[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[elpa] externals/pyim 7e8937010c 4/4: Merge pull request #432 from tumas
|
From: |
ELPA Syncer |
|
Subject: |
[elpa] externals/pyim 7e8937010c 4/4: Merge pull request #432 from tumashu/count-log |
|
Date: |
Thu, 6 Jan 2022 22:57:48 -0500 (EST) |
branch: externals/pyim
commit 7e8937010ca04224672d3ac97f3a58b0f6ca0607
Merge: 19a18b414b 1c864717d5
Author: tumashu <tumashu@163.com>
Commit: GitHub <noreply@github.com>
Merge pull request #432 from tumashu/count-log
Count log
---
pyim-candidates.el | 89 +++++++---------------
pyim-dcache.el | 1 +
pyim-dhashcache.el | 215 ++++++++++++++++++++++++++++++++++++++++++++--------
pyim-dregcache.el | 19 ++---
pyim-process.el | 1 -
tests/pyim-tests.el | 44 ++++++-----
6 files changed, 240 insertions(+), 129 deletions(-)
diff --git a/pyim-candidates.el b/pyim-candidates.el
index 063c733a4d..2d93cbb95a 100644
--- a/pyim-candidates.el
+++ b/pyim-candidates.el
@@ -52,29 +52,13 @@
细节信息请参考 `pyim-page-refresh' 的 docstring.")
-(defvar pyim-candidates-possible-chiefs nil
- "可能做第一位候选词的词条列表。")
-
(pyim-register-local-variables
'(pyim-candidates pyim-candidate-position))
;; ** 获取备选词列表
-(defun pyim-candidates-create-weight-table (words)
- "基于 WORDS 的先后顺序，创建一个用于候选词排序的 count 权重表。
-count 和 count 权重结合起来确定词条的先后顺序。"
- (let ((table (make-hash-table :test #'equal))
- ;; FIXME: 这个权重列表是想当然的数字，因为目前我也不知道这个合理的权重是
- ;; 什么，希望以后通过实际使用，可以总结出更合理的数字。
- (weights (list 1.3 1.2 1.1)))
- (dolist (weight weights)
- (let ((word (pop words)))
- (when word
- (puthash word weight table))))
- table))
-
-(defun pyim-candidates-sort (candidates &optional weight-table)
+(defun pyim-candidates-sort (candidates)
"对 CANDIDATES 进行排序。"
- (pyim-dcache-call-api 'sort-words candidates nil weight-table))
+ (pyim-dcache-call-api 'sort-words candidates))
(defun pyim-candidates-create (imobjs scheme-name &optional async)
"按照 SCHEME-NAME 对应的输入法方案， 从输入法内部对象列表:
@@ -85,45 +69,32 @@ IMOBJS 获得候选词条。"
(funcall (intern (format "pyim-candidates-create:%S" class))
imobjs scheme-name async)))))
-(defun pyim-candidates-add-possible-chief (word)
- "将 WORD 添加到 `pyim-candidates-possible-chiefs'."
- (push word pyim-candidates-possible-chiefs)
- (setq pyim-candidates-possible-chiefs
- (cl-subseq pyim-candidates-possible-chiefs 0
- (min 100 (length pyim-candidates-possible-chiefs)))))
-
(defun pyim-candidates-get-chief (scheme-name &optional personal-words
common-words)
"选取第一位候选词。"
- (let* ((class (pyim-scheme-get-option scheme-name :class))
- (words pyim-candidates-possible-chiefs)
- (length (length words))
- ;; NOTE: 网上传言，一段话平均70个字，按照一个词两个字估算，100个词大概
- ;; 为两段话。
- (words100 (cl-subseq words 0 (min 100 length)))
- ;; NOTE: 10个词大概1句话。
- (words10 (cl-subseq words 0 (min 10 length))))
- (cond
- ((equal class 'xingma)
- (or
- ;; 如果从公共词库里面获取到的第一个词条是汉字，就选择它。
- (when (= (length (car common-words)) 1)
- (car common-words))
- ;; 从个人词库里面按排列的先后顺序，获取一个汉字。
- (cl-find-if
- (lambda (word)
- (= (length word) 1))
- personal-words)))
- (t (or
- ;; 最近输入的10个词中出现一次以上。
- (cl-find-if (lambda (word)
- (> (cl-count word words10 :test #'equal) 1))
- personal-words)
- ;; 最近输入的100个词中出现过三次以上。
- (cl-find-if (lambda (word)
- (> (cl-count word words100 :test #'equal) 3))
- personal-words)
- ;; 个人词条中的第一个词。
- (car personal-words))))))
+ (let ((class (pyim-scheme-get-option scheme-name :class)))
+ (cond ((equal class 'xingma)
+ (or
+ ;; 如果从公共词库里面获取到的第一个词条是汉字，就选择它。
+ (when (= (length (car common-words)) 1)
+ (car common-words))
+ ;; 从个人词库里面按排列的先后顺序，获取一个汉字。
+ (cl-find-if
+ (lambda (word)
+ (= (length word) 1))
+ personal-words)))
+ (t (or
+ ;; 最近输入的10个不同的词中出现一次以上。
+ (cl-find-if
+ (lambda (word)
+ (> (or (car (pyim-dcache-get word 'iword2count-recent1)) 0)
1))
+ personal-words)
+ ;; 最近输入的50个不同的词中出现过三次以上。
+ (cl-find-if
+ (lambda (word)
+ (> (or (car (pyim-dcache-get word 'iword2count-recent2)) 0)
3))
+ personal-words)
+ ;; 个人词条中的第一个词。
+ (car personal-words))))))
(defun pyim-candidates-create:xingma (imobjs scheme-name &optional async)
"`pyim-candidates-create' 处理五笔仓颉等形码输入法的函数."
@@ -158,10 +129,7 @@ IMOBJS 获得候选词条。"
;; 1. 第一个词选择公共词库中的第一个词。
;; 2. 剩下的分成常用字和词，常用字优先排，字和词各按 count 大小排序。
(let* ((personal-words (pyim-dcache-get last-code
'(icode2word)))
- (weight-table (pyim-candidates-create-weight-table
personal-words))
- (personal-words
- (pyim-candidates-sort
- personal-words weight-table))
+ (personal-words (pyim-candidates-sort personal-words))
(common-words (pyim-dcache-get last-code '(code2word)))
(chief-word (pyim-candidates-get-chief scheme-name
personal-words common-words))
(common-words (pyim-candidates-sort common-words
weight-table))
@@ -314,8 +282,7 @@ IMOBJS 获得候选词条。"
;; 个人词条排序：使用词频信息对个人词库得到的候选词排序，第一个词条的位置
;; 比较特殊，不参与排序，具体原因请参考 `pyim-page-select-word' 中的
;; comment.
- (setq weight-table (pyim-candidates-create-weight-table personal-words))
- (setq personal-words (pyim-candidates-sort personal-words weight-table))
+ (setq personal-words (pyim-candidates-sort personal-words))
(setq chief-word (pyim-candidates-get-chief scheme-name personal-words))
;; 调试输出
diff --git a/pyim-dcache.el b/pyim-dcache.el
index 6c8f537714..343af278b4 100644
--- a/pyim-dcache.el
+++ b/pyim-dcache.el
@@ -220,6 +220,7 @@ non-nil，文件存在时将会提示用户是否覆盖，默认为覆盖模式"
如果 FORCE 为真，强制加载。"
(pyim-dcache-init-variables)
(when pyim-dcache-auto-update
+ (pyim-dcache-call-api 'update-iword2priority force)
(pyim-dcache-call-api 'update-personal-words force)
(let* ((dict-files (mapcar (lambda (x)
(unless (plist-get x :disable)
diff --git a/pyim-dhashcache.el b/pyim-dhashcache.el
index 14d42c02de..0a519c9015 100644
--- a/pyim-dhashcache.el
+++ b/pyim-dhashcache.el
@@ -40,35 +40,92 @@
(require 'pyim-dcache)
(require 'pyim-scheme)
+(defvar pyim-dhashcache-count-types
+ '((day
+ ;; 用于生成类似 :20220206 这样的 key.
+ :format ":%Y%m%d"
+ ;; 最多保存七天 count 到缓存。
+ :max-save-length 7
+ ;; 计算排序综合指标时，最近七天 count 对应的权重。
+ :weights (0.396 0.245 0.151 0.094 0.057 0.038 0.019)
+ ;; 获取前一天需要减去的天数。
+ :delta -1
+ ;; 计算日平均 count 需要乘的数字。
+ :factor 0.143))
+ "计算排序综合指数时，用到的基本信息。")
+
(defvar pyim-dhashcache-code2word nil)
(defvar pyim-dhashcache-code2word-md5 nil)
(defvar pyim-dhashcache-word2code nil)
(defvar pyim-dhashcache-iword2count nil)
+(defvar pyim-dhashcache-iword2count-log nil)
+(defvar pyim-dhashcache-iword2count-recent1 nil)
+(defvar pyim-dhashcache-iword2count-recent2 nil)
+(defvar pyim-dhashcache-iword2priority nil)
(defvar pyim-dhashcache-shortcode2word nil)
(defvar pyim-dhashcache-icode2word nil)
(defvar pyim-dhashcache-ishortcode2word nil)
(defvar pyim-dhashcache-update-shortcode2word-p nil)
(defvar pyim-dhashcache-update-ishortcode2word-p nil)
(defvar pyim-dhashcache-update-icode2word-p nil)
+(defvar pyim-dhashcache-update-iword2priority-p nil)
(defvar pyim-dhashcache-update-code2word-running-p nil)
-(defun pyim-dhashcache-sort-words (words-list &optional iword2count
count-weight-table)
- "对 WORDS-LIST 排序，词频大的排在前面.
-
-如果 IWORD2COUNT 为 nil, 排序将使用 `pyim-dhashcache-iword2count'
-中记录的词频信息
-
-COUNT-WEIGHT-TABLE 是一个哈希表，保存词条的 count 权重，在排序过
-程中， ‘count * 权重’ 的取值决定了排序先后顺序, 权重是一个不小于1
-的数字。"
- (let ((iword2count (or iword2count pyim-dhashcache-iword2count))
- (count-weight-table (or count-weight-table (make-hash-table :test
#'equal))))
+(defun pyim-dhashcache-sort-words (words-list)
+ "对 WORDS-LIST 排序"
+ (let ((iword2count pyim-dhashcache-iword2count)
+ (iword2priority pyim-dhashcache-iword2priority))
(sort words-list
(lambda (a b)
- (> (* (or (gethash a iword2count) 0)
- (or (gethash a count-weight-table) 1))
- (* (or (or (gethash b iword2count) 0)
- (or (gethash b count-weight-table) 1))))))))
+ (let ((n1 (or (gethash a iword2priority) 0))
+ (n2 (or (gethash b iword2priority) 0)))
+ (if (= n1 n2)
+ (let ((n3 (or (gethash a iword2count) 0))
+ (n4 (or (gethash b iword2count) 0)))
+ (> n3 n4))
+ (> n1 n2)))))))
+
+(defun pyim-dhashcache-get-counts-from-log (log-info &optional time)
+ "从 LOG-INFO 中获取所有的 count 值。
+
+比如： ((day :20220205 10
+ :20220204 6 => ((day 10 6 0 3 ...))
+ :20220202 3
+ ...))"
+ (mapcar (lambda (x)
+ (let* ((label (car x))
+ (plist (cdr x))
+ (format (plist-get plist :format))
+ (n (plist-get plist :max-save-length))
+ (delta (plist-get plist :delta))
+ (time (or time (current-time)))
+ output)
+ (dotimes (i n)
+ (let* ((time (time-add time (days-to-time (* i delta))))
+ (key (intern (format-time-string format time)))
+ (plist (cdr (assoc label log-info))))
+ (push (or (plist-get plist key) 0) output)))
+ `(,label ,@(reverse output))))
+ pyim-dhashcache-count-types))
+
+(defun pyim-dhashcache-calculate-priority (counts-info)
+ "根据 COUNTS-INFO 计算一个优先级指标，用于对词条进行排序。
+COUNTS-INFO 是一个 alist, 其结构类似：
+
+ ((day n1 n2 n3 ...))
+
+其中 (n1 n2 n3 ...) 代表从当前日期逐日倒推，每日 count 所组成的列表。"
+ (apply #'+ (mapcar (lambda (x)
+ (let* ((label (car x))
+ (plist (cdr x))
+ (weights (plist-get plist :weights))
+ (factor (plist-get plist :factor)))
+ (* (apply #'+ (cl-mapcar (lambda (a b)
+ (* (or a 0) b))
+ (cdr (assoc label
counts-info))
+ weights))
+ factor)))
+ pyim-dhashcache-count-types)))
(defun pyim-dhashcache-get-shortcodes (code)
"获取 CODE 所有的 shortcodes.
@@ -124,16 +181,15 @@ COUNT-WEIGHT-TABLE 是一个哈希表，保存词条的 count 权重，在排序
,@(pyim-dhashcache-async-inject-variables)
(require 'pyim-dhashcache)
(pyim-dcache-init-variable pyim-dhashcache-icode2word)
- (pyim-dcache-init-variable pyim-dhashcache-iword2count)
+ (pyim-dhashcache-init-count-and-priority-variables)
(pyim-dcache-save-variable
'pyim-dhashcache-ishortcode2word
(pyim-dhashcache-update-ishortcode2word-1
- pyim-dhashcache-icode2word
- pyim-dhashcache-iword2count)))
+ pyim-dhashcache-icode2word)))
(lambda (_)
(pyim-dcache-reload-variable pyim-dhashcache-ishortcode2word)))))
-(defun pyim-dhashcache-update-ishortcode2word-1 (icode2word iword2count)
+(defun pyim-dhashcache-update-ishortcode2word-1 (icode2word)
"`pyim-dhashcache-update-ishortcode2word' 内部函数."
(let ((ishortcode2word (make-hash-table :test #'equal)))
(maphash
@@ -147,7 +203,7 @@ COUNT-WEIGHT-TABLE 是一个哈希表，保存词条的 count 权重，在排序
icode2word)
(maphash
(lambda (key value)
- (puthash key (pyim-dhashcache-sort-words value iword2count)
+ (puthash key (pyim-dhashcache-sort-words value)
ishortcode2word))
ishortcode2word)
ishortcode2word))
@@ -166,16 +222,15 @@ COUNT-WEIGHT-TABLE 是一个哈希表，保存词条的 count 权重，在排序
,@(pyim-dhashcache-async-inject-variables)
(require 'pyim-dhashcache)
(pyim-dcache-init-variable pyim-dhashcache-code2word)
- (pyim-dcache-init-variable pyim-dhashcache-iword2count)
+ (pyim-dhashcache-init-count-and-priority-variables)
(pyim-dcache-save-variable
'pyim-dhashcache-shortcode2word
(pyim-dhashcache-update-shortcode2word-1
- pyim-dhashcache-code2word
- pyim-dhashcache-iword2count)))
+ pyim-dhashcache-code2word)))
(lambda (_)
(pyim-dcache-reload-variable pyim-dhashcache-shortcode2word)))))
-(defun pyim-dhashcache-update-shortcode2word-1 (code2word iword2count)
+(defun pyim-dhashcache-update-shortcode2word-1 (code2word)
"`pyim-dhashcache-update-shortcode2word' 的内部函数"
(let ((shortcode2word (make-hash-table :test #'equal)))
(maphash
@@ -197,7 +252,7 @@ COUNT-WEIGHT-TABLE 是一个哈希表，保存词条的 count 权重，在排序
code2word)
(maphash
(lambda (key value)
- (puthash key (pyim-dhashcache-sort-words value iword2count)
+ (puthash key (pyim-dhashcache-sort-words value)
shortcode2word))
shortcode2word)
shortcode2word))
@@ -349,10 +404,10 @@ code 对应的中文词条了。
,@(pyim-dhashcache-async-inject-variables)
(require 'pyim-dhashcache)
(pyim-dcache-init-variable pyim-dhashcache-icode2word)
- (pyim-dcache-init-variable pyim-dhashcache-iword2count)
+ (pyim-dhashcache-init-count-and-priority-variables)
(maphash
(lambda (key value)
- (puthash key (pyim-dhashcache-sort-words value
pyim-dhashcache-iword2count)
+ (puthash key (pyim-dhashcache-sort-words value)
pyim-dhashcache-icode2word))
pyim-dhashcache-icode2word)
(pyim-dcache-save-variable
@@ -404,22 +459,38 @@ code 对应的中文词条了。
(directory-files pyim-dcache-directory nil "-backup-"))
(message "PYIM: 在 %S 目录中发现备份文件的存在，可能是词库缓存文件损坏导致，请抓紧检查处理！！！"
pyim-dcache-directory))
- (pyim-dcache-init-variable pyim-dhashcache-iword2count)
+ (pyim-dhashcache-init-count-and-priority-variables)
(pyim-dcache-init-variable pyim-dhashcache-code2word)
(pyim-dcache-init-variable pyim-dhashcache-word2code)
(pyim-dcache-init-variable pyim-dhashcache-shortcode2word)
(pyim-dcache-init-variable pyim-dhashcache-icode2word)
(pyim-dcache-init-variable pyim-dhashcache-ishortcode2word))
+(defun pyim-dhashcache-init-count-and-priority-variables ()
+ "初始化 count 相关的变量。"
+ (pyim-dcache-init-variable pyim-dhashcache-iword2count)
+ (pyim-dcache-init-variable pyim-dhashcache-iword2count-log)
+ (pyim-dcache-init-variable pyim-dhashcache-iword2count-recent1)
+ (pyim-dcache-init-variable pyim-dhashcache-iword2count-recent2)
+ (pyim-dcache-init-variable pyim-dhashcache-iword2priority))
+
(defun pyim-dhashcache-save-personal-dcache-to-file ()
;; 用户选择过的词
(pyim-dcache-save-variable
'pyim-dhashcache-icode2word
pyim-dhashcache-icode2word 0.8)
- ;; 词频
+ ;; 词条总 count
(pyim-dcache-save-variable
'pyim-dhashcache-iword2count
- pyim-dhashcache-iword2count 0.8))
+ pyim-dhashcache-iword2count 0.8)
+ ;; 词条 count 日志
+ (pyim-dcache-save-variable
+ 'pyim-dhashcache-iword2count-log
+ pyim-dhashcache-iword2count-log 0.8)
+ ;; 词条优先级
+ (pyim-dcache-save-variable
+ 'pyim-dhashcache-iword2priority
+ pyim-dhashcache-iword2priority 0.8))
(defmacro pyim-dhashcache-put (cache code &rest body)
"将 BODY 的返回值保存到 CACHE 对应的 CODE 中。
@@ -437,8 +508,34 @@ code 对应的中文词条了。
(setq ,new-value (progn ,@body))
(puthash ,key ,new-value ,table))))
+(defun pyim-dhashcache-update-iword2count-recent (word n hash-table)
+ (let (words-need-remove)
+ (pyim-dhashcache-put
+ hash-table :all-words
+ (setq orig-value (remove word orig-value))
+ (push word orig-value)
+ (if (<= (length orig-value) n)
+ orig-value
+ (setq words-need-remove (nthcdr n orig-value))
+ (cl-subseq orig-value 0 n)))
+ (dolist (w words-need-remove)
+ (remhash w hash-table))
+ (pyim-dhashcache-put
+ hash-table word
+ (+ (or orig-value 0) 1))
+ hash-table))
+
(defun pyim-dhashcache-update-iword2count (word &optional wordcount-handler)
"保存词频到缓存."
+ ;; 更新最近输入 10 个词条的 count 表
+ (setq pyim-dhashcache-iword2count-recent1
+ (pyim-dhashcache-update-iword2count-recent
+ word 10 pyim-dhashcache-iword2count-recent1))
+ ;; 更新最近输入 50 个词条的 count 表
+ (setq pyim-dhashcache-iword2count-recent2
+ (pyim-dhashcache-update-iword2count-recent
+ word 50 pyim-dhashcache-iword2count-recent2))
+ ;; 更新总 count 表
(pyim-dhashcache-put
pyim-dhashcache-iword2count word
(cond
@@ -446,7 +543,59 @@ code 对应的中文词条了。
(funcall wordcount-handler (or orig-value 0)))
((numberp wordcount-handler)
wordcount-handler)
- (t (or orig-value 0)))))
+ (t (or orig-value 0))))
+ ;; 更新 count 日志表。
+ (pyim-dhashcache-put
+ pyim-dhashcache-iword2count-log word
+ (let (out)
+ (dolist (x pyim-dhashcache-count-types)
+ (let* ((label (car x))
+ (key (intern (format-time-string (plist-get (cdr x) :format))))
+ (n (plist-get (cdr x) :max-save-length))
+ (plist (cdr (assoc label orig-value)))
+ (value (plist-get plist key))
+ (output (if value
+ (plist-put plist key (+ 1 value))
+ (append (list key 1) plist)))
+ (length (length output))
+ (output (cl-subseq output 0 (min length (* 2 n)))))
+ (push `(,label ,@output) out)))
+ out))
+ ;; 更新优先级表
+ (pyim-dhashcache-put
+ pyim-dhashcache-iword2priority word
+ ;; Fix warn
+ (ignore orig-value)
+ (pyim-dhashcache-calculate-priority
+ (pyim-dhashcache-get-counts-from-log
+ (gethash word pyim-dhashcache-iword2count-log)))))
+
+(defun pyim-dhashcache-update-iword2priority (&optional force)
+ "更新词条优先级表，如果 FORCE 为真，强制更新。"
+ (interactive)
+ (when (or force (not pyim-dhashcache-update-iword2priority-p))
+ ;; NOTE: 这个变量按理说应该在回调函数里面设置，但 async 在某些情况下会卡死，
+ ;; 这个变量无法设置为 t, 导致后续产生大量的 emacs 进程，极其影响性能。
+ (setq pyim-dhashcache-update-iword2priority-p t)
+ (async-start
+ `(lambda ()
+ ,@(pyim-dhashcache-async-inject-variables)
+ (require 'pyim-dhashcache)
+ (pyim-dhashcache-init-count-and-priority-variables)
+ (maphash
+ (lambda (key value)
+ (puthash key
+ (pyim-dhashcache-calculate-priority
+ (pyim-dhashcache-get-counts-from-log
+ value))
+ pyim-dhashcache-iword2priority))
+ pyim-dhashcache-iword2count-log)
+ (pyim-dcache-save-variable
+ 'pyim-dhashcache-iword2priority
+ pyim-dhashcache-iword2priority)
+ nil)
+ (lambda (_)
+ (pyim-dcache-reload-variable pyim-dhashcache-iword2priority)))))
(defun pyim-dhashcache-delete-word (word)
"将中文词条 WORD 从个人词库中删除"
@@ -467,7 +616,9 @@ code 对应的中文词条了。
(puthash key new-value pyim-dhashcache-ishortcode2word)
(remhash key pyim-dhashcache-ishortcode2word)))))
pyim-dhashcache-ishortcode2word)
- (remhash word pyim-dhashcache-iword2count))
+ (remhash word pyim-dhashcache-iword2count)
+ (remhash word pyim-dhashcache-iword2count-log)
+ (remhash word pyim-dhashcache-iword2priority))
(defun pyim-dhashcache-insert-word-into-icode2word (word code prepend)
"将词条 WORD 插入到 icode2word 词库缓存 CODE 键对应的位置.
diff --git a/pyim-dregcache.el b/pyim-dregcache.el
index 4a7825163c..94c7764cdf 100644
--- a/pyim-dregcache.el
+++ b/pyim-dregcache.el
@@ -67,24 +67,15 @@
(insert-file-contents file)
(buffer-string)))))
-(defun pyim-dregcache-sort-words (words-list &optional iword2count
count-weight-table)
- "对 WORDS-LIST 排序，词频大的排在前面.
-
-如果 IWORD2COUNT 为 nil, 排序将使用 `pyim-dregcache-iword2count'
-中记录的词频信息
-
-COUNT-WEIGHT-TABLE 是一个哈希表，保存词条的 count 权重，在排序过
-程中， ‘count * 权重’ 的取值决定了排序先后顺序。"
- (let ((iword2count (or iword2count pyim-dregcache-iword2count))
- (count-weight-table (or count-weight-table (make-hash-table :test
#'equal))))
+(defun pyim-dregcache-sort-words (words-list)
+ "对 WORDS-LIST 排序，词频大的排在前面."
+ (let ((iword2count pyim-dregcache-iword2count))
(sort words-list
(lambda (a b)
(let ((a (car (split-string a ":")))
(b (car (split-string b ":"))))
- (> (* (or (gethash a iword2count) 0)
- (or (gethash a count-weight-table) 1))
- (* (or (or (gethash b iword2count) 0)
- (or (gethash b count-weight-table) 1)))))))))
+ (> (or (gethash a iword2count) 0)
+ (or (gethash b iword2count) 0)))))))
(defun pyim-dregcache-sort-icode2word ()
"对个人词库排序."
diff --git a/pyim-process.el b/pyim-process.el
index 5bbcd6f6bc..7eb1131441 100644
--- a/pyim-process.el
+++ b/pyim-process.el
@@ -575,7 +575,6 @@ BUG：拼音无法有效地处理多音字。"
(codes (pyim-cstring-to-codes
word scheme-name
(or criteria pyim-cstring-to-code-criteria))))
- (pyim-candidates-add-possible-chief word)
;; 保存对应词条的词频
(when (> (length word) 0)
(pyim-dcache-update-wordcount word (or wordcount-handler #'1+)))
diff --git a/tests/pyim-tests.el b/tests/pyim-tests.el
index 2bec8f6707..ece136030d 100644
--- a/tests/pyim-tests.el
+++ b/tests/pyim-tests.el
@@ -742,8 +742,8 @@ zuo-zuo-you-mang 作作有芒")
(should (equal (gethash "啊啊" output2) nil))))
(ert-deftest pyim-tests-pyim-dhashcache-update-shortcode2word ()
- (let ((code2word (make-hash-table :test #'equal))
- (iword2count (make-hash-table :test #'equal))
+ (let ((pyim-dhashcache-iword2count (make-hash-table :test #'equal))
+ (code2word (make-hash-table :test #'equal))
(shortcode2word (make-hash-table :test #'equal))
output)
@@ -755,7 +755,7 @@ zuo-zuo-you-mang 作作有芒")
(puthash "wubi/aaae" '("𧝣") code2word)
(setq shortcode2word
- (pyim-dhashcache-update-shortcode2word-1 code2word iword2count))
+ (pyim-dhashcache-update-shortcode2word-1 code2word))
(should (equal (gethash "wubi/aa" shortcode2word)
'(#("工" 0 1 (:comment "a"))
@@ -769,8 +769,8 @@ zuo-zuo-you-mang 作作有芒")
#("𧝣" 0 1 (:comment "e")))))))
(ert-deftest pyim-tests-pyim-dhashcache-update-ishortcode2word ()
- (let ((icode2word (make-hash-table :test #'equal))
- (iword2count (make-hash-table :test #'equal))
+ (let ((pyim-dhashcache-iword2count (make-hash-table :test #'equal))
+ (icode2word (make-hash-table :test #'equal))
ishortcode2word)
(puthash "ni" '("你" "呢") icode2word)
@@ -778,8 +778,7 @@ zuo-zuo-you-mang 作作有芒")
(puthash "ni-huai" '("你坏") icode2word)
(setq ishortcode2word
- (pyim-dhashcache-update-ishortcode2word-1
- icode2word iword2count))
+ (pyim-dhashcache-update-ishortcode2word-1 icode2word))
(should (equal (gethash "n-h" ishortcode2word)
'("你好" "呢耗" "你坏")))
@@ -860,27 +859,30 @@ yin-xing 因行
(ert-deftest pyim-tests-pyim-dhashcache-sort-words ()
(let ((pyim-dhashcache-iword2count (make-hash-table :test #'equal))
- (weight-table (make-hash-table :test #'equal))
words)
(puthash "你好" 3 pyim-dhashcache-iword2count)
(puthash "呢耗" 2 pyim-dhashcache-iword2count)
(puthash "你豪" 1 pyim-dhashcache-iword2count)
- (puthash "你好" 0.1 weight-table)
- (puthash "呢耗" 0.3 weight-table)
- (puthash "你豪" 5 weight-table)
-
(setq words (list "呢耗" "你豪" "你好"))
(should (equal (pyim-dhashcache-sort-words words)
- '("你好" "呢耗" "你豪")))
-
- (setq words (list "呢耗" "你豪" "你好"))
- (should (equal (pyim-dhashcache-sort-words words
pyim-dhashcache-iword2count)
- '("你好" "呢耗" "你豪")))
-
- (setq words (list "呢耗" "你豪" "你好"))
- (should (equal (pyim-dhashcache-sort-words words nil weight-table)
- '("你豪" "呢耗" "你好")))))
+ '("你好" "呢耗" "你豪")))))
+
+(ert-deftest pyim-tests-pyim-dhashcache-get-counts-from-log ()
+ (should (member (pyim-dhashcache-get-counts-from-log
+ '((day :20220107 10
+ :20220106 6
+ :20220104 3
+ :20220103 3))
+ ;; (date-to-time "2022-01-07")
+ '(25047 4608))
+ '(((day 6 0 3 3 0 0 0)) ;Fixme: In github-ci will result
this value, why?
+ ((day 10 6 0 3 3 0 0))))))
+
+(ert-deftest pyim-tests-pyim-dhashcache-calculate-priority ()
+ (should (equal (pyim-dhashcache-calculate-priority
+ '((day 3 7 6 4 5 9 1)))
+ 0.690833)))
;; ** pyim-dregcache 相关单元测试
(ert-deftest pyim-tests-pyim-general ()