2014-03-14 0:30 GMT+01:00 Alex Loukissas <alex at maginatics.com>: > I have verified that adding the said include fixes the issue (1-line patch): > > diff --git sqlite3.c sqlite3.c > index 7c6d3e2..a9399e4 100644 > --- sqlite3.c > +++ sqlite3.c > @@ -145656,6 +145656,7 @@ SQLITE_API int sqlite3_rtree_init( > #include <unicode/uregex.h> > #include <unicode/ustring.h> > #include <unicode/ucol.h> > +#include <unicode/utf8.h> > > /* #include <assert.h> */ > > Hope this helps others that run into this issue.
> This sounds strange to me (I compiled the SQLite ICU extension > for Cygwin, and didn't run into this issue). > According to the ICU documentation, <unicode/utypes.h> (which is > just above the <unicode/uregex.h>) includes <unicode/utf.h>, which > in turn includes <unicode/utf8.h>: > <http://icu.sourcearchive.com/documentation/4.2/utypes_8h.html> > So, the current order of includes should just work fine. > > What ICU version are you using? I used ICU 4.8 and 5.1, neither > of them have the problem you describe. Sorry..... > > Regards, > Jan Nijtmans Hi, I've been using icu4c, ver 52.1 (most current), which I build from source into a static library, using the following options (copying from my bash build script): CONFIG_OPTS="--enable-static --disable-shared --disable-strict \ --disable-extras --disable-samples --disable-tests" EXTRA_FLAGS="-DU_HAVE_NL_LANGINFO_CODESET=0 \ -DU_USING_ICU_NAMESPACE=0 \ -DU_CHARSET_IS_UTF8=1 \ -DU_NO_DEFAULT_INCLUDE_UTF_HEADERS=1 \ -DUCONFIG_NO_LEGACY_CONVERSION=1 \ -DUCONFIG_NO_TRANSLITERATION=0 \ -D__STDC_INT64__ \ -DU_TIMEZONE=0" GCC_FLAGS="-fno-short-wchar -fno-short-enums" I'll try to see whether the pointer you gave is still valid in 52.1 and perhaps one of my macros is blocking an include with an ifdef. Thanks! Alex -- Alex Loukissas www.maginatics.com _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users