Toolbox

In the process of creating a wikified version of the LKSF, i had the problem of importing all those files into the Wiki. As wikis.sun.com is based on the Atlassian Confluence i couldn´t use the already available scripts for Mediawiki for example . I wasn´t really fond of the idea to do this manually, thus i´ve asked via Twitter for a solution and i´ve got a hint for the xmlrpc interface. At the end, i´ve created a script with stolen ... sorry ... copied components of many other scripts (like this one or this one). My script compares the current version with the new version, uploads the new version and creates new pages if nescessary.It takes three command line parameters: the name of the Wikipage, the name of the space (something like a subwiki or a administrative domain in Confluence) and the file you want to upload. Sorry for the programming style ... python isn´t really my native programming language, but the xmlrpc is solved really good in python and it was a good reason to refresh my really rusty python skills.

Okay ... let´s get to the usage. For example:

./export.py sat.lksf_sat_crashcore.tex lksf sat/lksf_sat_crashcore.tex.wikified

This uploads the file sat/lksf_sat_crashcore.tex.wikified into the space lksf with the page name sat.lksf_sat_crashcore.tex>. This leads to the wiki page http://wikis.sun.com/display/lksf/sat.lksf_sat_crashcore.texNow to the source of the script export.py:

#!/usr/bin/python

import sys
import xmlrpclib
import hashlib

title = sys.argv[1]
space = sys.argv[2]
file  = sys.argv[3]
content = ''

print "================="
print file

chash = hashlib.sha256()

try:
 datei = open(file, 'r')
 for line in datei:
    content = content + line
    if line == "BREAKBREAKBREAK\n":
    	break
 datei.close()
except IOError:
 print  file , "not readable"
 sys.exit(1)

chash.update(content)
contenthash = chash.hexdigest();
print "new content hash:", contenthash

s = xmlrpclib.Server("http://wikis.sun.com/rpc/xmlrpc")
token = s.confluence1.login("username", "password")

newpage= "false"

try:
 	page2 = s.confluence1.getPage(token, space, title)
 	oldcontent = page2["content"]
	ochash = hashlib.sha256()
	ochash.update(oldcontent)
	oldcontenthash = ochash.hexdigest();
	print "old content hash:", oldcontenthash
	if oldcontenthash == contenthash:
		print "no update needed"
		sys.exit(0)
except xmlrpclib.Fault, err:
    print "A fault occurred"
    print "Fault code: %d" % err.faultCode
    newpage = "true"

if newpage == "true":
	page={}
	page["title"] = title
	page["content"] = content
	page["space"] = space
else:
    page=page2
    page["content"] = content

try: 
	s.confluence1.storePage(token, page)
except xmlrpclib.Fault, err:
    print "A fault occurred"
    print "Fault code: %d" % err.faultCode
    print "Fault string: %s" % err.faultString

print title, "updated"

Just in case you wonder because of this "BREAKBREAKBREAK" line. It´s just for debugging purposes. As i wrote yesterday, the xmlrpc is a little bit bitchy about charsets and the error message wasn´t helpful. It was easier to find the problem by bisecting the file again and again. And this was done by the "BREAKBREAKBREAK" line

Ben Rockwood wrote an excellent article about measuring I/O bandwidth: IO Benchmarking: How, Why and With What.

Kris hat einen ein sehr guten Artikel über die Abläufe beim Starten und Beenden von Prozessen in einem unixoiden Operating Environment geschrieben: fork, exec, wait und exit. Eindeutige Leseempfehlung.

Ich weiss, es gibt Safari Books schon länger. Ich habe es trotzdem erst jetzt für mich entdeckt. Schon praktisch, wenn man sich Fachbücher nur ausleiht so lange man sie defacto braucht. Ich habe etwa ein Billy-Regal voll mit Fachbüchern, die ich nur seltenst anpacke und sowieso nie da habe, wo ich sie brauche.

Man nehme zwei Linksys WRT54GL, die DD-WRT Firmware mit VPN-Unterstützung sowie ein paar kleinere Startup-Scripte und schon kann man SunRay-Sessions auch ueber WAN benutzen. Ich werde darueber in den nächsten Tagen eine Anleitung im Wiki schreiben.

Okay, you´ve downloaded the Sun One Directory server, want to use an opensource application with schema extensions for OpenLdap... but ... the schema files of openldap are non-standard, so you are not able to add them to the Sun One Directory Server. You have to rewrite them. Or convert them with a little script.

Der Admin des LDAP-Servers ist beim Mittag oder hat gerade keine Lust die Audit-Logs herauszugeben. Oder es stehen Requests von hunderten von Servern auf dem Server ein und die Analyse geraet zum Nadel im Nadelhaufen suchen.

/opt/csw/bin/tethereal -V port 389 | /usr/xpg4/bin/egrep "(Filter|Base)"