Koji Ishii-san said that there are some issues about "find in page" vs. <ruby>. A word in <ruby> may be separated by ruby-text. For example: <ruby>T<rt>たまご</rt>K<rt>かけ</rt>G<rt>ごはん</rt></ruby> this will be rendered as (after bug 256274 and bug 33339): たまごかけごはん T K G Then, users cannot find the original word "TKG" by find. Furthermore, even if web pages use simplest ruby style, such as: <ruby>TKG<rt>たまごかけごはん</rt></ruby>を食べた In this case, users cannot find "TKGを". I guess <ruby> will be used in many pages after we implement it by Shiozawa-san's work (bug 256274) because we're last implementer in major web browser vendors. Then, this might become major issue of find. In the latter case, there is a very difficult issue. Users should be able to find both "TKGを" and "ごはんを". So, there are parallel text content around ruby element. I have no idea about the way to find from both text. Do you have good ideas about these issues?
I think it should be done in http://dxr.mozilla.org/mozilla-central/source/embedding/components/find/nsFind.cpp
You need to log in before you can comment on or make changes to this bug.