+- Southperry.net (https://www.southperry.net)
+-- Forum: Social (https://www.southperry.net/forumdisplay.php?fid=14)
+--- Forum: Rubik's Cube (https://www.southperry.net/forumdisplay.php?fid=58)
+--- Thread: Quick help with regular expressions (/showthread.php?tid=13206)
Quick help with regular expressions - Fiel - 2009-07-06
Okay, here is my string I'm working with.
"Todd taught me to move towards upper right-hand side to attack and kill #bJr. Sentinel#k"
I want it to turn into this:
"Todd taught me to move towards upper right-hand side to attack and kill [noparse]Jr. Sentinel[/noparse]"
Here's how I'm currently doing it:
Regular expression --> #b[\w ]*#k
Then snip off the ends (#b and #k)
Then add my tags ([noparse] and [/noparse])
Then do a string replace("[noparse]Jr. Sentinel[/noparse]" for "#bJr. Sentinel#k")
There's got to be a faster, more efficient way of doing this in Python. Any ideas?
Quick help with regular expressions - Dusk - 2009-07-06
replace [noparse]/#b([\w ])#k/ with /\Q\E\1\Q/.[/noparse]
Does Python have a substitution function?
sed/anything similar =
Actually I guess that should be [noparse]s:#b(.*?)#k:\1:g[/noparse] which looks much nicer.
Explanation: \Q starts a literal string, \E ends it. When you place a segment of a regex in parentheses it gets stored in the buffers \1..\9.
Quick help with regular expressions - Stereo - 2009-07-06
import re
re.sub(#b([\w ]+)#k,$1,string)
etc. I'm not sure if you need to escape the [] in the replacement string cause I tested it using some online thingy. And I don't know if \w will match a "." which is in Jr. Sentinel. Might just want [.]
Brackets() let you refer to each match by $1, consecutive brackets are $2 etc. like if you look for (#[br])([\w ]*)#[kb] then $1 would be a b, $2 the "Jr. Sentinel".
Quick help with regular expressions - Dusk - 2009-07-06
Stereo Wrote:Might just want [.]
Oh, it doesn't. Shouldn't that be [\.\w ] though? It'd mess up if there's more than one tag on a line, i.e.
#bTodd#k taught me to move towards upper right-hand side to attack and kill #bJr. Sentinel#k
edit: k that does it below lol. Do global find/replace though.
Quick help with regular expressions - Nikkey - 2009-07-06
[noparse]re.sub("#b(.*?)#k", "\\1", x)[/noparse]
brb dinner.
Quick help with regular expressions - Fiel - 2009-07-06
Ah, didn't know about the buffers in the re.sub() function. No wonder why I couldn't make heads from tails on it. Devil's sub() function worked perfectly.
One question:
"(.*?)" <--- Isn't the *? redundant? Couldn't it just as well be "(.*)"? What am I missing here?
Also, Python's regex parser leans heavily on Perl. I wouldn't be surprised if Dusk's expression worked too.
Quick help with regular expressions - Nikkey - 2009-07-06
Fiel Wrote:Ah, didn't know about the buffers in the re.sub() function. No wonder why I couldn't make heads from tails on it. Devil's sub() function worked perfectly.
One question:
"(.*?)" <--- Isn't the *? redundant? Couldn't it just as well be "(.*)"? What am I missing here?
Also, Python's regex parser leans heavily on Perl. I wouldn't be surprised if Dusk's expression worked too.
(.*?) is lazy matching. Trying to find a match through (.*) would yield a greedy match (as big match as possible). Greedy matches would work okay in sentences where there are only one match, e.g.
"Todd taught me to move towards upper right-hand side to attack and kill #bJr. Sentinel#k"
However, whenever there are several matches, greedy matches would fail:
"Todd taught me to move towards upper right-hand side to attack and kill #bJr. Sentinel#k and #bDrum Bunnies#k"
Would be transformed into
[noparse]"Todd taught me to move towards upper right-hand side to attack and kill Jr. Sentinel#k and #bDrum Bunnies"[/noparse]
With the "?", you add in lazy matching, the match would translate into
[noparse]"Todd taught me to move towards upper right-hand side to attack and kill Jr. Sentinel and Drum Bunnies"[/noparse]
And as stated earlier, () are brackets, and may be referred to through litteral "\1". (Litteral --> backslashes equal backslashes. Thus, to refer to a backslash, you have to double-backslash. And to actually refer to a "\"-replacement, you'd have to write four backslashes.). \\1 thus refers to the match.
Though, why not actually just
[noparse]x.replaceAll("#b", "").replaceAll("#k", "")[/noparse]
?
Quick help with regular expressions - Fiel - 2009-07-06
Devil's Sunrise Wrote:[noparse]x.replaceAll("#b", "").replaceAll("#k", "")[/noparse]
?
Because that would break other expressions.
"#rTodd#k taught me to move towards upper right-hand side to attack and kill #bJr. Sentinel#k"
where #r#k = [noparse][highlight][/highlight][/noparse]
Quick help with regular expressions - Stereo - 2009-07-07
Isn't #b for blue text anyway? I think bold is #e.
Quick help with regular expressions - KajitiSouls - 2009-07-07
Stereo Wrote:Isn't #b for blue text anyway? I think bold is #e.
I'm with Stereo, because my friend sees #e all over the place in the ToS stuff, which has no blue text in sight, but bolds in various places.
I should add that the Guest ToS (if there are other ToS versions, I dunno) is easily available since all you have to do to see it is to click the Guest Login at the MS Login screen =)