Hi, Have you looked into this idzebra bug at all?
Sorry that I can't easily contribute to bugs.debian.org. Some observations for you: 1. There are three places in the source code where a char is passed to isalpha or isspace. You should cast to unsigned char here, I think, to avoid getting different behaviour on different architectures. Plain char is unsigned on arm64, and also on PPC, I think. 2. This warning is one that shouldn't, I think, be ignored: d1_sumout.c: In function 'data1_nodetosummary': d1_sumout.c:104:25: warning: assignment from incompatible pointer type case 14: res->rank = f_integer(c, o); break; ^ 3. The test passes on arm64 if you insert a space before the '-' in the line containing "<abstract>" in test/api/test_icu_indexing.c. I'm not suggesting this as a fix, but it might help understand what's happening. Sorry if any of this is obvious or irrelevant, but I thought it might help. Edmund -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org