can i use the statement both "$msg contains" and "$fromhost-ip startswith" in rsyslog config?
when i use the follow for rsyslog config, it work!
if $msg contains 'src_port=6699' then -?DynFileA
& ~
if $fromhost-ip startswith '10.10.10.1' then -?DynFileB
& ~
but when i combine both of them, it was fail!
if $fromhost-ip startswith '10.10.10.1' and $msg !contains 'src_port=6699' then -?DynFileA
& ~
i search many article from internet, especially rsyslog office document,but didn't find any useful suggestion!
can any one tell me how to do this?
Hmm... are you sure about your substitution ? At first sight, if your host IP is 10.10.10.1, you'll apply -?DynFileA instead of -?DynFileB. And your first example held two distincts operations... (don't know what it is btw)
regards,
I would just go full rsyslog Reiner script now, your first example was a or, not an and with two different dynamic file templates, if that is what you wanted just replace the and below with an or
if ($fromhost-ip startswith '10.10.10.1') and ($msg !contains 'src_port=6699') then {
-?DynFileA
stop
}
Related
I have the question about the proper handling of the config file. I'm trying to solve my issue for a couple of days now but with the best will, I just can't find out how to do it. I know that this question is maybe quite similar with all the others here and I really tried to use them - but I didn't really get it. I hope that some things about how snakemake works will be more clear when I solved this problem.
I'm just switching to snakemake and I thought I just can easily convert my bash script. To get familiar with snakemake I started trying a simple Data-Processing pipeline. I know I could solve my case while defining every variable within the snakefile. But I want to use an external config file.
First is to say, for better understanding I decided just to post the code which I thought would work somehow. I already played around with different versions for a "rule all" and the "lambda" functions, but nothing worked so far and it just would be confusing. I'm really a bit embarrassed and confused about why I can't get this working. The variable differs from the key because I aways had a version where I redefine the variable, like:
$ sample=config["samples"]
I would be incredibly thankful for an example code.
What I'd like to have is:
The config file:
samples:
- SRX1232390
- SRX2312380
names:
- SomeData
- SomeControl
adapters:
- GATCGTAGC
- GATCAGTCG
And then I thought I can just call the keys like different variables.
rule download_fastq:
output:
"fastq/{name}.fastq.gz"
shell:
"fastq-dump {wildcards.sample} > {output}"
later there will be more rules, so I thought for them I also just need a key:
rule trimming_cutadapt:
input:
"fastq/{name}.fastq"
output:
"ctadpt_{name}.fastq"
shell:
"cutadapt -a {adapt}"
I also tried something with a config file like this:
samples:
Somedata: SRX1232131
SomeControl: SRX12323
But in the end I also didn't find the final solution nor would I know how to add a third "variable" then.
I hope it is somehow understandable what I want to have. It would be very awesome if someone could help me.
EDIT:
Ok - I reworked my code and tried to dig into everything. I fear my understanding lacks in connecting the things I read in this case. I would really appreciate some tips which will probably help me to understand my confusion.
First of all: Rather than try to download data from a pipeline I decided to do this in a config step. I tried out two different versions now:
Based on this answer I tried version one. I like the version with the two files. But I'm stuck in how to deal with the variables now in things like using them with the lambda function or everything where you normally would write "config["sample"]".
So my problem here is that I don't knwo ho to proceed or how the correct syntax is now to call the variables.
#version one
configfile: "config.yaml"
sample_file = config["sample_file"]
import pandas as pd
sample = pd.read_table(sample_file)['samples']
adapt = pd.read_table(sample_file)['adapters']
rule trimming_cutadapt:
input:
data=expand("../data/{sample}.fastq", name = pd.read_table(sample_file)['names']),
lambda wildcards: ???
output:
"trimmed/ctadpt_{sample}.fastq"
shell:
"cutadapt -a {adapt}"
So I went back to try to understand using and defining the wildcards. So (among other things) I looked into the example snakefile and the example rules of Johannes. And of course into the man. Oh and the Thing about the zip function.
At least I don't get an error anymore that it can't deal with wildcards or whatever. Now it's just doing nothing. And I can't find out why because I don't get any information. Additionaly I marked some points which I don't understand.
#version two
configfile: "config_ChIP_Seq_Pipeline.yaml"
rule all:
input:
expand("../data/{sample}.fastq", sample=config["samples"])
#when to write the lambda or the expand in a rule all and when into the actual rule?
rule trimming_cutadapt:
input:
"../data/{sample}.fastq"
params:
adapt=lambda wildcards: config[wildcards.sample]["adapt"] #why do I have to write .samle? when I have to use wildcard.XXX in the shell part?
output:
"trimmed/ctadpt_{sample}.fastq"
shell:
"cutadapt -a {params.adapt}"
As a testfile I used this one.
My configfile in version 1:
sample_file: "sample.tab"
and the tab file:
samples names adapters
test_1 input GACCTA
and the configfile from version two:
samples:
- test_1
adapt:
- GTACGTAG
Thanks for your help and patients!
Cheers
You can look at this post to see how to store and access sample information.
Then you can look at Snakemake documentation here, more specifically at the zip function, which might help you as well.
I wonder if it is possible to remap "$" and "#" to other keys.
sample:
#set( $foo = "bar" )
I want to use other keys because those interfere with another syntax of a script I am using.
$ and # characters are not configurable in Velocity. Even at compile time, it would at least imply to recompile the parser, and make a full code review for standalone $ and # chars...
That said:
Velocity does cope pretty well with syntax fragments it cannot parse, like jQuery $ object. It just render them as is, and most of the time it does the job.
You can escape your other script's sensitive characters whenever needed, for instance by using the EscapeTool: ${esc.d} for dollar, ${esc.h} for hash.
I am trying to get my CGI scripts running on my web host (which runs on FreeBSD). To debug why I keep getting the dreaded "premature end of script headers" error, their support recommended that I redirect all my output to stderr, rather than printing it. Looking up how to do this, I came across a very old RAMBO ticket about it, but it looks like it was never implemented.
Per some of the answers to this question, it seems like I should be able to do a call {echo Hello, world >&2} to achieve this, but it doesn't work.
How can I write to stderr in REBOL2?
For my CGI-specific scenario, I have a truly awful workaround. Since writing to stderr in Perl (with which I am entirely unfamiliar) is a one-liner, I'm currently calling the REBOL script from Perl and printing its output to stderr from there:
#!/usr/bin/perl
use strict;
use warnings;
use CGI;
# Note the backticks
my $the_string = `/home/public/rebol -csw test-reb.cgi`;
print STDERR $the_string;
This webpage has some suggestions http://www.liquidweb.com/kb/apache-error-premature-end-of-script-headers/
to solve your real problem. Perhaps you did not have the headers printed as first thing in your script, this must be the first thing to do. Maybe the rights are not sufficient, or the .r file type was not properly added in your .htaccess as cgi able file. Your (correct!) rebol core exe has not the correct rights. Or your script ends up in an endless loop?
Some hints to redirect errors for Rebol cgi script:
http://www.rebol.com/docs/core23/rebolcore-2.html#section-6.2
Better late than never... I've just implemented it for Rebol3 in my Rebol fork.
https://github.com/Oldes/Rebol-issues/issues/2468
The syntax will be probably changed a little bit, because I don't like that the system console port is named input, although it is not just for the input.
So far it is:
print 1 ;<- std_out
modify system/ports/input 'error on
print 2 ;<- std_err
modify system/ports/input 'error off
print 3 ;<- std_out
Yakuake provides a hotkey and a GUI way to rename commandline tabs/sessions.
I'd like to do the same via the command line, so I can script it and use it in an alias. (My goal is that if I use an alias which does an SSH to some server, then the tab is renamed according to this servers name...)
I tried the suggestions shown here Renaming a Konsole session from commandline after ssh so far no luck.
Since KDE4, one should use qdbus to control KDE apps (instead of deprecated and deleted DCOP).
For example, to change a title of the first session one may use:
qdbus org.kde.yakuake /Sessions/1 org.kde.konsole.Session.setTitle 1 "New title"
To explore available interfaces, methods and properties one may use qdbusviewer.
As a homework try to get a list of active sessions (before you going to change smth).
Like #fgysin pointed out, his command also works for me. BUT it needs the ` character and not " for the subcommand :
qdbus org.kde.yakuake /yakuake/sessions org.kde.yakuake.activeSessionId
It gives :
qdbus org.kde.yakuake /yakuake/tabs org.kde.yakuake.setTabTitle `qdbus org.kde.yakuake /yakuake/sessions org.kde.yakuake.activeSessionId` "NEW TAB TITLE";
EDIT: I have just changed the if to an if else
IF "%CD:~1%"==":\" (
set FIR=%CD:~0,2%
) ELSE (
set FIR=%CD%
)
Trying to use a wildcard to detect if a batch file is on the root directory (As root causes there to be a backslash at the end of the DIR)
I need to determine with a simple if statement whether it is *:\ or not. I have tried using wildcards *:\ and ?:\ but to no avail. If I use the drive letter, example: G:\ it works.
if %CD%==G:\ echo SOMETHING WRONG HERE
works perfectly well, until the drive letter changes. Where as
if %CD%==?:\ echo SOMETHING WRONG HERE
or
if "%CD%"=="?:\" echo SOMETHING WRONG HERE
Doesn't work.
EDIT: After reading up it turns out if and wildcards are not compatible!
try this:
if "%CD:~-2%"==":\" (echo root) else echo NOT root
see docu help set.
Wildcards are not usable within IF statements in command.