Yogsototh

Blog switch

yogsototh — Fri, 11 Sep 2009 21:34:12 +0000

I completely switched my blog. Now all new entries should be at http://yannesposito.com

I almost all done from Scratch using a web page generator engine called nanoc (http://nanoc.stoneship.org). I use the old 2.2 version.

I’ve got some work to enable all feature wordpress had given to me. I’ve almost finished doing so.

GitHub behind the evil corporate firewall

yogsototh — Wed, 10 Jun 2009 09:50:18 +0000

Thanks to github providing git using port 443, because it is mostly open.

Simple and efficient, create an ssh alias

> cat ~/.ssh/config
Host github
User git
Port 443
Hostname ssh.github.com

edit the file .git/config into your project :

...
[remote "github"]
url = git@github:user/Project.git
fetch = +refs/heads/*:refs/remotes/github/*

Multi-line grep

yogsototh — Mon, 04 May 2009 14:17:26 +0000

After many, many times I use many different tricks to implement a multi-line grep, I’ve done one myself I called it mlegrep:

#!/usr/bin/env zsh

separator='\n\n'
while [[ ${1[1]} = '-' ]]; do
    case ${1[2]} in
        F) separator="${1[3,-1]}" ; shift ;;
        *) print "unknown option : $1" >&2; exit 2 ;;
    esac
done

if (($#>1)); then     
    print   "usage: $0:t [-Fseparator] REGEXP [FILE ...]" >&2
    exit 1
fi

value="$1"
shift
perl -ne 'BEGIN{$/="'$separator'";} if ( $_ =~ m/'$value'/ ) { print }' "$@"

hope this help,
Y.

YPassword v1.4

yogsototh — Thu, 02 Apr 2009 08:27:44 +0000

I updated the last version of YPassword to 1.4, mostly bug correction.

More informations or direct download.

Blogged with the Flock Browser

get disk usage by file type

yogsototh — Wed, 18 Mar 2009 14:51:40 +0000

A simple and usefull script to evaluate usage of your disk by filetype. Should be improved
to manage extension with more than 3 letters. For this you can just replace the /\.{2,3}$/ by

/\..{2-6}$/ to accepts extension from 2 to 6 letters long.

#!/usr/bin/env zsh
if (($#<1)); then
    cible='.'
else
    cible=( "$@" )
fi
find $cible -type f -exec du -k {} \;  |  awk --posix '
{
    if ($2 ~ /\.[^/]{2,3}$/) {
        ext=$2;
        sub(/.*\./,"",ext);
    } else {
        ext="withoutExtension";
    }
    taille[ext]+=$1 ;
    nb[ext]++ ;
}
END{
    for (x in taille) {
        printf "%10d ", nb[x];
        print x" for "taille[x]" Ko";
    }
}' | sort -n

hope this is usefull

Blogged with the Flock Browser

Better passwords for YPassword

yogsototh — Tue, 17 Mar 2009 07:43:21 +0000

I recently updated YPassword to 1.3. Just added the base64 output in order to make my password shorter. See my blog entry on yannesposito.com.

From fluid to fixed layout and why

yogsototh — Sat, 14 Mar 2009 17:13:54 +0000

I recently changed my mind about fixed layout in web design. Click to see why.

How to manage accessibility with iWeb

yogsototh — Mon, 02 Mar 2009 11:22:58 +0000

I discovered this week-end iWeb use JavaScript to manage the navigation bar. The resulting problem is: no blog post is referenced in Google.

Now I did manage the problem with a simple trick.

First I opened the Web directory in the interface (I have a bug, and have to remove the Web link). It then reappear and create a second mount: yannesposito-1 which contain the Web directory on my iDisk.
My site’s name is YBlog

Once my iDisk Web folder was mounted then I opened a terminal and did :

cd /Volumes/yannesposito-1/Web/Sites/

{
print ''
for fic in YBlog/**/*.html(.); do
    print ''">'$fic:t'
'
done
print ''
} > /tmp/hiddenReferences.txt

I then “copy/paste” the content of /tmp/hiddenReferences.txt into an HTML snippet in my “welcome” page.

But tonight, I’ll do a much better thing (to automatize a little more), and to prevent the problem that, in fact the iFrame is not searched by Google.

I’ll make a simple text box, relatively small and hidden, containing only one link to:

http://yannesposito.com/hiddenReferences.html

And I’ll create a script called updateiWebReferences.sh with the following content

#!/usr/bin/env zsh
WebDirectory="/Volumes/yannesposito-1/Web/Sites"
cd $WebDirectory

{
: ###### Print XHTML 1.0 Strict prefix #####
cat << END_TAG
'
 
    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="fr" xml:lang="fr">
    <head>
        <meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
        all site references
    
    
       All pages of my site
       
END_TAG

: ###### Add all pages with their title #####
for fic in YBlog/**/*.html(.); do
    pageTitle=$( perl -pe 'if(m#(.*)# ) {
            if ( $1 ) { $_=$1."\n";}
            else { $_="" ; } }
        else { $_=""; }' <  $fic )
    if [[ $pageTitle = "" ]]; then
        pageTitle=span style="font-weight:bold;color:#aa55aa;">"$fic:t:gs#_# #"
    fi
    print ''">'$pageTitle''
done

: ###### End of the XHTML page #####
cat << END_TAG
        
    

END_TAG
} > $WebDirectory/hiddenReferences.html

And I’ll just have to launch this script once per new blog entry, or each time I change a blog post title, just after I published my iWeb website. And now all my blog entries are referenced in google.

Tell me if you founded this trick usefull.

Blogged with the Flock Browser

join in zsh

yogsototh — Tue, 23 Dec 2008 14:01:49 +0000

Unlike the join utility which make a join similar to SQL ones. Here is a function to make the same work as the join function in javascript, Perl or Python:

> join ';' 'hello world!' "I'm here!" "I'm happy"

hello world!;I'm here!;I'm happy

Here is a short solution:

function join {
    local res=""
    if (($#>0)); then
        local sep="$1"
        shift
        if (($#>0)); then
            res=$1
            shift
            while (($#>0)); do
                res=$res$sep$1
                shift
            done
        fi
    fi
    print $res
}

Blogged with the Flock Browser

My two cents on Cappuccino

yogsototh — Wed, 17 Dec 2008 16:12:15 +0000

Here is one of my long post: My two cents on Cappuccino.

This was my first impression of Cappuccino after trying to do the tutorials.

In substance: I like the concept and I dislike the implementation. But there is very few to add to convince me.

Blogged with the Flock Browser