Menu ▾ ▴

#69 Wrong frequency information for some kanji

open
nobody
None
1
2019-12-25
2016-02-06
jchrbe
No

(Debian/Jessie)

When a kanji has no frequency information (Fnnn in kanji dictionary), the radical information can take a wrong information from another parameter, as DFnnn.

Ex:
斤 3654 U65a4 B69 G8 S4 J1 N2076 V2379 H2949 DK1871 L1125 K2012 DO1818 MN13534 MP5.0618 E1176 IN1897 DF1397 DT1034 DJ1914 DG953 DM1134 P3-2-2 I0a4.3 Q7222.1 DR2654 ZPP3-3-1 ZPP4-4-4 Yjin1 Wgeun キン T2 おの {axe} {1.32 lb} {catty} {counter for loaves of bread} {axe radical (no. 69)}

As no Fxxx exists, the radical information returns 1397 from DF1397, which doesn't mean any frequency use.

Discussion

  • jchrbe

    jchrbe - 2016-03-17

    As I'm not able to fix the problem in the sources (the last version doesn't compile on my system, lacking several libraries), I add ' F9999' in any line missing a ' Fnnn" (with a space before 'F'), quick and dirty fix!

    In a GNU/Linux Debian/GNOME, Kanji file is in /home/{usr}/.config/gwaei/dictionaries/kanji/ 

    #! /usr/bin/python

han=open("Kanji")
lines=han.readlines() # into a list
han.close()

for i in range(len(lines)):
  line=lines[i]
  if " F" not in line: # lacking ' Fnnn' sequence
    line=line.replace(" U"," F9999 U") # as ' U' exists in any line
    lines[i]=line

han=open("Kanji","w")
han.writelines(lines) # saving from the list
han.close()

    Note: copy and save your "Kanji" file before processing.

     

    Last edit: jchrbe 2016-03-17

  • jchrbe

    jchrbe - 2019-12-25

    Back to the problem...

    Gwaei 3.6.2-5b1 (stable) on Debian 10 Buter still has a problème for kanji-frequency. Using the kanji-files "Kanji", gwaei reads the first combination Fn(nn) of the kanji line, and not always the " Fnnn " expression. For instance,
    伊 ｜ ヨ 化 304B U4f0a B9 G9 S6 F703 J1... frequency will be 0 (UAf0a) instead of 703
    依 衣 化 亠 304D U4f9d B9 G8 S8 F906 J2... frequency will be 9 (U4f9d) instead of 906

    It should be easy to fix the bug, no?

     

Log in to post a comment.

MongoDB Logo MongoDB