Page 1 of 3
"Copy Path" regexp transformation?
Posted: 15 Mar 2008 14:54
by ryanagler
So I have been using XY and Directory Opus for over a year now, and I find myself switching back and forth because each one has its specialty. I find XY to be orders of magnitude faster browsing linux network shares, which is what I spend mot of my time working with, so Id prefer using XY for as much as I can.
There is one thing Opus has that I just cant do without, and that is being to copy a file or folder name to the clipboard and have a regexp transform it to a unix-style path for pasting. For example, say i copy a folder that normally would put the text string "S:\usr\local\apache2\conf" into the clipboard. I am looking for something in XY that would automatically apply a regexp to that string to make it "/usr/local/apache2/conf" Does such a feature exist in XY?
Re: "Copy Path" regexp transformation?
Posted: 15 Mar 2008 15:13
by admin
ryanagler wrote:For example, say i copy a folder that normally would put the text string "S:\usr\local\apache2\conf" into the clipboard. I am looking for something in XY that would automatically apply a regexp to that string to make it "/usr/local/apache2/conf" Does such a feature exist in XY?
Currently not, but it's very easy to add, of course. There is only one puzzle stone missing here, and that's a scripting command for string manipulation using regexp. This is actually planned for scripting 2.0 (which will happen in v7.2 or so)!
PS: Is this always like your example? I mean: cut the drive letter and change the slashes? In that case I could add it immediately...
Code: Select all
S:\usr\local\apache2\conf
/usr/local/apache2/conf
Re: "Copy Path" regexp transformation?
Posted: 15 Mar 2008 15:25
by jacky
admin wrote:There is only one puzzle stone missing here, and that's a scripting command for string manipulation using regexp. This is actually planned for scripting 2.0 (which will happen in v7.2 or so)!
/me gets all excited, takes note : v7.2
Re: "Copy Path" regexp transformation?
Posted: 15 Mar 2008 15:34
by ryanagler
admin wrote:Currently not, but it's very easy to add, of course. There is only one puzzle stone missing here, and that's a scripting command for string manipulation using regexp. This is actually planned for scripting 2.0 (which will happen in v7.2 or so)!
awesome!
admin wrote:PS: Is this always like your example? I mean: cut the drive letter and change the slashes? In that case I could add it immediately...
There are some exceptions, but 90% of the time, yes! If you could add it that would be great!
Re: "Copy Path" regexp transformation?
Posted: 15 Mar 2008 20:11
by admin
ryanagler wrote:admin wrote:Currently not, but it's very easy to add, of course. There is only one puzzle stone missing here, and that's a scripting command for string manipulation using regexp. This is actually planned for scripting 2.0 (which will happen in v7.2 or so)!
awesome!
admin wrote:PS: Is this always like your example? I mean: cut the drive letter and change the slashes? In that case I could add it immediately...
There are some exceptions, but 90% of the time, yes! If you could add it that would be great!
90% is not enough. So, what RegExp do you use for this job?
Re: "Copy Path" regexp transformation?
Posted: 16 Mar 2008 09:02
by admin
jacky wrote:admin wrote:There is only one puzzle stone missing here, and that's a scripting command for string manipulation using regexp. This is actually planned for scripting 2.0 (which will happen in v7.2 or so)!
/me gets all excited, takes note : v7.2
Okay, I added it right now.
Not sure about the name of the function.
regexpreplace is a bit long... what about
regrep??
Re: "Copy Path" regexp transformation?
Posted: 16 Mar 2008 09:15
by j_c_hallgren
admin wrote:Not sure about the name of the function.
regexpreplace is a bit long... what about regrep??
I'd suggest: regexrpl ...or... regexrep as I think regex needs to be there.
As I initially read regrep as re-grep...
Re: "Copy Path" regexp transformation?
Posted: 16 Mar 2008 09:26
by admin
j_c_hallgren wrote:admin wrote:Not sure about the name of the function.
regexpreplace is a bit long... what about regrep??
I'd suggest: regexrpl ...or... regexrep as I think regex needs to be there.
As I initially read regrep as re-grep...
Yep. I made a little research and found that "RegExReplace" is a quite common name for it, so I take it.
It will look like this:
Code: Select all
RegExReplace OutputVar, InputString, RegExPattern, Replacement, MatchCase
Re: "Copy Path" regexp transformation?
Posted: 16 Mar 2008 11:38
by admin
admin wrote:PS: Is this always like your example? I mean: cut the drive letter and change the slashes? In that case I could add it immediately...
Code: Select all
S:\usr\local\apache2\conf
/usr/local/apache2/conf
v6.80.0112 can do this now. Create a UDC "Run Script" and use the following multiline script (use Edit... button to enter it):
Code: Select all
// copy current item as unix path
mid $p, <curitem>, 3;
replace $p, $p, "", "/";
copytext $p;
Assign a KS, e.g. Ctrl+U, and it will copy the currently selected item as unix path to the clipboard.
PS: why was I so keen to add it already now? Because now I can do this:
Code: Select all
// copy current item as xyplorer.com URL
replace $p, <curitem>, "", "/";
replace $p, $p, "D:/www/xyplorer.com/code", "http://www.xyplorer.com";
copytext $p;
Re: "Copy Path" regexp transformation?
Posted: 16 Mar 2008 13:42
by jacky
admin wrote:jacky wrote:/me gets all excited, takes note : v7.2
Okay, I added it right now.
Awesome!!
Haven't played with it much, but I couldn' resist a few tries... I have a question though : it seems that caret (^) & dollar ($) only match the very start and very end of the entire string, I think it'd be better if it could match the beginning & end of every line instead (or at least add an option to switch to this mode).
(On your end, I beleive all it takes is to put Multiline to True.)
PS: this is pretty cool stuff!!
Re: "Copy Path" regexp transformation?
Posted: 16 Mar 2008 18:04
by admin
jacky wrote:(On your end, I beleive all it takes is to put Multiline to True.)
Okay, done. All wishes should be that easy!
Re: "Copy Path" regexp transformation?
Posted: 16 Mar 2008 22:56
by jacky
admin wrote:jacky wrote:(On your end, I beleive all it takes is to put Multiline to True.)
Okay, done. All wishes should be that easy!
Cool, thanks
So, couple of (quick) suggestions that seems to come naturally...
- mid : might be nice to have <start> to support negative values, to start from the end, so mid $a, "Maxi", -2 would return "xi"
- locate : returns the position on a string. locate $i, "Maxi", "a" would return 2
- inc : to increment an integer value. set $a, 2; inc $a, 3; msg $a; inc $a, -7; msg $a; would say "5" the first time, and "-2" the second time
- regexp : to use a regexp to search/extract a string, so regexp $a, "C:\Folder\Sub\Here", ":\\(.+?)\", "Main folder is $1" would return "Main folder is Folder"
Posted: 17 Mar 2008 21:45
by PeterH
Some notes from me, too...
On "SubStr", the very last character is addressed by Start = -1, while the very first is Start = 0. I wouldn't really expect this.
(And on a string-instruction in a script I would prefer to count the first char as 1 - I don't think of an offset here, but of a position in the string?)
Isn't the "Incr"-instruction "Add"? It adds second and third operand, and returns result as the first?
(Only, if you don't plan to define "Set result, var1+var2+var3" or so. But I don't expect that...)
Is there a reason to limit "AddStr" to 3 variables? Some more could be quite handy. (While 3 will do in most cases...)
For "functions" requiring a number (like a pos, length, ...): what happens, if it is alfa or so?
But one thing must be said: the abilities of scripting are enhanced by these few variable-manipulating instructions by a whole world!
Enjoy your bootcamp, or how did you call it
Posted: 17 Mar 2008 21:58
by jacky
PeterH wrote:On "SubStr", the very last character is addressed by Start = -1, while the very first is Start = 0. I wouldn't really expect this.
(And on a string-instruction in a script I would prefer to count the first char as 1 - I don't think of an offset here, but of a position in the string?)
As Don said, it goes along with the substr function of PHP. It is common usage, so it makes sense to find it here in XY as well.
PeterH wrote:For "functions" requiring a number (like a pos, length, ...): what happens, if it is alfa or so?
As hinted in a few examples, if the first characters of the string are numbers, their value will be used, otherwise zero.
"foo" => 0
"foo2" => 0
"2foo3" => 2
PeterH wrote:But one thing must be said: the abilities of scripting are enhanced by these few variable-manipulating instructions by a whole world!
Agreed!
BTW, the next logical step here now (but for
after v7.0 of course) has to be condition (IF..THEN..ELSE..) and loops (WHILE), right?
Posted: 18 Mar 2008 02:07
by PeterH
jacky wrote:As Don said, it goes along with the substr function of PHP. It is common usage, so it makes sense to find it here in XY as well.
I'm afraid there are a lot of systems, some count strings from position 0, some from pos 1. Wouldn't it be best to take the better version - whatever? And I wouldn't have posted my question, if there wouldn't be the discrepancy of 1 = 2nd character, but -1 = 1st character from the end. So if 0 is the first char, I would ask for -0 to be the last char...
Thanks for the info about mis-formatted numbers, very usefull to know what to expect.
And yes: If...Then...Else... and some kind of loop (like I said above: different systems have different commands, like While, For, Do, ...). But what I also would like to see are Do...End groups. (Some use Begin...End instead, don't know what PHP does...)