07-18-2016
I've learned the hard way that compressing my code to the extreme just confuses myself when I need to change it. Do I really want 2 hours of extricating myself for a few extra lines of clear code that cost no noticeable time.
There is a case for a one-liner if you can get a tool such as awk to do several things at once rather than calling a loop with multiple greps or cuts being called within it. For me it comes down to using it for processing efficiency. Even if I can make my code fit into fewer blocks on disk, it's just not worth it. If I do, then it gets an explicit comment, often with example code it is logically replacing. If I can adjust it in future in a longer method, then conversion would be simpler if I get my requirements clearer first.
It's the same with debugging logs for batch programs - write lots of info to the log so that when there is a problem, the trace is there. Just make sure you clean up old logs after a short time to avoid running out of space. It's far better than trying to trace or reproduce an error, potentially altering your data for the worse on each attempt or having to put trace information in when an error occurs and you are bleary-eyed.
Robin
4 More Discussions You Might Find Interesting
1. UNIX for Dummies Questions & Answers
First of all, apologies to the admins for not reading the rules totally and missing the bit about ranting off about other OSs.
But that raises a question. Where do you go to have a good rant, to vent your disgust at various corporations and thier hideous behaviour?
:confused: (2 Replies)
Discussion started by: u6ik
2 Replies
2. Shell Programming and Scripting
Guys,
I have a requirement like this.
A file has >5K records always. Separated by "|", it has 30 fields for each line. In some lines, I am getting an odd field. say, the 15th field is supposed to be 2 characters but comes in as >2. In this case, for resolving this I need to copy the value of... (6 Replies)
Discussion started by: PikK45
6 Replies
3. What is on Your Mind?
I wanted to say LOL and punch my face when I saw post#11 (where Don_Cragun even reduced the string manipulation with a simple regex) in the thread https://www.unix.com/shell-programming-scripting/220553-add-0-start-filename-2.html
I mean, when things can be done with just a one liner, sometimes I... (6 Replies)
Discussion started by: ahamed101
6 Replies
4. What is on Your Mind?
Hi guys...
(Apologies for any typos etc...)
This is basically a rant.
I have been doing kids level projects and writing code to suit since around 1982, for the uProfessor, for the Sinclair Spectrum and later for the QL, IBM-XT in MS-DOS and after that for a 386DX40 up to Windows 95, until I... (3 Replies)
Discussion started by: wisecracker
3 Replies
LEARN ABOUT DEBIAN
rant-import
rant-import(1) rant-import rant-import(1)
NAME
Rant - Flexible, Ruby based make
SYNOPSIS
The rant-import command creates a monolithic rant script tailored to the needs of your project and thus removes the dependency on an Rant
installation (but of course one person needs an Rant installation to run rant-import).
Run the command with the --help option to get a brief help message:
% rant-import --help
Probably the easiest way to create your monolithic rant script is with the --auto option:
% rant-import --auto ant
This will write a monolithic rant script to the file ant in the current directory. To determine which plugins and imports your project is
using, it performs step 2 of the rant command as described in doc/rant.rdoc, which means that it loads the Rantfile in the current direc-
tory.
That one command should be enough:
% ruby ant
This script has the same behaviour as the rant command. Distribute it with your project and nobody else but you needs an Rant installation.
FURTHER INFORMATION
For more information on Rant see: <http://make.rubyforge.org/>
<http://make.rubyforge.org/files/doc/rant-import_rdoc.html>
AUTHOR
Rant is by Stefan Lang
This package is maintained by Ralph Amissah
SEE
rant(1)
version 0.5.6 July 26, 2006 rant-import(1)