| # This is a sample wordlist that can be converted to a binary dictionary |
| # for use by the Latin IME. |
| # The file is essentially a CSV file, with indent level denoting nesting. |
| # |
| # The file starts with a single CSV line with the header attributes. Whatever |
| # the content, these are included as is in the binary file. The first attribute |
| # of the file should be `dictionary'. Usual fields are `locale', `description', |
| # `date', `version', `options'. |
| # |
| # Each word has a `word' entry and at least a `f' argument denoting its |
| # probability, as an integer between 0 and 255 on a logarithmic scale, with |
| # 255 meaning 1 and each decrement in 1 dividing probability by 1.15. |
| # As a special case, a weight of 0 is taken to mean profanity - words that |
| # should not be considered a typo, but that should never be suggested |
| # explicitly. An entry may be made not a word by adding a `not_a_word' |
| # field with a value of `true'. The main reason for putting such entries |
| # into the dictionary is to add shortcut targets and maybe a whitelist |
| # replacement. |
| # |
| # Each word may or may not have any number of shortcut target lines |
| # starting with a `shortcut' entry and having at least a `f' frequency |
| # value between 0 and 14, or the special value `whitelist' which becomes |
| # 15, which is then taken to be the whitelist target of this word. |
| # |
| # Each word may also have any number of bigram lines starting with a |
| # `bigram' entry containing the following word whose frequency should |
| # override the unigram frequency when following the word this bigram is |
| # for. |
| # |
| dictionary=main:en,locale=en,description=Sample wordlist,date=1351495318,version=1 |
| word=sample,f=200 |
| bigram=wordlist,f=243 |
| word=wordlist,f=180 |
| word=shortcut,f=176 |
| shortcut=target,f=10 |
| word=witelisted,f=10,not_a_word=true |
| shortcut=whitelisted,f=whitelist |
| word=profanity,f=0 |
