<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<p>Hi Tim,</p>
<p>many, many kudo's. You were right, after installing <a
title="Package Information Summary"
href="https://pkg.openindiana.org/hipster/info/0/system%2Flibrary%2Ficonv%2Funicode%400.5.11%2C5.11-2022.0.0.20866%3A20220109T015823Z">system/library/iconv/unicode</a>
all tests passed</p>
<p><font face="monospace">+Test project $(@D)<br>
+ Start 1: libclamav<br>
+1/5 Test #1: libclamav ........................ Passed
19.76 sec<br>
+ Start 2: clamscan<br>
+2/5 Test #2: clamscan ......................... Passed
4.83 sec<br>
+ Start 3: clamd<br>
+3/5 Test #3: clamd ............................ Passed
21.79 sec<br>
+ Start 4: freshclam<br>
+4/5 Test #4: freshclam ........................ Passed
2.07 sec<br>
+ Start 5: sigtool<br>
+5/5 Test #5: sigtool .......................... Passed
0.63 sec<br>
+<br>
+100% tests passed, 0 tests failed out of 5 <br>
</font></p>
<p>and even <br>
</p>
<p>builduser@userland:~$ iconv -f PCK -t UTF-8 "/tmp/ja"<br>
縺薙ｓ縺ｫ縺｡縺ｯ縺薙ｓ縺ｫ縺｡縺ｯ</p>
<p>now delivers a reasonable result.</p>
<p>Another question. Jim Klimov initially created this package. He
also prepared a README stating:</p>
<p><i>It is expected that ultimate deployments might not want to run
all of</i><i><br>
</i><i>the components in the same operating environment. In
particular, the</i><i><br>
</i><i>maintenance of a local copy of the virus definitions
database adds a</i><i><br>
</i><i>considerable storage and internet-traffic overhead which
would be not</i><i><br>
</i><i>needed on systems that do not run `clamd` or other
implementations of</i><i><br>
</i><i>the scanning engine directly.</i><i><br>
</i></p>
<p>So, he created several packages to allow distinct installations.
Currently I've created only one package, but for update reasons
should I stick to the original packages (I think storage shouldn't
be a problem nowadays, despite I'm an educated engineer trying to
be as efficient to environment as possible, basically it is a
consideration of minimum space requirements and kind of
convenience, also without updating databases a virus scanner on a
regular basis is kind of useless, right)?</p>
<p>kind regards,</p>
<p> Fritz<br>
</p>
<p><br>
</p>
<div class="moz-cite-prefix">Am 11.01.2022 um 21:15 schrieb Tim
Mooney via oi-dev:<br>
</div>
<blockquote type="cite"
cite="mid:%3C4b35fbd-a56a-e5ef-ed0-5a8de7aecc8@ndsu.edu%3E">In
regard to: Re: [oi-dev] clamav update, Friedrich Kink via oi-dev
said...:
<br>
<br>
<blockquote type="cite">But maybe this code page is simply not
supported by openindiana because I tried to play around with
iconv, too (it seems there is nothing similar to
JAPANESE_SHIFT_JIS) :
<br>
</blockquote>
<br>
I just ran into this a couple weeks ago with a different test
suite.
<br>
Based on how I read iconv_unicode(5) (and iconv(5), for non
unicode),
<br>
OI's iconv has a limited set of supported conversions.
<br>
<br>
The tables in iconv_unicode(5) list "SJIS", though, which is the
alias
<br>
OI uses, so in theory it should be possible to go from UTF-8 to
SJIS.
<br>
<br>
That it's not showing up in the output from iconv -l has me
wondering if
<br>
it's the case of a missing package. My guess is there should be a
module
<br>
somewhere in /usr/lib/iconv/ that has "SJIS" (or perhaps some
variant of
<br>
Kanji, since that's listed as the name, and SJIS is apparently the
alias?)
<br>
somewhere in its name.
<br>
<br>
Note the 'eucJP' does show up in the iconv output, and that is an
<br>
alternate (incompatible) pre-Unicode encoding.
<br>
<br>
This page has a lot of good info on the various encodings for
Japan:
<br>
<br>
<a class="moz-txt-link-freetext" href="https://www.sljfaq.org/afaq/encodings.html">https://www.sljfaq.org/afaq/encodings.html</a>
<br>
<br>
Tim
<br>
</blockquote>
</body>
</html>