cjkvi / cjkvi-dict Goto Github PK

In swjz.xml, most seal head entries refer the image files, like,
嘏
but, there is an exception. "十" is described as
十

looking at
http://archive.wul.waseda.ac.jp/kosho/ho04/ho04_00026/ho04_00026_0003/ho04_00026_0003_p0009.jpg
we can understand as the image file for "十" would be "088b-07".

But crawling the image files at kanji-database.sourceforge.net, it seems that
there is no "088b-07.png" (or "088b-07s.png"). I guess, XML was manipulated
after the cropping work, and the author could not find the corresponding seal
image, and leave img attribute as "nil". I think, the missing image file should
be supplied, and the nil-attribute problem should be fixed. How we can contribute
the missing image file?

In kanji-database.sourceforge.net, the seal glyphs cropped from 説文解字注 were hosted, like,
http://kanji-database.sourceforge.net/dict/swjz/swjz-img/088b-06.png (big)
http://kanji-database.sourceforge.net/dict/swjz/swjz-img/088b-06s.png (small)
but it seems that they are not migrated to github. Is it possible to provide them via github repository?

I created page numbers for the Morohashi and after careful checking noticed a bunch of differences between my data and the data in dkw2ucs.txt-file. The ones relating to the *.1 numbers (i.e.: like number 01368.1 or 13681' in Morohashi) are probably due to differences in editions. I have the 1984 edition (the third revision), yours is probably based and the 1st or 2nd edition. Everything else is probably an error in dkw2ucs.txt. I included a dump with the following format:
- .
checkmorohashi2a.txt

If you want the data in any other format, please let me know. I also included the mysqldump of the relevant table in my dbase.
moro_order.sql.gz

Kind regards,

Reinaert

The description says it all. The characters marked # missing shouldn't have a DP-field.