SugarMemo Rating算法

目標:
通過人性化的算法, 用戶只需要按 Fail/Pass/Bingo，
由系統自動轉換成sm2算法中的 A/B/C/D/E/F

變量說明:
Bn, 第N次按鈕選擇, Bn = Fail/Pass/Bingo
{noRating, Bingo, Pass, Fail} = {0, 1, 2, 3)
我傾向于
{noRating, Fail, Pass, Bingo} = {0, 1, 2, 3)

ts.1, ts.firstRating 的簡寫, ts.1 = Fail/Pass/Bingo
ts.2, ts.secondRating 的簡寫, ts.2 = Fail/Pass/Bingo
ts.1 記錄復習的第一次Bn
ts.2 記錄復習的第一次之后的Bn中最小的一次

現有規則, Bn = Pass/Bingo, 則退出復習, 同時進行 Rating
由于復習可以選模式,使得Rating的判斷相對復雜, 并且還有多選模式

考慮最簡單的情況, 只有一種復習, 比如說回想釋義
這種情況下 Bn = Pass/Bingo 的時候, 就只有 ts.1, 不會有 ts.2
這時, 現有的Rating模式實際上只可能Rating成 B,C
這應該來說不太合理

我來說說我的想法, 應該主要依據 ts.1 來判斷
ts.1 = Bingo, Rate = A
ts.1 = Pass,  Rate = B

totalFail = 1
ts.1 = Fail, ts.2 = Bingo, Rate = B
ts.1 = Fail, ts.2 = Pass,  Rate = C

totalFail = 2
ts.1 = Fail, ts.2 = Bingo, Rate = C
ts.1 = Fail, ts.2 = Pass,  Rate = D

totalFail > 2 and totalFail < MAXFAIL
ts.1 = Fail, ts.2 = Bingo, Rate = D
ts.1 = Fail, ts.2 = Pass,  Rate = E

totalFail >= MAXFAIL, Rate = F

再考慮稍微復雜一點的情況, 同時回想釋義和單詞
ts.1 = Bingo, ts.2 = Bingo, Rate = A

ts.1 = Bingo, ts.2 = Pass, Rate = B
ts.1 = Pass, ts.2 = Bingo, Rate = B

ts.1 = Pass, ts.2 = Pass, Rate = C

ts.1 = Fail 和以上算法一樣

如果再包含多選模式, Rating 的時候就更加難以合理了