means from the lowest to the highest level, --- meaning from 10 to 1. -allKanji :: [S.Set Kanji] -allKanji = map f ks - where f = foldr (\k s -> bool s (S.insert (Kanji k) s) $ isKanji k) mempty - ks = [tenth, ninth, eighth, seventh, sixth, - fifth, fourth, third, preSecond, second] - --- | Is the `Level` of a given `Kanji` known? -hasLevel :: Kanji -> Bool -hasLevel k = has _Just $ level k - --- | What is the density @d@ of Kanji characters in a given String-like --- type, where @0 <= d <= 1@? -kanjiDensity :: Int -> [Kanji] -> Float -kanjiDensity len ks = fromIntegral (length ks) / fromIntegral len - --- | How much of the Kanji found are learned in elementary school in Japan? --- --- > elementaryDen . levelDist :: [Kanji] -> Float -elementaryDen :: [(Rank,Float)] -> Float -elementaryDen dists = sum $ dists ^.. each . inRank [Five, Six ..] - --- | How much of the Kanji found are learned by the end of middle school? --- --- > middleDen . levelDist :: [Kanji] -> Float -middleDen :: [(Rank,Float)] -> Float -middleDen dists = sum $ dists ^.. each . inRank [Three, Four ..] - --- | How much of the Kanji found are learned by the end of high school? --- --- > highDen . levelDist :: [Kanji] -> Float -highDen :: [(Rank,Float)] -> Float -highDen dists = sum $ dists ^.. each . inRank [PreTwo, Three ..] - --- | How much of the Kanji found should be able to read by the average person? --- --- > adultDen . levelDist :: [Kanji] -> Float -adultDen :: [(Rank,Float)] -> Float -adultDen dists = sum $ dists ^.. each . inRank [Two, PreTwo ..] - -inRank :: [Rank] -> Traversal' (Rank,Float) Float -inRank rs f (r,n) | r `elem` rs = (r,) <$> f n - | otherwise = pure (r,n) - --- | All `Level`s, with all their `Kanji`, ordered from Level-10 to Level-2. -levels :: [Level] -levels = zipWith Level allKanji [Ten ..] - --- | What `Level` does a Kanji belong to? -level :: Kanji -> Maybe Level -level = level' levels - where level' [] _ = Nothing - level' (q:qs) k | isKanjiInLevel q k = Just q - | otherwise = level' qs k - --- | Does a given `Kanji` belong to the given `Level`? -isKanjiInLevel :: Level -> Kanji -> Bool -isKanjiInLevel q k = S.member k $ _allKanji q - --- | Is there a `Level` that corresponds with a given `Rank` value? -levelFromRank :: Rank -> Maybe Level -levelFromRank = levelFromRank' levels - where levelFromRank' [] _ = Nothing - levelFromRank' (q:qs) qn | _rank q == qn = Just q - | otherwise = levelFromRank' qs qn - --- | Find the average `Level` of a given set of `Kanji`. -averageLevel :: [Kanji] -> Float -averageLevel ks = average ranks - where ranks = ks ^.. each . to level . _Just . to _rank . to fromRank - average ns = sum ns / fromIntegral (length ns) - --- | How much of each `Level` is represented by a group of Kanji? -levelDist :: [Kanji] -> [(Rank,Float)] -levelDist ks = map toNumPercentPair $ group sortedRanks - where sortedRanks = sort $ ks ^.. each . to level . _Just . to _rank - toNumPercentPair qns = (head qns, length' qns / length' ks) - length' n = fromIntegral $ length n - --- | The distribution of each `Kanji` in a set of them. --- The distribution values must sum to 1. -percentSpread :: [Kanji] -> [(Kanji,Float)] -percentSpread ks = map getPercent kQuants - where getPercent (k,q) = (k, fromIntegral q / totalKanji) - kQuants = kanjiQuantities ks - totalKanji = fromIntegral $ foldl (\acc (_,q) -> q + acc) 0 kQuants - --- | Determines how many times each `Kanji` appears in given set of them. -kanjiQuantities :: [Kanji] -> [(Kanji,Int)] -kanjiQuantities = map (head &&& length) . group . sort - --- | Which Kanji appeared from each Level in the text? -uniques :: [Kanji] -> [(Rank,[Kanji])] -uniques = M.toList . S.foldl h M.empty . S.fromList - where h a k = maybe a (\l -> M.insertWith (++) (_rank l) [k] a) $ level k diff --git a/Data/Kanji/Levels.hs b/Data/Kanji/Levels.hs deleted file mode 100644 --- a/Data/Kanji/Levels.hs +++ /dev/null @@ -1,170 +0,0 @@ --- | --- Module : Data.Kanji.Levels --- Copyright : (c) Colin Woodbury, 2015, 2016 --- License : GPL3 --- Maintainer: Colin Woodbury --- --- All Kanji from levels 10 to Pre-2. - -module Data.Kanji.Levels where - -import qualified Data.Set as S - ---- - --- | The Kanji unique to Level-10, studied by the end of 1st grade in --- Japanese elementary schools. -tenth :: S.Set Char -tenth = S.fromDistinctAscList $ - "一七三上下中九二五人休先入八六円出力十千" ++ - "口右名四土夕大天女子字学小山川左年手文日" ++ - "早月木本村林校森正気水火犬玉王生田男町白" ++ - "百目石空立竹糸耳花草虫見貝赤足車金雨青音" - --- | The Kanji unique to Level-9, studied by the end of 2nd grade in --- Japanese elementary schools. -ninth :: S.Set Char -ninth = S.fromDistinctAscList $ - "万丸交京今会体何作元兄光公内冬刀分切前北" ++ - "午半南原友古台合同回図国園地場声売夏外多" ++ - "夜太妹姉室家寺少岩工市帰広店弓引弟弱強当" ++ - "形後心思戸才教数新方明星春昼時晴曜書朝来" ++ - "東楽歌止歩母毎毛池汽活海点父牛理用画番直" ++ - "矢知社秋科答算米紙細組絵線羽考聞肉自船色" ++ - "茶行西親角言計記話語読谷買走近通週道遠里" ++ - "野長門間雪雲電頭顔風食首馬高魚鳥鳴麦黄黒" - --- | The Kanji unique to Level-8, studied by the end of 3rd grade in --- Japanese elementary schools. -eighth :: S.Set Char -eighth = S.fromDistinctAscList $ - "丁世両主乗予事仕他代住使係倍全具写列助勉" ++ - "動勝化区医去反取受号向君味命和品員商問坂" ++ - "央始委守安定実客宮宿寒対局屋岸島州帳平幸" ++ - "度庫庭式役待急息悪悲想意感所打投拾持指放" ++ - "整旅族昔昭暑暗曲有服期板柱根植業様横橋次" ++ - "歯死氷決油波注泳洋流消深温港湖湯漢炭物球" ++ - "由申界畑病発登皮皿相県真着短研礼神祭福秒" ++ - "究章童笛第筆等箱級終緑練羊美習者育苦荷落" ++ - "葉薬血表詩調談豆負起路身転軽農返追送速進" ++ - "遊運部都配酒重鉄銀開院陽階集面題飲館駅鼻" - --- | The Kanji unique to Level-7, studied by the end of 4th grade in --- Japanese elementary schools. -seventh :: S.Set Char -seventh = S.fromDistinctAscList $ - "不争付令以仲伝位低例便信倉候借停健側働億" ++ - "兆児共兵典冷初別利刷副功加努労勇包卒協単" ++ - "博印参史司各告周唱喜器囲固型堂塩士変夫失" ++ - "好季孫完官害察巣差希席帯底府康建径徒得必" ++ - "念愛成戦折挙改救敗散料旗昨景最望未末札材" ++ - "束松果栄案梅械極標機欠歴残殺毒氏民求治法" ++ - "泣浅浴清満漁灯無然焼照熱牧特産的省祝票種" ++ - "積競笑管節粉紀約結給続置老胃脈腸臣航良芸" ++ - "芽英菜街衣要覚観訓試説課議象貨貯費賞軍輪" ++ - "辞辺連達選郡量録鏡関陸隊静順願類飛飯養験" - --- | The Kanji unique to Level-6, studied by the end of 5th grade in --- Japanese elementary schools. -sixth :: S.Set Char -sixth = S.fromDistinctAscList $ - "久仏仮件任似余価保修俵個備像再刊判制券則" ++ - "効務勢厚句可営因団圧在均基報境墓増夢妻婦" ++ - "容寄富導居属布師常幹序弁張往復徳志応快性" ++ - "恩情態慣承技招授採接提損支政故敵断旧易暴" ++ - "条枝査格桜検構武比永河液混減測準演潔災燃" ++ - "版犯状独率現留略益眼破確示祖禁移程税築精" ++ - "素経統絶綿総編績織罪群義耕職肥能興舌舎術" ++ - "衛製複規解設許証評講謝識護豊財貧責貸貿賀" ++ - "資賛質輸述迷退逆造過適酸鉱銅銭防限険際雑" ++ - "非預領額飼" - --- | The Kanji unique to Level-5, studied by the end of 6th grade in --- Japanese elementary schools. -fifth :: S.Set Char -fifth = S.fromDistinctAscList $ - "並乱乳亡仁供俳値傷優党冊処刻割創劇勤危卵" ++ - "厳収后否吸呼善困垂城域奏奮姿存孝宅宇宗宙" ++ - "宝宣密寸専射将尊就尺届展層己巻幕干幼庁座" ++ - "延律従忘忠憲我批担拝拡捨探推揮操敬映晩暖" ++ - "暮朗机枚染株棒模権樹欲段沿泉洗派済源潮激" ++ - "灰熟片班異疑痛皇盛盟看砂磁私秘穀穴窓筋策" ++ - "簡糖系紅納純絹縦縮署翌聖肺背胸脳腹臓臨至" ++ - "若著蒸蔵蚕衆裁装裏補視覧討訪訳詞誌認誕誠" ++ - "誤論諸警貴賃遺郵郷針鋼閉閣降陛除障難革頂" ++ - "骨" - --- | The Kanji unique to Level-4, studied during middle school in Japan. -fourth :: S.Set Char -fourth = S.fromDistinctAscList $ - "丈与丘丹乾互井介仰伺依侵俗倒偉傍傾僧儀兼" ++ - "冒凡凶刈到刺剣剤劣勧匹占即却及叫召吐含吹" ++ - "咲唐嘆噴圏坊執堅堤塔壁壊壱奇奥奴妙姓威娘" ++ - "婚寂寝尋尽尾屈峠峰巡巨帽幅幾床弐弾彩影彼" ++ - "征御微徴忙怒怖恋恐恒恥恵悩惑惨慎慢慮憶戒" ++ - "戯扇払扱抗抜抱抵押拍拓拠振捕掘描握援搬摘" ++ - "撃攻敏敷斜旨旬是普暇暦曇更替朱朽杯枯柄柔" ++ - "桃欄歓歳殖殿汗汚沈沖沢沼況泊浜浮浸涙淡添" ++ - "渡溶滴漫澄濁濃為烈煙煮燥爆狂狩狭猛獣獲玄" ++ - "珍環甘畳疲療皆盆盗監盤盾眠瞬矛砲祈秀称稲" ++ - "稿突端箇範粒紋紫紹絡継維網緯縁繁繰罰翼耐" ++ - "肩肪胴脂脚脱腐腕腰膚致舗舞舟般芋芝茂荒菓" ++ - "蓄薄薪被襲触訴詰詳誇誉謡豪販賦贈越趣距跡" ++ - "跳踊踏躍軒較載輝輩込迎迫逃透途遅違遣避郎" ++ - "釈鈍鉛鋭鎖鑑闘陣陰隠隣隷雄雅雌離雷需震霧" ++ - "露響項頼飾香駆騒驚髪鬼鮮麗黙鼓齢" - --- | The Kanji unique to Level-3, studied by the end of middle school --- in Japan. -third :: S.Set Char -third = S.fromDistinctAscList $ - "乏乙了企伏伐伴伸佳侍促倣倹偶催債克免冗冠" ++ - "凍凝刑削励勘募匠匿卑卓卸厘又双吉吏哀哲啓" ++ - "喚喫嘱坑埋塊塗墜墨墳墾壇奉契奪如妨姫娯婆" ++ - "婿嫁嬢孔孤宴審寿封尿岐岳峡崩巧帆帝幻幽廉" ++ - "廊弧彫徐忌怠怪恨悔悟悦惜愚慈慌慕慨慰憂憎" ++ - "憩房抑択抽拘掃掌排掛控措掲揚換揺携搾摂撮" ++ - "擁擦敢斗斤斥施既昇晶暫架某桑棄棋楼概欧欺" ++ - "殊殴没泌浪湾湿滅滑滝滞漂漏潜潤濫瀬炉炊炎" ++ - "焦牲犠猟獄甲畔畜疾痘癖硬碑礎祉稚穂穏穫窒" ++ - "符篤簿籍粋粗粘糧紛紺絞綱緊締緩縛縫繕翻聴" ++ - "肝胆胎胞脅膜膨芳苗菊華葬藩虐虚蛮衝衰袋裂" ++ - "裸覆訂託詠該誘請諮諾謀譲豚貫賊賢赦赴超軌" ++ - "軸辛辱逮遂遇遭遵邦邪郊郭酔酵鋳錠錬錯鍛鎮" ++ - "鐘閲阻陪陳陵陶隆随隔隻雇零霊顧飽餓駐騎髄" ++ - "魂魅魔鯨鶏" - --- | The Kanji unique to Level-Pre2, considerend "mid high school" level. -preSecond :: S.Set Char -preSecond = S.fromDistinctAscList $ - "且丙亜享亭仙伯但佐併侮侯俊俸倫偏偵偽傑傘" ++ - "僕僚儒償充准凸凹刃剖剛剰劾勅勲升厄叔叙吟" ++ - "呈呉唆唇唯喝喪嗣嚇囚坪垣培堀堕堪塀塁塑塚" ++ - "塾壌壮奔奨妃妄妊妥姻娠媒嫌嫡宜宰宵寛寡寧" ++ - "寮尉尚尼履屯岬崇崎帥幣庶庸廃廷弊弔弦彰循" ++ - "徹忍恭悠患悼惰愁愉慶憤憾懇懐懲懸戻扉扶抄" ++ - "把披抹拐拒拙括拷挑挟挿捜据搭摩撤撲擬斉斎" ++ - "旋昆暁曹朕朴杉析枠枢柳栓核栽桟棚棟棺槽款" ++ - "殉殻汁江沸泡泥泰洞津洪浄浦涯涼淑渇渉渋渓" ++ - "渦溝滋漆漠漬漸潟濯煩爵猫献猶猿珠琴璽瓶甚" ++ - "畝疎疫症痢痴癒盲眺睡督矯砕硝硫碁磨礁祥禅" ++ - "禍租秩稼窃窮窯竜筒粛粧糾紡索累紳緒縄繊繭" ++ - "缶罷羅翁耗肌肖肢肯臭舶艇艦茎荘菌薦薫藻虜" ++ - "虞蚊蛇蛍融衡衷裕褐褒襟覇訟診詐詔誓諭謁謄" ++ - "謙謹譜貞貢賄賓賜賠購践軟轄迅迭逐逓逝逸遍" ++ - "遮遷還邸酌酢酪酬酷醜醸釣鈴鉢銃銘閑閥附陥" ++ - "隅雰霜靴韻頑頒頻顕飢駄騰麻" - --- | The Kanji unique to Level-2, considered "standard adult" level. -second :: S.Set Char -second = S.fromDistinctAscList $ - "串丼乞亀伎侶俺傲僅冥冶凄刹剥勃勾匂叱呂呪" ++ - "咽哺唄唾喉喩嗅嘲埼堆塞填奈妖妬媛嫉宛尻岡" ++ - "崖嵐巾弄弥彙怨恣惧慄憧憬戚戴拉拭拳拶挨挫" ++ - "捉捗捻摯斑斬旦旺昧曖曽枕柵柿栃桁梗梨椅椎" ++ - "楷毀氾汎汰沃沙淫湧溺潰煎熊爪爽牙狙玩瑠璃" ++ - "璧瓦畏畿痕痩瘍眉睦瞭瞳稽窟箋箸籠綻緻罵羞" ++ - "羨肘股脇脊腎腫腺膝膳臆臼舷艶芯苛茨萎葛蓋" ++ - "蔑蔽藍藤虎虹蜂蜜袖裾訃詣詮誰諦諧謎貌貪貼" ++ - "賂賭踪蹴辣遜遡那酎醒采釜錦錮鍋鍵鎌闇阜阪" ++ - "隙韓頃須頓頬顎餅餌駒骸鬱鶴鹿麓麺" diff --git a/Data/Kanji/Types.hs b/Data/Kanji/Types.hs deleted file mode 100644 --- a/Data/Kanji/Types.hs +++ /dev/null @@ -1,104 +0,0 @@ -{-# LANGUAGE FlexibleInstances #-} - --- | --- Module : Data.Kanji.Types --- Copyright : (c) Colin Woodbury, 2015, 2016 --- License : GPL3 --- Maintainer: Colin Woodbury --- --- Types for this library. Note that typeclass instances for `ByteString`s --- are not provided, as `SB.pack` garbles characters above 0xFF. - -module Data.Kanji.Types where - -import Data.Char (ord) -import Data.Kanji.Types.Internal () -import Data.Maybe (fromJust) -import qualified Data.Set as S -import qualified Data.Text as ST -import qualified Data.Text.Lazy as LT -import Lens.Micro - ---- - --- | Anything that can be transformed into a list of Kanji. -class AsKanji a where - - -- | Traverse into this type to find 0 or more Kanji. - -- - -- Despite what the Haddock documentation says, this is part of the - -- minimal complete definition. - _Kanji :: Traversal' a Kanji - - -- | Transform this string type into a list of Kanji. The source string - -- and the resulting list might not have the same length, if there - -- were `Char` in the source that did not fall within the legal - -- UTF8 range for Kanji. - asKanji :: a -> [Kanji] - asKanji a = a ^.. _Kanji - -instance AsKanji Char where - _Kanji f c = if isKanji c then _kanji <$> f (Kanji c) else pure c - {-# INLINE _Kanji #-} - -instance AsKanji [Char] where - _Kanji = traverse . _Kanji - {-# INLINE _Kanji #-} - -instance AsKanji ST.Text where - _Kanji = each . _Kanji - {-# INLINE _Kanji #-} - -instance AsKanji LT.Text where - _Kanji = each . _Kanji - {-# INLINE _Kanji #-} - --- | A single symbol of Kanji. Japanese Kanji were borrowed from China --- over several waves during the past millenium. Japan names 2136 of --- these as their standard set, with rarer characters being the domain --- of academia and esoteric writers. --- --- Japanese has several Japan-only Kanji, including: --- --- * 畑 (a type of rice field) --- * 峠 (a narrow mountain pass) --- * 働 (to do physical labour) -newtype Kanji = Kanji { _kanji :: Char } deriving (Eq, Ord, Show) - --- | A Level or "Kyuu" (級) of Japanese Kanji ranking. There are 12 of these, --- from 10 to 1, including intermediate levels between 3 and 2, and 2 and 1. --- --- Japanese students will typically have Level-5 ability by the time they --- finish elementary school. Level-5 accounts for 1006 characters. --- --- By the end of middle school, they would have covered up to Level-3 --- (1607 Kanji) in their Japanese class curriculum. --- --- While Level-2 (2136 Kanji) is considered "standard adult" ability, --- many adults could not pass the Level-2, or even the Level-Pre2 (1940 Kanji) --- exam without considerable study. --- --- Level data for Kanji above Level-2 is currently not provided by --- this library. -data Level = Level { _allKanji :: S.Set Kanji - , _rank :: Rank - } deriving (Eq, Show) - --- | A numeric representation of a `Level`. -data Rank = Ten | Nine | Eight | Seven | Six | Five | Four | Three | PreTwo - | Two | PreOne | One deriving (Eq, Ord, Enum, Read, Show) - --- | Discover a Rank's numeric representation, as a `Float`. -fromRank :: Rank -> Float -fromRank = fromJust . flip lookup rankMap - --- | A mapping of Ranks to their numeric representation. -rankMap :: [(Rank,Float)] -rankMap = zip [Ten ..] [10,9,8,7,6,5,4,3,2.5,2,1.5,1] - --- | Legal Kanji appear between UTF8 characters 19968 and 40959. -isKanji :: Char -> Bool -isKanji c = lowLimit <= c' && c' <= highLimit - where c' = ord c - lowLimit = 19968 -- This is `一` - highLimit = 40959 -- I don't have the right fonts to display this. diff --git a/Data/Kanji/Types/Internal.hs b/Data/Kanji/Types/Internal.hs deleted file mode 100644 --- a/Data/Kanji/Types/Internal.hs +++ /dev/null @@ -1,58 +0,0 @@ -{-# LANGUAGE FlexibleInstances #-} -{-# LANGUAGE MultiParamTypeClasses #-} -{-# LANGUAGE TypeFamilies #-} -{-# LANGUAGE UndecidableInstances #-} - --- | --- Module : Data.Kanji.Types.Internal --- Copyright : (c) Colin Woodbury, 2015, 2016 --- License : GPL3 --- Maintainer: Colin Woodbury --- --- Instances needed for using `each` over `Text` types. --- This code is borrowed from `microlens-platform`: --- http://hackage.haskell.org/package/microlens-platform - -module Data.Kanji.Types.Internal () where - -import Data.Monoid (Endo) -import qualified Data.Text as T -import qualified Data.Text.Lazy as TL -import Lens.Micro -import Lens.Micro.Internal - ---- - -instance (a ~ Char, b ~ Char) => Each T.Text T.Text a b where - each = strictText - {-# INLINE each #-} - -instance (a ~ Char, b ~ Char) => Each TL.Text TL.Text a b where - each = lazyText - {-# INLINE each #-} - -strictUnpacked :: Lens' T.Text String -strictUnpacked f t = T.pack <$> f (T.unpack t) -{-# INLINE strictUnpacked #-} - -strictText :: Traversal' T.Text Char -strictText = strictUnpacked . traversed -{-# INLINE [0] strictText #-} - -{-# RULES -"strict text -> map" strictText = sets T.map :: ASetter' T.Text Char; -"strict text -> foldr" strictText = foldring T.foldr :: Getting (Endo r) T.Text Char; - #-} - -lazyUnpacked :: Lens' TL.Text String -lazyUnpacked f t = TL.pack <$> f (TL.unpack t) -{-# INLINE lazyUnpacked #-} - -lazyText :: Traversal' TL.Text Char -lazyText = lazyUnpacked . traversed -{-# INLINE [0] lazyText #-} - -{-# RULES -"lazy text -> map" lazyText = sets TL.map :: ASetter' TL.Text Char; -"lazy text -> foldr" lazyText = foldring TL.foldr :: Getting (Endo r) TL.Text Char; - #-} diff --git a/README.md b/README.md new file mode 100644 --- /dev/null +++ b/README.md @@ -0,0 +1,92 @@ +Kanji +==== + +`kanji` is a Japanese Kanji library and analysation program written in Haskell. Its main +function is to tell what Kanji belong to what Level of the Japanese National +Kanji Examination (漢字検定). + +`kanji` can be used to: + - determine what Level individual Kanji belong to + - determine the average Level (difficulty, in other words) of a group of Kanji + - apply the above to whole files of Japanese + +INSTALLING `kanji` +--------------- +First, get the source files from: + +https://github.com/fosskers/kanji + +`kanji` is written in Haskell and uses the +[stack](http://docs.haskellstack.org/en/stable/README.html) tool. Once +`stack` is installed, move to the source directory and perform: + + stack install + +USAGE +----- +Assuming you've made it so that you can run the executable, the following +command-line options are available: + +``` +Usage: kanji [-u|--unknowns] [-d|--density] [-e|--elementary] [-l|--leveldist] + [-a|--average] [-s|--splits] ((-f|--file ARG) | JAPANESE) + +Available options: + -h,--help Show this help text + -u,--unknowns Find Kanji whose Level couldn't be determined + -d,--density Find how much of the input is made of Kanji + -e,--elementary Find density of Kanji learnt in elementary school + -l,--leveldist Find the distribution of Kanji levels + -a,--average Find the average Level of all Kanji present + -s,--splits Show which Level each Kanji belongs to + -f,--file ARG Take input from a file +``` + +#### NOTES ON CLOs + * All options above can be mixed to include their analysis result + in the output JSON. + * `-h` will over-ride any other options or arguments, discarding them and + printing a help message. + +#### Examples +*Single Kanji* +``` +$> kanji -s 日 +{ + "levelSplit": { + "Ten": [ + "日" + ] + } +} +``` + +*A Japanese sentence* +``` +$> kanji -s これは日本語 +{ + "levelSplit": { + "Nine": ["語"], + "Ten": ["本", "日"] + } +} +``` + +*All options* +``` +$> kanji -leadus これは日本語 +{ + "levelSplit": { + "Nine": ["語"], + "Ten": ["本", "日"] + }, + "elementary": 1, + "average": 9.666667, + "density": 0.5, + "unknowns": [], + "distributions": { + "Nine": 0.33333334, + "Ten": 0.6666667 + } +} +``` diff --git a/Setup.hs b/Setup.hs deleted file mode 100644 --- a/Setup.hs +++ /dev/null @@ -1,2 +0,0 @@ -import Distribution.Simple -main = defaultMain diff --git a/bench/Bench.hs b/bench/Bench.hs new file mode 100644 --- /dev/null +++ b/bench/Bench.hs @@ -0,0 +1,26 @@ +module Main where + +import Criterion.Main +import Data.Foldable (fold) +import Data.Kanji +import Data.Kanji.Types +import qualified Data.Set as S + +--- + +main :: IO () +main = defaultMain + [ bgroup "Analysis" + [ + bench "testKanji - uniques" $ nf uniques testKanji + , bench "bigKanji - uniques" $ nf uniques bigKanji + , bench "level 一" $ nf level (Kanji '一') + , bench "level 串" $ nf level (Kanji '串') + ] + ] + +testKanji :: [Kanji] +testKanji = map Kanji "本猫机鉛筆紙帽子包丁床天井壁光電気棚箱靴家階段戸屋根地面" + +bigKanji :: [Kanji] +bigKanji = S.toList $ fold allKanji diff --git a/kanji.cabal b/kanji.cabal --- a/kanji.cabal +++ b/kanji.cabal @@ -1,35 +1,92 @@ -name: kanji -version: 3.0.2 -synopsis: Perform 漢字検定 (Japan Kanji Aptitude Test) level analysis on Japanese Kanji. -description: Perform 漢字検定 (Japan Kanji Aptitude Test) level analysis on Japanese Kanji. -license: GPL-3 -license-file: LICENSE -author: Colin Woodbury -maintainer: colingw@gmail.com - -homepage: https://github.com/fosskers/nanq --- copyright: -category: Data -build-type: Simple --- extra-source-files: -cabal-version: >=1.10 +-- This file has been generated from package.yaml by hpack version 0.20.0. +-- +-- see: https://github.com/sol/hpack +-- +-- hash: c5331657544f2fd01e194d0fe17da254a4f957bf458c13dca0c5da45ea3566db -library - exposed-modules: Data.Kanji - , Data.Kanji.Types - , Data.Kanji.Levels +name: kanji +version: 3.1.0 +synopsis: Perform 漢字検定 (Japan Kanji Aptitude Test) level analysis on Japanese Kanji +description: Perform 漢字検定 (Japan Kanji Aptitude Test) level analysis on Japanese Kanji. +category: Data +homepage: https://github.com/fosskers/kanji +author: Colin Woodbury +maintainer: colingw@gmail.com +copyright: 2011 - 2018 Colin Woodbury +license: BSD3 +license-file: LICENSE +build-type: Simple +cabal-version: >= 1.10 - other-modules: Data.Kanji.Types.Internal +extra-source-files: + CHANGELOG.md + README.md - other-extensions: FlexibleInstances +library + hs-source-dirs: + lib + ghc-options: -fwarn-unused-imports -fwarn-unused-binds -fwarn-name-shadowing -fwarn-unused-matches -fwarn-incomplete-patterns -fwarn-incomplete-uni-patterns + build-depends: + aeson >=1.2 && <1.3 + , base >=4.7 && <5 + , containers + , deepseq + , hashable + , text + exposed-modules: + Data.Kanji + Data.Kanji.Levels + Data.Kanji.Types + default-language: Haskell2010 - build-depends: base >= 4.8 && < 4.11 - , containers >= 0.5 && < 0.6 - , bytestring >= 0.10 && < 0.11 - , microlens >= 0.4.1 && < 0.5 - , text >= 1.2 && < 1.3 +executable kanji + main-is: nanq.hs + hs-source-dirs: + nanq + default-extensions: NoImplicitPrelude OverloadedStrings + ghc-options: -fwarn-unused-imports -fwarn-unused-binds -fwarn-name-shadowing -fwarn-unused-matches -fwarn-incomplete-patterns -fwarn-incomplete-uni-patterns -O2 + build-depends: + aeson >=1.2 && <1.3 + , aeson-pretty + , base >=4.7 && <5 + , containers + , kanji + , microlens >=0.4 && <0.5 + , microlens-aeson >=2.2 && <2.3 + , microlens-platform + , optparse-applicative >=0.14 && <0.15 + , protolude >=0.2 && <0.3 + , text + default-language: Haskell2010 - -- hs-source-dirs: - default-language: Haskell2010 +test-suite kanji-test + type: exitcode-stdio-1.0 + main-is: Test.hs + hs-source-dirs: + test + ghc-options: -fwarn-unused-imports -fwarn-unused-binds -fwarn-name-shadowing -fwarn-unused-matches -fwarn-incomplete-patterns -fwarn-incomplete-uni-patterns -threaded + build-depends: + HUnit-approx >=1.1 && <1.2 + , aeson >=1.2 && <1.3 + , base >=4.7 && <5 + , containers + , kanji + , tasty >=0.11 && <1.1 + , tasty-hunit >=0.9 && <0.11 + , text + default-language: Haskell2010 - ghc-options: -O2 +benchmark kanji-bench + type: exitcode-stdio-1.0 + main-is: Bench.hs + hs-source-dirs: + bench + ghc-options: -fwarn-unused-imports -fwarn-unused-binds -fwarn-name-shadowing -fwarn-unused-matches -fwarn-incomplete-patterns -fwarn-incomplete-uni-patterns -threaded -O2 + build-depends: + aeson >=1.2 && <1.3 + , base >=4.7 && <5 + , containers + , criterion >=1.1 && <1.3 + , kanji + , text + default-language: Haskell2010 diff --git a/lib/Data/Kanji.hs b/lib/Data/Kanji.hs new file mode 100644 --- /dev/null +++ b/lib/Data/Kanji.hs @@ -0,0 +1,114 @@ +-- | +-- Module : Data.Kanji +-- Copyright : (c) Colin Woodbury, 2015 - 2018 +-- License : GPL3 +-- Maintainer: Colin Woodbury +-- +-- A library for analysing the density of Kanji in given texts, +-- according to their "Level" classification, as defined by the +-- Japan Kanji Aptitude Testing Foundation (日本漢字能力検定協会). + +module Data.Kanji + ( + -- * Kanji + Kanji + , kanji, _kanji + , allKanji + , isKanji + -- * Levels + , Level(..) + , level + -- * Analysis + , percentSpread + , levelDist + , averageLevel + , uniques + -- ** Densities + , kanjiDensity + , elementaryDen + , middleDen + , highDen + , adultDen + ) where + +import Control.Applicative ((<|>)) +import Control.Arrow hiding (second) +import Data.Bool (bool) +import Data.Foldable (foldl') +import Data.Kanji.Levels +import Data.Kanji.Types +import Data.List (sort, group) +import qualified Data.Map.Strict as M +import Data.Maybe (catMaybes) +import Data.Semigroup ((<>)) +import qualified Data.Set as S + +--- + +-- | All Japanese Kanji, grouped by their Level (級). +allKanji :: M.Map Level (S.Set Kanji) +allKanji = M.fromList . zip [ Ten .. ] $ map (S.map Kanji) ks + where ks = [ tenth, ninth, eighth, seventh, sixth + , fifth, fourth, third, preSecond, second ] + +-- | What `Level` does a Kanji belong to? +level :: Kanji -> Maybe Level +level k = M.foldlWithKey' (\acc l ks -> acc <|> bool Nothing (Just l) (S.member k ks)) Nothing allKanji + +-- | Given the length of some String-like type and a list of `Kanji` found therein, +-- what percentage of them were Kanji? +kanjiDensity :: Int -> [Kanji] -> Float +kanjiDensity len ks = fromIntegral (length ks) / fromIntegral len + +-- | How much of the Kanji found are learnt in elementary school in Japan? +-- +-- > elementaryDen . levelDist :: [Kanji] -> Float +elementaryDen :: M.Map Level Float -> Float +elementaryDen m = M.foldl' (+) 0 . M.restrictKeys m $ S.fromList [ Five, Six .. ] + +-- | How much of the Kanji found are learnt by the end of middle school? +-- +-- > middleDen . levelDist :: [Kanji] -> Float +middleDen :: M.Map Level Float -> Float +middleDen m = M.foldl' (+) 0 . M.restrictKeys m $ S.fromList [ Three, Four .. ] + +-- | How much of the Kanji found are learnt by the end of high school? +-- +-- > highDen . levelDist :: [Kanji] -> Float +highDen :: M.Map Level Float -> Float +highDen m = M.foldl' (+) 0 . M.restrictKeys m $ S.fromList [ PreTwo, Three .. ] + +-- | How much of the Kanji found should be able to be read by the average person? +-- +-- > adultDen . levelDist :: [Kanji] -> Float +adultDen :: M.Map Level Float -> Float +adultDen m = M.foldl' (+) 0 . M.restrictKeys m $ S.fromList [ Two, PreTwo .. ] + +-- | Find the average `Level` of a given set of `Kanji`. +averageLevel :: [Kanji] -> Float +averageLevel ks = average . map numericLevel . catMaybes $ map level ks + where average ns = foldl' (+) 0 ns / fromIntegral (length ns) + +-- | How much of each `Level` is represented by a group of Kanji? +-- The distribution values must sum to 1. +levelDist :: [Kanji] -> M.Map Level Float +levelDist ks = M.fromList . map percentPair . group . sort . catMaybes $ map level ks + where percentPair qns = (head qns, fromIntegral (length qns) / totalKs) + totalKs = fromIntegral $ length ks + +-- | The distribution of each `Kanji` in a set of them. +-- The distribution values must sum to 1. +percentSpread :: [Kanji] -> M.Map Kanji Float +percentSpread ks = getPercent <$> kQuants + where getPercent q = fromIntegral q / totalKanji + kQuants = kanjiQuantities ks + totalKanji = fromIntegral $ M.foldl' (+) 0 kQuants + +-- | Determines how many times each `Kanji` appears in given set of them. +kanjiQuantities :: [Kanji] -> M.Map Kanji Int +kanjiQuantities = M.fromList . map (head &&& length) . group . sort + +-- | Which Kanji appeared from each Level in the text? +uniques :: [Kanji] -> M.Map Level (S.Set Kanji) +uniques = S.foldl' h M.empty . S.fromList + where h a k = maybe a (\l -> M.insertWith (<>) l (S.singleton k) a) $ level k diff --git a/lib/Data/Kanji/Levels.hs b/lib/Data/Kanji/Levels.hs new file mode 100644 --- /dev/null +++ b/lib/Data/Kanji/Levels.hs @@ -0,0 +1,170 @@ +-- | +-- Module : Data.Kanji.Levels +-- Copyright : (c) Colin Woodbury, 2015 - 2018 +-- License : GPL3 +-- Maintainer: Colin Woodbury +-- +-- All Kanji from levels 10 to Pre-2. + +module Data.Kanji.Levels where + +import qualified Data.Set as S + +--- + +-- | The Kanji unique to Level-10, studied by the end of 1st grade in +-- Japanese elementary schools. +tenth :: S.Set Char +tenth = S.fromDistinctAscList $ + "一七三上下中九二五人休先入八六円出力十千" ++ + "口右名四土夕大天女子字学小山川左年手文日" ++ + "早月木本村林校森正気水火犬玉王生田男町白" ++ + "百目石空立竹糸耳花草虫見貝赤足車金雨青音" + +-- | The Kanji unique to Level-9, studied by the end of 2nd grade in +-- Japanese elementary schools. +ninth :: S.Set Char +ninth = S.fromDistinctAscList $ + "万丸交京今会体何作元兄光公内冬刀分切前北" ++ + "午半南原友古台合同回図国園地場声売夏外多" ++ + "夜太妹姉室家寺少岩工市帰広店弓引弟弱強当" ++ + "形後心思戸才教数新方明星春昼時晴曜書朝来" ++ + "東楽歌止歩母毎毛池汽活海点父牛理用画番直" ++ + "矢知社秋科答算米紙細組絵線羽考聞肉自船色" ++ + "茶行西親角言計記話語読谷買走近通週道遠里" ++ + "野長門間雪雲電頭顔風食首馬高魚鳥鳴麦黄黒" + +-- | The Kanji unique to Level-8, studied by the end of 3rd grade in +-- Japanese elementary schools. +eighth :: S.Set Char +eighth = S.fromDistinctAscList $ + "丁世両主乗予事仕他代住使係倍全具写列助勉" ++ + "動勝化区医去反取受号向君味命和品員商問坂" ++ + "央始委守安定実客宮宿寒対局屋岸島州帳平幸" ++ + "度庫庭式役待急息悪悲想意感所打投拾持指放" ++ + "整旅族昔昭暑暗曲有服期板柱根植業様横橋次" ++ + "歯死氷決油波注泳洋流消深温港湖湯漢炭物球" ++ + "由申界畑病発登皮皿相県真着短研礼神祭福秒" ++ + "究章童笛第筆等箱級終緑練羊美習者育苦荷落" ++ + "葉薬血表詩調談豆負起路身転軽農返追送速進" ++ + "遊運部都配酒重鉄銀開院陽階集面題飲館駅鼻" + +-- | The Kanji unique to Level-7, studied by the end of 4th grade in +-- Japanese elementary schools. +seventh :: S.Set Char +seventh = S.fromDistinctAscList $ + "不争付令以仲伝位低例便信倉候借停健側働億" ++ + "兆児共兵典冷初別利刷副功加努労勇包卒協単" ++ + "博印参史司各告周唱喜器囲固型堂塩士変夫失" ++ + "好季孫完官害察巣差希席帯底府康建径徒得必" ++ + "念愛成戦折挙改救敗散料旗昨景最望未末札材" ++ + "束松果栄案梅械極標機欠歴残殺毒氏民求治法" ++ + "泣浅浴清満漁灯無然焼照熱牧特産的省祝票種" ++ + "積競笑管節粉紀約結給続置老胃脈腸臣航良芸" ++ + "芽英菜街衣要覚観訓試説課議象貨貯費賞軍輪" ++ + "辞辺連達選郡量録鏡関陸隊静順願類飛飯養験" + +-- | The Kanji unique to Level-6, studied by the end of 5th grade in +-- Japanese elementary schools. +sixth :: S.Set Char +sixth = S.fromDistinctAscList $ + "久仏仮件任似余価保修俵個備像再刊判制券則" ++ + "効務勢厚句可営因団圧在均基報境墓増夢妻婦" ++ + "容寄富導居属布師常幹序弁張往復徳志応快性" ++ + "恩情態慣承技招授採接提損支政故敵断旧易暴" ++ + "条枝査格桜検構武比永河液混減測準演潔災燃" ++ + "版犯状独率現留略益眼破確示祖禁移程税築精" ++ + "素経統絶綿総編績織罪群義耕職肥能興舌舎術" ++ + "衛製複規解設許証評講謝識護豊財貧責貸貿賀" ++ + "資賛質輸述迷退逆造過適酸鉱銅銭防限険際雑" ++ + "非預領額飼" + +-- | The Kanji unique to Level-5, studied by the end of 6th grade in +-- Japanese elementary schools. +fifth :: S.Set Char +fifth = S.fromDistinctAscList $ + "並乱乳亡仁供俳値傷優党冊処刻割創劇勤危卵" ++ + "厳収后否吸呼善困垂城域奏奮姿存孝宅宇宗宙" ++ + "宝宣密寸専射将尊就尺届展層己巻幕干幼庁座" ++ + "延律従忘忠憲我批担拝拡捨探推揮操敬映晩暖" ++ + "暮朗机枚染株棒模権樹欲段沿泉洗派済源潮激" ++ + "灰熟片班異疑痛皇盛盟看砂磁私秘穀穴窓筋策" ++ + "簡糖系紅納純絹縦縮署翌聖肺背胸脳腹臓臨至" ++ + "若著蒸蔵蚕衆裁装裏補視覧討訪訳詞誌認誕誠" ++ + "誤論諸警貴賃遺郵郷針鋼閉閣降陛除障難革頂" ++ + "骨" + +-- | The Kanji unique to Level-4, studied during middle school in Japan. +fourth :: S.Set Char +fourth = S.fromDistinctAscList $ + "丈与丘丹乾互井介仰伺依侵俗倒偉傍傾僧儀兼" ++ + "冒凡凶刈到刺剣剤劣勧匹占即却及叫召吐含吹" ++ + "咲唐嘆噴圏坊執堅堤塔壁壊壱奇奥奴妙姓威娘" ++ + "婚寂寝尋尽尾屈峠峰巡巨帽幅幾床弐弾彩影彼" ++ + "征御微徴忙怒怖恋恐恒恥恵悩惑惨慎慢慮憶戒" ++ + "戯扇払扱抗抜抱抵押拍拓拠振捕掘描握援搬摘" ++ + "撃攻敏敷斜旨旬是普暇暦曇更替朱朽杯枯柄柔" ++ + "桃欄歓歳殖殿汗汚沈沖沢沼況泊浜浮浸涙淡添" ++ + "渡溶滴漫澄濁濃為烈煙煮燥爆狂狩狭猛獣獲玄" ++ + "珍環甘畳疲療皆盆盗監盤盾眠瞬矛砲祈秀称稲" ++ + "稿突端箇範粒紋紫紹絡継維網緯縁繁繰罰翼耐" ++ + "肩肪胴脂脚脱腐腕腰膚致舗舞舟般芋芝茂荒菓" ++ + "蓄薄薪被襲触訴詰詳誇誉謡豪販賦贈越趣距跡" ++ + "跳踊踏躍軒較載輝輩込迎迫逃透途遅違遣避郎" ++ + "釈鈍鉛鋭鎖鑑闘陣陰隠隣隷雄雅雌離雷需震霧" ++ + "露響項頼飾香駆騒驚髪鬼鮮麗黙鼓齢" + +-- | The Kanji unique to Level-3, studied by the end of middle school +-- in Japan. +third :: S.Set Char +third = S.fromDistinctAscList $ + "乏乙了企伏伐伴伸佳侍促倣倹偶催債克免冗冠" ++ + "凍凝刑削励勘募匠匿卑卓卸厘又双吉吏哀哲啓" ++ + "喚喫嘱坑埋塊塗墜墨墳墾壇奉契奪如妨姫娯婆" ++ + "婿嫁嬢孔孤宴審寿封尿岐岳峡崩巧帆帝幻幽廉" ++ + "廊弧彫徐忌怠怪恨悔悟悦惜愚慈慌慕慨慰憂憎" ++ + "憩房抑択抽拘掃掌排掛控措掲揚換揺携搾摂撮" ++ + "擁擦敢斗斤斥施既昇晶暫架某桑棄棋楼概欧欺" ++ + "殊殴没泌浪湾湿滅滑滝滞漂漏潜潤濫瀬炉炊炎" ++ + "焦牲犠猟獄甲畔畜疾痘癖硬碑礎祉稚穂穏穫窒" ++ + "符篤簿籍粋粗粘糧紛紺絞綱緊締緩縛縫繕翻聴" ++ + "肝胆胎胞脅膜膨芳苗菊華葬藩虐虚蛮衝衰袋裂" ++ + "裸覆訂託詠該誘請諮諾謀譲豚貫賊賢赦赴超軌" ++ + "軸辛辱逮遂遇遭遵邦邪郊郭酔酵鋳錠錬錯鍛鎮" ++ + "鐘閲阻陪陳陵陶隆随隔隻雇零霊顧飽餓駐騎髄" ++ + "魂魅魔鯨鶏" + +-- | The Kanji unique to Level-Pre2, considerend "mid high school" level. +preSecond :: S.Set Char +preSecond = S.fromDistinctAscList $ + "且丙亜享亭仙伯但佐併侮侯俊俸倫偏偵偽傑傘" ++ + "僕僚儒償充准凸凹刃剖剛剰劾勅勲升厄叔叙吟" ++ + "呈呉唆唇唯喝喪嗣嚇囚坪垣培堀堕堪塀塁塑塚" ++ + "塾壌壮奔奨妃妄妊妥姻娠媒嫌嫡宜宰宵寛寡寧" ++ + "寮尉尚尼履屯岬崇崎帥幣庶庸廃廷弊弔弦彰循" ++ + "徹忍恭悠患悼惰愁愉慶憤憾懇懐懲懸戻扉扶抄" ++ + "把披抹拐拒拙括拷挑挟挿捜据搭摩撤撲擬斉斎" ++ + "旋昆暁曹朕朴杉析枠枢柳栓核栽桟棚棟棺槽款" ++ + "殉殻汁江沸泡泥泰洞津洪浄浦涯涼淑渇渉渋渓" ++ + "渦溝滋漆漠漬漸潟濯煩爵猫献猶猿珠琴璽瓶甚" ++ + "畝疎疫症痢痴癒盲眺睡督矯砕硝硫碁磨礁祥禅" ++ + "禍租秩稼窃窮窯竜筒粛粧糾紡索累紳緒縄繊繭" ++ + "缶罷羅翁耗肌肖肢肯臭舶艇艦茎荘菌薦薫藻虜" ++ + "虞蚊蛇蛍融衡衷裕褐褒襟覇訟診詐詔誓諭謁謄" ++ + "謙謹譜貞貢賄賓賜賠購践軟轄迅迭逐逓逝逸遍" ++ + "遮遷還邸酌酢酪酬酷醜醸釣鈴鉢銃銘閑閥附陥" ++ + "隅雰霜靴韻頑頒頻顕飢駄騰麻" + +-- | The Kanji unique to Level-2, considered "standard adult" level. +second :: S.Set Char +second = S.fromDistinctAscList $ + "串丼乞亀伎侶俺傲僅冥冶凄刹剥勃勾匂叱呂呪" ++ + "咽哺唄唾喉喩嗅嘲埼堆塞填奈妖妬媛嫉宛尻岡" ++ + "崖嵐巾弄弥彙怨恣惧慄憧憬戚戴拉拭拳拶挨挫" ++ + "捉捗捻摯斑斬旦旺昧曖曽枕柵柿栃桁梗梨椅椎" ++ + "楷毀氾汎汰沃沙淫湧溺潰煎熊爪爽牙狙玩瑠璃" ++ + "璧瓦畏畿痕痩瘍眉睦瞭瞳稽窟箋箸籠綻緻罵羞" ++ + "羨肘股脇脊腎腫腺膝膳臆臼舷艶芯苛茨萎葛蓋" ++ + "蔑蔽藍藤虎虹蜂蜜袖裾訃詣詮誰諦諧謎貌貪貼" ++ + "賂賭踪蹴辣遜遡那酎醒采釜錦錮鍋鍵鎌闇阜阪" ++ + "隙韓頃須頓頬顎餅餌駒骸鬱鶴鹿麓麺" diff --git a/lib/Data/Kanji/Types.hs b/lib/Data/Kanji/Types.hs new file mode 100644 --- /dev/null +++ b/lib/Data/Kanji/Types.hs @@ -0,0 +1,92 @@ +{-# LANGUAGE DeriveAnyClass, DeriveGeneric #-} + +-- | +-- Module : Data.Kanji.Types +-- Copyright : (c) Colin Woodbury, 2015, 2016 +-- License : GPL3 +-- Maintainer: Colin Woodbury +-- +-- Types for this library. While a constructor for `Kanji` is made available +-- here, you should prefer the `kanji` "smart constructor" unless you know +-- for sure that the `Char` in question falls within the correct UTF8 range. + +module Data.Kanji.Types where + +import Control.DeepSeq (NFData) +import Data.Aeson +import Data.Aeson.Encoding (text) +import Data.Bool (bool) +import Data.Char (ord) +import Data.Hashable +import qualified Data.Map.Strict as M +import Data.Maybe (fromJust) +import qualified Data.Text as T +import GHC.Generics + +--- + +-- | A single symbol of Kanji. Japanese Kanji were borrowed from China +-- over several waves during the last 1,500 years. Japan names 2,136 of +-- these as their standard set, with rarer characters being the domain +-- of academia and esoteric writers. +-- +-- Japanese has several Japan-only Kanji, including: +-- +-- * 畑 (a type of rice field) +-- * 峠 (a narrow mountain pass) +-- * 働 (to do physical labour) +newtype Kanji = Kanji { + _kanji :: Char -- ^ The original `Char` of a `Kanji`. + } deriving (Eq, Ord, Show, Generic, Hashable, NFData) + +instance ToJSON Kanji where + toJSON (Kanji c) = String $ T.singleton c + +instance FromJSON Kanji where + parseJSON = withText "Kanji" $ \t -> case T.uncons t of + Nothing -> fail "No Kanji given" + Just (c, t') | not (T.null t') -> fail "More than one Kanji given in a single String" + | otherwise -> pure $ Kanji c + +-- | Construct a `Kanji` value from some `Char` if it falls in the correct UTF8 range. +kanji :: Char -> Maybe Kanji +kanji c = bool Nothing (Just $ Kanji c) $ isKanji c + +-- | A Level or "Kyuu" (級) of Japanese Kanji ranking. There are 12 of these, +-- from 10 to 1, including intermediate levels between 3 and 2, and 2 and 1. +-- +-- Japanese students will typically have Level-5 ability by the time they +-- finish elementary school. Level-5 accounts for 1,006 characters. +-- +-- By the end of middle school, they would have covered up to Level-3 +-- (1607 Kanji) in their Japanese class curriculum. +-- +-- While Level-2 (2,136 Kanji) is considered "standard adult" ability, +-- many adults could not pass the Level-2, or even the Level-Pre2 (1940 Kanji) +-- exam without considerable study. +-- +-- Level data for Kanji above Level-2 is currently not provided by +-- this library. +data Level = Ten | Nine | Eight | Seven | Six | Five | Four | Three | PreTwo + | Two | PreOne | One + deriving (Eq, Ord, Enum, Show, Generic, Hashable, NFData, ToJSON, FromJSON) + +instance ToJSONKey Level where + toJSONKey = ToJSONKeyText f g + where f = T.pack . show + g = text . T.pack . show + +-- | Discover a `Level`'s numeric representation, as a `Float`. +numericLevel :: Level -> Float +numericLevel = fromJust . flip M.lookup rankMap + +-- | A mapping of Ranks to their numeric representation. +rankMap :: M.Map Level Float +rankMap = M.fromList $ zip [Ten ..] [10,9,8,7,6,5,4,3,2.5,2,1.5,1] + +-- | Legal Kanji appear between UTF8 characters 19968 and 40959. +isKanji :: Char -> Bool +isKanji c = lowLimit <= c' && c' <= highLimit + where c' = ord c + lowLimit = 19968 -- This is `一` + highLimit = 40959 -- I don't have the right fonts to display this. diff --git a/nanq/nanq.hs b/nanq/nanq.hs new file mode 100644 --- /dev/null +++ b/nanq/nanq.hs @@ -0,0 +1,106 @@ +{-# LANGUAGE FlexibleContexts #-} +{-# LANGUAGE TupleSections #-} + +module Main ( main ) where + +import Data.Aeson +import Data.Aeson.Encode.Pretty +import Data.Kanji +import qualified Data.Map.Strict as M +import Data.Maybe (isNothing) +import qualified Data.Set as S +import qualified Data.Text as T +import qualified Data.Text.IO as TIO +import Data.Text.Lazy.Builder (toLazyText) +import Lens.Micro +import Lens.Micro.Aeson +import Lens.Micro.Platform () +import Options.Applicative +import Protolude hiding (to) + +--- + +data Flags = Flags [Operation] (Either FilePath Text) deriving (Eq) + +data Operation = Unknowns | Density | Elementary + | Distribution | Average | Splits deriving (Eq) + +data Env = Env { _allKs :: [Kanji] + , _original :: Text } deriving Eq + +-- | Long, Short, Help +lsh l s h = long l <> short s <> help h + +flags :: Parser Flags +flags = Flags <$> operations <*> (file <|> japanese) + where file = Left <$> strOption (lsh "file" 'f' "Take input from a file") + japanese = Right <$> argument str (metavar "JAPANESE") + +operations :: Parser [Operation] +operations = catMaybes <$> ops + where ops = traverse optional + [ flag' Unknowns $ + lsh "unknowns" 'u' "Find Kanji whose Level couldn't be determined" + , flag' Density $ + lsh "density" 'd' "Find how much of the input is made of Kanji" + , flag' Elementary $ + lsh "elementary" 'e' "Find density of Kanji learnt in elementary school" + , flag' Distribution $ + lsh "leveldist" 'l' "Find the distribution of Kanji levels" + , flag' Average $ + lsh "average" 'a' "Find the average Level of all Kanji present" + , flag' Splits $ + lsh "splits" 's' "Show which Level each Kanji belongs to" ] + +-- | Shortcut for singleton objects +ob :: ToJSON v => Text -> v -> Value +ob k v = object [ k .= v ] + +averageLev :: Reader Env Value +averageLev = ob "average" . averageLevel <$> asks _allKs + +splits :: Reader Env Value +splits = ob "levelSplit" . S.foldl' f mempty . S.fromList <$> asks _allKs + where f a k = maybe a (\l -> M.insertWith (++) l [k] a) $ level k + +unknowns :: Reader Env Value +unknowns = ob "unknowns" . S.map _kanji . S.filter (isNothing . level) . S.fromList <$> asks _allKs + +distribution :: Reader Env Value +distribution = ob "distributions" . levelDist <$> asks _allKs + +density :: Reader Env Value +density = do + d <- kanjiDensity <$> asks (T.length . _original) <*> asks _allKs + pure $ ob "density" d + +elementaryDensity :: Reader Env Value +elementaryDensity = ob "elementary" . elementaryDen . levelDist <$> asks _allKs + +-- | All operations return JSON, to be aggregated into a master Object. +execOp :: Operation -> Reader Env Value +execOp Unknowns = unknowns +execOp Density = density +execOp Elementary = elementaryDensity +execOp Distribution = distribution +execOp Average = averageLev +execOp Splits = splits + +output :: Value -> IO () +output = putLText . toLazyText . encodePrettyToTextBuilder + +-- | Dispatch on each `Operation` given. Aggregates the resulting JSON. +work :: (Env, [Operation]) -> Value +work (e, os) = Object $ vals ^. each . _Object + where vals = runReader (traverse execOp os) e + +env :: Flags -> IO (Env, [Operation]) +env (Flags os inp) = case inp of + Right t -> pure (e t, os) + Left f -> (, os) . e <$> TIO.readFile f + where e t = Env (t ^.. each . to kanji . _Just) t + +main :: IO () +main = execParser opts >>= env >>= output . work + where opts = info (helper <*> flags) + (fullDesc <> header "nanq - Kanji analysis of Japanese text") diff --git a/test/Test.hs b/test/Test.hs new file mode 100644 --- /dev/null +++ b/test/Test.hs @@ -0,0 +1,42 @@ +{-# LANGUAGE ImplicitParams #-} +{-# LANGUAGE OverloadedStrings #-} +{-# LANGUAGE TypeApplications #-} + +module Main ( main ) where + +import Data.Aeson +import Data.Kanji +import Data.Kanji.Types +import qualified Data.Map.Strict as M +import Test.HUnit.Approx +import Test.Tasty +import Test.Tasty.HUnit + +--- + +main :: IO () +main = defaultMain suite + +suite :: TestTree +suite = testGroup "Unit Tests" + [ testGroup "Analysis" + [ + testCase "percentSpread totals 1.0" $ M.foldl' (+) 0 (percentSpread testKanji) @?~ 1.0 + , testCase "levelDist totals 1.0" $ M.foldl' (+) 0 (levelDist testKanji) @?~ 1.0 + , testCase "levelDist == adultDen" $ M.foldl' (+) 0 (levelDist testKanji) @?~ adultDen (levelDist testKanji) + ] + , testGroup "JSON" + [ + testCase "Empty String should fail to parse" $ assert . isError $ fromJSON @Kanji (String "") + , testCase "String w/ multiple Kanji should fail" $ assert . isError $ fromJSON @Kanji (String "泥棒猫") + , testCase "String w/ one Kanji should succeed" $ fromJSON (String "猫") @?= Success (Kanji '猫') + ] + ] + where ?epsilon = 0.001 + +testKanji :: [Kanji] +testKanji = map Kanji "本猫机鉛筆紙帽子包丁床天井壁光電気棚箱靴家階段戸屋根地面" + +isError :: Result a -> Bool +isError (Error _) = True +isError _ = False