My installing namazu for wwwoffle on debian experience: Indexing my gigabytes of cache took several hours: Size (bytes): 297,458,983 #that's just the index Total Documents: 57,775 Total Keywords: 3,797,251 Wakati: module_kakasi -ieuc -oeuc -w #left it on. Guess it didn't hurt Time (sec): 25,689 #date +%R -d'0 + 25689 sec' gives 07:08, i.e. 7 hours File/Sec: 2.25
All seems OK, and better than htdig, I can even look up Chinese big5 most of the time it seems, good here in Taiwan. I just have to remember to put stars on both sides of search strings: *string*, to find them inside sentences. I didn't change $LANG before doing the indexing. I didn't see any special Chinese features. To install on debian I had to get http://www.naney.org/comp/distrib/mknmz-wwwoffle/archive/mknmz-wwwoffle_0.7.2-1_i386.deb and install debian packages namazu2-index-tools libfile-mmagic-perl namazu2 libnmz3 namazu2-common . I had to change /etc/ppp/ip-down.d/99wwwoffle to run /usr/share/wwwoffle/search/htdig/wwwoffle-htdig-lasttime . I did /usr/share/wwwoffle/search/namazu/wwwoffle-mknmz-full once by hand. I don't use any of the cron scripts. So that's my report from my first day. Sure hope my massive namazu index doesn't get blown away one day like htdig's did. Day 2: Wait, it's my first run of wwwoffle-htdig-lasttime, and minor problem: Use of uninitialized value in concatenation (.) or string at /usr/bin/mknmz-wwwoffle line 367. Use of uninitialized value in concatenation (.) or string at /usr/bin/mknmz line 238. both are due to $HOME being expected. So I'll set HOME in /etc/ppp/ip-down.d/99wwwoffle until naney fixes it. Hey, what's this Deleted Documents: 57,775 why, that all that were there ... all gone ... uh huh. Good thing I kept a backup. It couldn't have been that my documents were ALL too old. It seems that /usr/share/wwwoffle/search/namazu/wwwoffle-mknmz-lasttime is missing a --no-delete ! Ok, trying... OK ... otherwise all those hours of indexing down the drain, and one can't keep an index for more than of just the last online session. By the way, I notice some graphics seemingly left lying around: /var/cache/wwwoffle/search/namazu/db/NMZ.word24.jpg Not noted in /usr/share/doc/namazu2/en/nmz.html
