IDSgrep assumes UTF-8 throughout. When multiple files are specified, so that output lines are to be prefixed with the filenames enclosed in colons, idsgrep runs the filenames through its escaping code, in order to make sure that the result will be valid EIDS syntax. (In particular, literal colons in the filenames have to be escaped.) This is a problem if running on a non-UTF-8 host, where the filenames might possibly contain forbidden bytes.
We should figure out what happens in this case, and (harder) what SHOULD happen. If possible, it would be preferable to be able to keep the current "we are blind to locales" policy; but without knowing the locale, it's probably not going to be possible to translate the filenames to Unicode in a way that will survive a round trip.
mskala 
作りましょうのフォントファミリとＩＤＳｇｒｅｐ
IDSgrep assumes UTF-8 throughout. When multiple files are specified, so that output lines are to be prefixed with the filenames enclosed in colons, idsgrep runs the filenames through its escaping code, in order to make sure that the result will be valid EIDS syntax. (In particular, literal colons in the filenames have to be escaped.) This is a problem if running on a non-UTF-8 host, where the filenames might possibly contain forbidden bytes.
We should figure out what happens in this case, and (harder) what SHOULD happen. If possible, it would be preferable to be able to keep the current "we are blind to locales" policy; but without knowing the locale, it's probably not going to be possible to translate the filenames to Unicode in a way that will survive a round trip.