summaryrefslogtreecommitdiff
path: root/ext/mixed/js
diff options
context:
space:
mode:
authorsiikamiika <siikamiika@users.noreply.github.com>2019-11-03 05:08:57 +0200
committersiikamiika <siikamiika@users.noreply.github.com>2019-11-23 17:45:44 +0200
commit41020289ab68ef22a0691a9f268a79d6a706df6b (patch)
tree0cd10c38b37cc475dc306c5cf95e8a2e4247a98a /ext/mixed/js
parent3881457e4ed3f9c7833ac21a5e7fc44c2ba00b0f (diff)
add mecab support
Diffstat (limited to 'ext/mixed/js')
-rw-r--r--ext/mixed/js/japanese.js35
1 files changed, 32 insertions, 3 deletions
diff --git a/ext/mixed/js/japanese.js b/ext/mixed/js/japanese.js
index d24f56a6..78c419b2 100644
--- a/ext/mixed/js/japanese.js
+++ b/ext/mixed/js/japanese.js
@@ -61,12 +61,11 @@ function jpDistributeFurigana(expression, reading) {
const group = groups[0];
if (group.mode === 'kana') {
- if (reading.startsWith(group.text)) {
- const readingUsed = reading.substring(0, group.text.length);
+ if (jpKatakanaToHiragana(reading).startsWith(jpKatakanaToHiragana(group.text))) {
const readingLeft = reading.substring(group.text.length);
const segs = segmentize(readingLeft, groups.splice(1));
if (segs) {
- return [{text: readingUsed}].concat(segs);
+ return [{text: group.text}].concat(segs);
}
}
} else {
@@ -95,3 +94,33 @@ function jpDistributeFurigana(expression, reading) {
return segmentize(reading, groups) || fallback;
}
+
+function jpDistributeFuriganaInflected(expression, reading, source) {
+ const output = [];
+
+ let stemLength = 0;
+ const shortest = Math.min(source.length, expression.length);
+ const sourceHiragana = jpKatakanaToHiragana(source);
+ const expressionHiragana = jpKatakanaToHiragana(expression);
+ while (
+ stemLength < shortest &&
+ // sometimes an expression can use a kanji that's different from the source
+ (!jpIsKana(source[stemLength]) || (sourceHiragana[stemLength] === expressionHiragana[stemLength]))
+ ) {
+ ++stemLength;
+ }
+ const offset = source.length - stemLength;
+
+ for (const segment of jpDistributeFurigana(
+ source.slice(0, offset === 0 ? source.length : source.length - offset),
+ reading.slice(0, offset === 0 ? reading.length : reading.length - expression.length + stemLength)
+ )) {
+ output.push(segment);
+ }
+
+ if (stemLength !== source.length) {
+ output.push({text: source.slice(stemLength)});
+ }
+
+ return output;
+}