I posted version 0.3 of IDSgrep last night; follow the link to download it from SourceForge.JP. As you may recall, IDSgrep is my kanji structural-query software. Some of the ideas behind it were discussed on this Web log back in December. The general idea is that this is software to answer queries about layout and visual components of Han-script (Japanese, Chinese, etc.) characters.
The main new things in version 0.3 are support for regular expression searches; the inclusion of a bundled dictionary (based on the IDS decompositions of the CHISE project); and a "cooked" output mode.
At this point I think IDSgrep as such - the search program - is basically finished. As bug reports and practical experience accumulate, there may eventually be a 0.4 or 1.0 release, but all the features I think it needs to have are now in place and it seems to serve pretty well the original purpose for which I needed to develop it.