Hi,
I think this problem should be fixed on w3m side because adding a character
conversion facility to migemo causes a conversion twice.
(w3m converts a string to the system's encoding(UTF-8) and migemo convert back
to EUC-JP.)
An attached patch is for w3m.
This patch changes the "to encoding" to EUC-JP from system's encoding.
It works on EUC-JP and UTF-8 locales.
Thanks,
--
Morita Sho <[EMAIL PROTECTED]>
diff -u w3m-0.5.2.orig/search.c w3m-0.5.2/search.c
--- w3m-0.5.2.orig/search.c 2004-03-24 01:44:02.000000000 +0900
+++ w3m-0.5.2/search.c 2008-06-08 21:40:59.000000000 +0900
@@ -58,7 +58,7 @@
if (migemor == NULL || migemow == NULL)
if (open_migemo(migemo_command) == 0)
return str;
- fprintf(migemow, "%s\n", conv_to_system(str));
+ fprintf(migemow, "%s\n", wc_conv_strict(str, InnerCharset, WC_CES_EUC_JP)->ptr);
again:
if (fflush(migemow) != 0) {
switch (errno) {
@@ -68,7 +68,7 @@
goto err;
}
}
- tmp = Str_conv_from_system(Strfgets(migemor));
+ tmp = wc_Str_conv(Strfgets(migemor), WC_CES_EUC_JP, InnerCharset);
Strchop(tmp);
if (tmp->length == 0)
goto err;
