In-place regex substitution

lmocsi has asked for the wisdom of the Perl Monks concerning the following question:

Replies are listed 'Best First'.
Re: In-place regex substitution by toolic (Bishop) on Dec 09, 2016 at 21:39 UTC
If you want to change the string in place, you can use s///ge: use warnings; use strict; my $content = '{ "geometry": { "type": "Polygon", "coordinates": [ [ [ + 19.054804912278406, 47.485785556135411 ], [ 19.057857836771483, 47.4 +87322542030711 ], [ 19.06025597925397, 47.488491565765209 ], [ 19.060 +347248086835, 47.488539642204628 ], [ 19.060463310421543, 47.48457828 +7406251 ], [ 19.054804912278406, 47.485785556135411 ] ] ] } }'; $content =~ s/(\d\d\.\d*)/substr($1,0,9)/ge; print "$content\n"; __END__ { "geometry": { "type": "Polygon", "coordinates": [ [ [ 19.054804, 47. +485785 ], [ 19.057857, 47.487322 ], [ 19.060255, 47.488491 ], [ 19.06 +0347, 47.488539 ], [ 19.060463, 47.484578 ], [ 19.054804, 47.485785 ] + ] ] } } [download]	[reply] [d/l]
Re^2: In-place regex substitution by AnomalousMonk (Archbishop) on Dec 10, 2016 at 00:06 UTC
Or without explicit capturing or evaluation (or rounding!) (update: and with dynamic control of truncation threshold), but needs Perl version 5.10+ for `\K` operator: c:\@Work\Perl\monks>perl -wMstrict -le "use 5.010; ;; my $content = qq/{ \"geometry\": { \"type\": \"Polygon\", \"coordinates\": [ [ +[ 19.05480 4912278406, \n/ . qq/47.485785556135411 ], [ 19.057857836771483, 47.487322542030711 + ], [ 19.06025597925397, \n/ . qq/47.488491565765209 ], [ 19.060347248086835, 47.488539642204628 + ], [ 19.060463310421543, \n/ . qq/47.484578287406251 ], [ 19.054804912278406, 47.485785556135411 + ] ] ] } } \n/ ; print qq{[[$content]]}; ;; my $n = 7; $content =~ s{ \d [.] \d{$n} \K \d+ }{}xmsg; ;; print qq{<<$content>>}; " [[{ "geometry": { "type": "Polygon", "coordinates": [ [ [ 19.054804912 +278406, 47.485785556135411 ], [ 19.057857836771483, 47.487322542030711 ], [ 19 +.06025597925397, 47.488491565765209 ], [ 19.060347248086835, 47.488539642204628 ], [ 19 +.060463310421543, 47.484578287406251 ], [ 19.054804912278406, 47.485785556135411 ] ] ] } + } ]] <<{ "geometry": { "type": "Polygon", "coordinates": [ [ [ 19.0548049, 47.4857855 ], [ 19.0578578, 47.4873225 ], [ 19.0602559, 47.4884915 ], [ 19.0603472, 47.4885396 ], [ 19.0604633, 47.4845782 ], [ 19.0548049, 47.4857855 ] ] ] } } >> [download] Give a man a fish: `<%-{-{-{-<`	[reply] [d/l] [select]
Re^2: In-place regex substitution by lmocsi (Novice) on Dec 09, 2016 at 21:45 UTC
Great, thx! :)	[reply]
Re^3: In-place regex substitution by LanX (Saint) on Dec 09, 2016 at 21:59 UTC
Are you sure that a number can't have more than 9 digits before the point? You might prefer `sprintf` instead of `substr` in order to just shorten the digits after the point. Cheers Rolf _{(addicted to the Perl Programming Language and ☆☆☆☆ :) Je suis Charlie!}	[reply]
Re: In-place regex substitution by Marshall (Canon) on Dec 10, 2016 at 07:54 UTC
To demo one idea from LanX, with the 'e', execute option on the regex, arbitrary code can be executed. I coded the idea of using sprintf(). The code below will round to 6 decimal places instead of truncating at 6 decimal places. Example: `47.488539642204628 truncates to: 47.488539 rounds to: 47.488540` [download] use warnings; use strict; my $content = '{ "geometry": { "type": "Polygon", "coordinates": [ [ [ + 19.054804912278406, 47.485785556135411 ], [ 19.057857836771483, 47.4 +87322542030711 ], [ 19.06025597925397, 47.488491565765209 ], [ 19.060 +347248086835, 47.488539642204628 ], [ 19.060463310421543, 47.48457828 +7406251 ], [ 19.054804912278406, 47.485785556135411 ] ] ] } }'; $content =~ s/(\d\d\.\d*)/sprintf("%.6f",$1);/ge; #round to 6 decimal +digits print "$content\n"; __END__ Truncated: { "geometry": { "type": "Polygon", "coordinates": [ [ [ 19.054804, 47. +485785 ], [ 19.057857, 47.487322 ], [ 19.060255, 47.488491 ], [ 19.06 +0347, 47.488539 ], [ 19.060463, 47.484578 ], [ 19.054804, 47.485785 ] + ] ] } } Rounded: { "geometry": { "type": "Polygon", "coordinates": [ [ [ 19.054805, 47. +485786 ], [ 19.057858, 47.487323 ], [ 19.060256, 47.488492 ], [ 19.06 +0347, 47.488540 ], [ 19.060463, 47.484578 ], [ 19.054805, 47.485786 ] + ] ] } } [download]	[reply] [d/l] [select]
Re^2: In-place regex substitution by AnomalousMonk (Archbishop) on Dec 10, 2016 at 15:49 UTC
`$content =~ s/(\d\d\.\d)/sprintf("%.6f",$1);/ge; #round to 6 decimal +digits` [download] One can avoid hard-coding the rounding precision by using either `my $n = $some_integer;` `... sprintf("%.${n}f", $1) ...` or the "wildcard" (if that's the right term (update:* maybe "placeholder"?)) format specifier `... sprintf('%.f', $n, $1) ...` Update:* In the code above, I've implied that `$n` must be an integer. Interestingly (for some definition of "interesting"), if it is not, then the first code example (`$n` interpolated) screws up, but the second (`*` wildcard specifier) behaves "properly", DWIMishly ignoring the fractional part of the numeric value. (I haven't checked yet to see if this behavior is the same as in C/C++/D/etc.) Give a man a fish: `<%-{-{-{-<`	[reply] [d/l] [select]
Re^2: In-place regex substitution by LanX (Saint) on Dec 10, 2016 at 15:36 UTC
Thanks, but a minor nitpick. :) > `s/(\d\d\.\d)/`* this will exclude numbers smaller than ten (already so in marto's post) Supposing all numbers have a decimal point, I'd say `s/(\d+\.\d*)/` It's really hard to guess which formats are possible and where they'll appear. In the end the safest way is use a JSON parser. Cheers Rolf _{(addicted to the Perl Programming Language and ☆☆☆☆ :) Je suis Charlie!}	[reply] [d/l] [select]