I am new to the community since I discovered the LanguageTool some weeks ago. First of all: Many thanks to Daniel and all you others for the great work!
Then I have got a question of which I hope you can help me: I want to add some own words to the dictionary. I guess there are three ways:
In the ignore.txt (If I do this the words won’t be marked wrong any more)
In the de_DE.dic (If I do this the words will be marked wrong nevertheless)
In the added.txt (If I do this the word’s won’t be marked wrong any more)
Hence, options 1 and 3 work. But unfortunately the words are not taken into the system’s consideration concerning suggestions in case of wrong words. If I add the word “Caspers” to the dictionary and I write “Casppers” in the text then the system should suggest me “Caspers”.
Could you give me your opinion concerning which way is propably the best and how I could realize the suggestion function? I searched into the forum but did not find an answer for this problem.
Am 21.05.2013 14:30, schrieb tjf [via LanguageTool User Forum]:
In the de_DE.dic (If I do this the words will be marked wrong
nevertheless)
This is the right place. Are you sure you added the word at the right
place (alphabetical order) and didn’t make any other mistakes? For
example, the file is not in UTF-8 encoding, unlike almost all other
files.
Yes, I inserted the word “Caspers” between “Caspar/S” and “Cassius”. If I now write “Caspers” in my text everything is fine. If I write “Casppers” I would expect that the system proposes “Caspers”. But it proposes “Korrektur: Jaspers; Campers; Kaspers; Camphers; Rappers; Caspars; Choppers”.
If I write “Capsers” it proposes “Japsers; Gipsers; Caesars; Hassers; Calwers; Hopsers; Campers; Japsens; Camphers; Japser; Capes; Japsern; Caspars; Jaspers; Kaisers; Kaspers; Lasers; Lausers; Papers; Papiers
Text: Capsers”.
Am 21.05.2013 20:18, schrieb tjf [via LanguageTool User Forum]:
Unfortunately, “Caspers” is no example given.
Sorry, I forgot about a special case we use for some languages:
although we make use of hunspell, we don’t use it for getting
suggestions because it’s too slow for that. The file used for
suggestions is de_DE.dict (note the ‘t’). This is a binary file for
faster lookups. You could create it from de_DE.dic using the
create_dict.sh, but that requires a Linux environment, the unmunch
command and maybe more. In other words, it’s quite brittle and I
wouldn’t recommend it.