PECL and generated files

As some of you may know, im working on a Unicode based I18N extension for PHP based on ICU (http://www-124.ibm.com/icu/) the results thus far are at http://www.pearfr.org/~cshmoove . the extension itself is progressing fine, but i have some questions concerning external data and packaging. ICU provides tools which converts data from Unicode.org (and elsewhere) into a binary form which is usable by the library. this includes character properties and names, locale data and timezone information. i was a bit optimistic in my assessment of the time i had/needed to convert these tools to PHP. keep in mind that ICU is primarily written in C++, and although i have recoded the implemented segments in C, ICU is a HUGE package with quite a few internal dependencies. im afraid after more than a year, i dont have the additional energy to move the tools over. so the question is : is it OK to have generated files included in the extension tarball or is it always required to be able to generate the binary data from the original ? my though is to use the tools as is on my platform to generate the files, then convert them into C byte array source code to be included inline. thanks in advance.. l0t3k