Reader: word search problem

[jimoow]

Thank you for making the time to report a bug and make Aedict better!

Thank you for your interest in Aedict. Prior submitting bugs, please make sure that you have newest Aedict from Google Play.

A screenshot is often worth a thousand words, so please make sure to include one ;)

Please provide:

1. Your phone brand and maker Freetel RAIJIN FTJ162E

2. The exact Android version. 7.0 Security patch level 5 Aug 2017

3. A screenshot if possible; even better, a youtube video demonstrating the issue.

4. A full description of the issue you're experiencing.

Word search by aedict reader sometimes not useful results and inconsistent with aedict. I will add more examples but one might be enough... text contains 見直し Select 見直 Reader finds two results = 見 and 直 Select 見直し Reader finds one result = 見直し Search aedict for 見直 Aedict will find 見直す、見直し

Happy if this is marked "needs more info" and I will add more examples.

---

Want to back this issue? Post a bounty on it! We accept bounties via Bountysource.

[mvysny]

The issue here is that Aedict by default searches in the "starts with" mode. So if you search for 見直 in Aedict with "starts-with", Aedict will indeed find 見直す、見直し. However, Reader can't do that since it must only lookup words that are exactly present in the text as-is. Therefore, if the trailing し is missing in the copied text, Reader searches for 見直 in "exact" mode, it doesn't find anything and it thus breaks the text into two kanjis.

You can reproduce this behaviour in Aedict as well: please try to search for 見直 in Aedict with "Exact" filtering option. You can open the filter popup in Aedict by pressing the upper-right filtering icon (looks a bit like a tree upside-down :-) )

Please let me know if this explanation makes sense ;)

[jimoow]

Thank you for the explanation which makes sense. I still think there is a problem but I cant describe it properly. I can file another issue if I work it out.

Consider another example, if I lookup 超え from this text 政府は『いざなみ景気超え』を強調し Reader =超(ちょう)+え If I lookup 超 Reader returns a useful result including 超える. Looking at the preceding two examples in Reader, to get a useful result in one case I must INCLUDE the trailing kana (し), in the other case I must EXCLUDE the trailing kana (え). I will need to think about it some more.

[mvysny]

If I lookup 超 Reader returns a useful result including 超える.

If you're only looking up one single kanji, Aedict returns info from Kanjidic (a directory of kanjis). Kanjidic is formatted a bit differently than JMDict (a word dictionary) and contains more information about individual kanjis. Kanjidic is formatted as follows: ko.eru means that the kanji itself is read as ko but as a word it's usually present as 超える.

Consider another example, if I lookup 超え Reader will return 超(ちょう)+え

In this mode Aedict will try to search for 超え and it fails to to deinflect it to 超える (please file a separate bug for this); thus it breaks it up into two "words" 超 and え and looks them up in the JMDict dictionary. In this mode, Aedict prefers results from JMDict since it's trying to break sentence into words. Since JMDict also coincidentally contains an entry for 超, it is included in the analysis (instead of a Kanjidic entry for 超).

You can toggle the dictionary in Aedict as well: just open the "filter" popup and check "Search in:":

• Dict means JMDict
• Kanjidic is Kanjidic
• Examples is the Tatoeba Examples dictionary.

You can find more about Kanjidic here: http://www.edrdg.org/wiki/index.php/KANJIDIC_Project ; more about JMDict can be found at http://www.edrdg.org/jmdict/edict_doc.html