11 releases
new 0.1.10 | Mar 28, 2025 |
---|---|
0.1.9 | Mar 17, 2025 |
0.1.5 | Feb 9, 2025 |
0.1.4 | Jan 31, 2025 |
0.1.0 | Dec 23, 2022 |
#291 in Text processing
550 downloads per month
65KB
1.5K
SLoC
furigana
Contains functionality for correctly mapping furigana to a word given a reading, optionally using kanji reading data.
Usage
for mapping in furigana::map_naive("物の怪", "もののけ") {
println!("{mapping}");
}
prints out the following mappings:
物の怪 物の怪
The second mapping is correct one, but based only on a word and its reading there's no way to determine that.
If given information about kanji readings (for example, from KANJIDIC2), furigana::map
is able to grade the potential mappings:
let mut kanji_to_readings = HashMap::new();
kanji_to_readings.insert("物".to_string(), vec!["もの".to_string()]);
kanji_to_readings.insert("怪".to_string(), vec!["け".to_string()]);
let mapping = furigana::map("物の怪", "もののけ", &kanji_to_readings)[0];
println!("{mapping}");
Here, the incorrect mapping is rejected using the knowledge given about kanji readings, so that only the correct mapping is printed:
物の怪
Notes
-
The algorithm used is recursive and not optimised, so it may be inefficient for very long inputs and certain edge cases that produce a large amount of potential mappings. When using real data and dividing it into shorter segments (e.g. by word or by sentence) there should be no issue.
-
Irregular readings such as おとな for 大人 and とおか 10日 are handled case by case so these may be mapped in correctly in some cases. Issues on these are appreciated.
-
If the library fails to produce the correct mapping, or if its accuracy is lower than that of an incorrect mapping, an issue is much appreciated!
License
Licensed under the Mozilla Public License Version 2.0.