Erik, thanks for the correction, you might indeed be a guru! /Lars > -----Original Message----- > From: Erik Christiansen [mailto:dva...@internode.on.net] > Sent: den 24 juni 2011 10:00 > To: emc-users@lists.sourceforge.net > Subject: Re: [Emc-users] Axis naming X Y Z E, can I have that? > > On 23.06.11 22:19, Lars Andersson wrote: > > Nice to have the file conversion integrated in AXIS. > > No need to change the axis name then. > > Thank you for the nicely potted integration example. > I'll certainly keep it for reference. > > [...] > > > This is E2A that was run from outside of AXIS before > > #!/bin/bash > > sed -e 's/\( E\)\([0-9]*\)/ A\2/' $1 > > Guru comments on this? > > Now that's asking for opinion as well as help. ;-) > > Here's 1.5c worth of one or the other: The above regex is in BRE (Basic > Regular Expression) form, and so is partly obscured by a flurry of > backslashes. > Readability is improved (innit?) if we allow sed to use ERE (Extended > Regular > Expressions), like this: > > sed -re 's/( E)([0-9]*)/ A\2/' $1 > > Now it's human readable, yet has identical behaviour: > > echo G1 X53.6223 Y37.9513 E49.6224 F2300 | sed -re 's/( E)([0-9]*)/ > A\2/' > G1 X53.6223 Y37.9513 A49.6224 F2300 > > Now it's also easier to see that the regex may have a weakness, in that > [0-9]* matches "zero or more digits". That means it is no more > selective > than: > > sed -re 's/ E/ A/' > > As we see here: > > $ echo G1 X53.6223 Y37.9513 Eggnog F2300 | sed -re 's/( E)([0-9]*)/ > A\2/' > G1 X53.6223 Y37.9513 Aggnog F2300 > > $ echo G1 X53.6223 Y37.9513 Eggnog 49.6224 F2300 | sed -re 's/ E/ A/' > G1 X53.6223 Y37.9513 Aggnog 49.6224 F2300 > > To make the second regex atom do anything, we need to ask for "one or > more", like so: > > $ echo G1 X53.6223 Y37.9513 E49.6224 F2300 | sed -re 's/( E)([0-9]+)/ > A\2/' > G1 X53.6223 Y37.9513 A49.6224 F2300 > > $ echo G1 X53.6223 Y37.9513 Eggnog F2300 | sed -re 's/( E)([0-9]+)/ > A\2/' > G1 X53.6223 Y37.9513 Eggnog F2300 > > Now it only acts in the desired context. (Hopefully I haven't > belaboured > the explanation too much, in trying to be explicit.) > > Since sed, awk, grep, etc all understand ERE as well as BRE, I've never > found a downside to using the more readable form. Less typing too. > (Even vim can be weaned off its tedious backslash addiction by > prepending \v to the regex expression.) > > There's also this advice in "man 7 regex": > > » > Regular expressions ("RE"s), as defined in POSIX.2, come in two forms: > modern REs (roughly those of egrep; POSIX.2 calls these "extended" REs) > and obsolete REs (roughly those of ed(1); POSIX.2 "basic" REs). > Obsolete REs mostly exist for backward compatibility in some old > programs; they will be discussed at the end. > « > > Whew! I'm glad we get to keep the good ones. > > Erik > > -- > The wonderful thing about standards is that there are so many of them. > - Andy > Tanenbaum > > > ----------------------------------------------------------------------- > ------- > All the data continuously generated in your IT infrastructure contains > a > definitive record of customers, application performance, security > threats, fraudulent activity and more. Splunk takes this data and makes > sense of it. Business sense. IT sense. Common sense.. > http://p.sf.net/sfu/splunk-d2d-c1 > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users
------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity and more. Splunk takes this data and makes sense of it. Business sense. IT sense. Common sense.. http://p.sf.net/sfu/splunk-d2d-c1 _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
