<br>Hello ECL developers,<br><br>I am currently wrapping up the code of a Java JNI to Common Lisp interface,<br>a bit à la "jfli" but with a different twist to it. I would like to shortly announce it as<br>working on ECL (as well as SBCL) under Linux/x86 or MS-Windows XP (x86).<br>
<br>In order to do that I definitely need to have ECL support C99 type "long long"<br>on those platforms since it is the only proper mapping for Java's "long" type<br>in that context.<br><br>I hereby attach a patch file that I did against ECL 8.12 that adds the support<br>
I need. I hope it is thorough enough. BTW, I am a bit annoyed by the fact<br>that this patch seems to break fasl compatibility but I have a feeling that this<br>is pretty much unavoidable.<br><br>Would it be possible to merge this patch with the current ECL main code base?<br>
<br>One caveat: I have not implemented "long long" on lisp callbacks yet but<br>I have an idea as to how to do it simply and quickly. To make sure I would<br>need to know what Lisp type would be converted to "long long" on callback<br>
return. Is it only fixnum or am I missing something else?<br><br>P.S.: The :long-long in *features* is just a quick and dirty way to identify<br>that this instance of ECL has "long long" support. I don't care much about<br>
it as long as I have a proper way to know that "long long" support is there.<br>It could very well disappear in a future release of ECL if it is well known that<br>"long long" support is a feature of that release.<br>
<br>P.P.S.: I just do "patch -p0 <ecl-8.12.0_long_long_.patch" to apply the patch<br>against the root of a fresh ECL 8.12. <br><br>Cheers,<br><br>Jean-Claude Beaudoin<br><br>