I am working on a Automator service and in my situation I have stdin as
B-Funny Flash Nonfiction 202105131635 and I want to get to B-Funny Flash Nonfiction 202105131636 incriminating the "5" by 1 to "6".
I'd think I'd first want to separate the text from the number before doing the add 1 then rejoin them?
Would egrep or sed or awk be best?
Tips?
Bash has simple integer arithmetic built in.
str='B-Funny Flash Nonfiction 202105131635'
# parse into prefix and number
num=${str##*[!0-9]}
prefix=${str%$num}
echo "$prefix$((num+1))"
The parameter expansion ${var#pat} produces the value of the variable var with any prefix matching pat removed; % does the same for suffixes, and doubling the operator changes to matching the longest possible pattern match instead of the shortest. The pattern *[!0-9] matches a string which ends on a character which isn't a number; in this context, it retrieves the prefix, i.e. everything up to just before the first digit. (If your prefix could contain numbers, too, this needs tweaking. Probably switch to removing all digits from the end, then extracting the removed numbers; but I guess this will require an unattractive temporary variable.)
Finally, the secret sauce which evaluates an arithmetic expression is the $((...)) arithmetic context.
For more involved number crunching, try bc or Awk. In fact, this could be a one-liner in Awk:
awk '{ $NF +=1 }1' <<<"$str"
The here string passes the value as standard input to Awk, which increments the last field $NF. The final 1 is a common Awk shorthand for "print all input lines to output".
I don't know the bash tools well enough to give a cool one-line answer, so here is a python script instead.
Usage
Save the code in a file increment.py;
Make the file executable with chmod +x increment.py;
Run the script with ./increment.py blablabla 123.
Code
#!/usr/bin/env python3
import sys
def print_help(argv0):
print('increment numbers by 1')
print('example usage:')
print(' {} B-Funny Flash Nonfiction 202105131635'.format(argv0))
print(' B-Funny Flash Nonfiction 202105131636')
def main(argv):
if len(argv) < 2:
print_help(argv[0])
else:
for s in argv[1:]:
if s.isnumeric():
print(int(s) + 1, end=' ')
else:
print(s, end=' ')
print()
if __name__=='__main__':
main(sys.argv)
Explanation
In a python program called from the command-line, the command-line arguments are stored in the array sys.argv.
The first element of the array, with index 0, is the name that was used to call the program, most likely "./increment.py" in our case.
The remaining elements are the parameters that were passed to the program; the words "B-Funny", "Flash", "Nonfiction", "202105131635" in our case.
The for-loop for s in argv[1:]: iterates on the elements of argv, but starting with the element 1 (thus ignoring the element 0). Each of these elements is a string; the method .isnumeric is used to check whether this string represents a number or not. Refer to the documentation on .isnumeric.
If the string is not numeric, we print is as-is. If the string is numeric, we compute the number it represents by calling int(s), then we add 1, and we print the result.
Apart from that, the line if len(argv): checks whether argv contains at least two elements; if it doesn't, that means it only contains its element 0, which is "./increment.py"; in this case, instead of printing the arguments, the script calls the function print_help which explains how to use the program.
Finally, the bit about if __name__ == '__main__': is a python idiom to check whether the file increment.py was run as a program or as a module imported by another file. Refer to this question.
I know that one can get the length of an array in bash by doing ${#arrayname[#]}.
My question is: is this just something that I have to memorize, or can this syntax be broken down into understandable parts? For instance, what does the # symbol mean where one would expect to find the index? Why the #?
# at the beginning of a variable reference means to get the length of the variable's value. For a normal variable this means its length in characters. # is the "number" sign, so you can remember this as meaning "the number of things in the variable".
# or * in an array index means to use the whole array, not a specific element, and instead of returning the number of characters, it returns the number of array elements. * is used as a wildcard in many contexts, so this should be easy to remember. Also, $* and $# are used to mean all the arguments to a shell script, so the parallel with all the array elements should be obvious.
You can't just write ${#arrayname} because when you use an array variable without a subscript, it's equivalent to element 0 of the array. So ${#arrayname} is the same as ${#arrayname[0]}, which is the number of characters in the first element of the array.
You should memorize. :) The # usually means number. e.g. the
$# - is the number of arguments
${#str} - length of the string $str
${#arr[#]}" - length (number of elements) of the array arr
${#arr} - the length of the 1st element of the array (like the str above)
Unfortunately the ${parameter#word} or ${parameter##word} has nothing with numbers. (it removes the shortest/longest word from the beginning of the parameter.
And also, the # .... is comment ;)
In general usage of form ${#PARAMETER} returns the length in number of characters and NOT bytes of the parameter's value.
myString="Hello StackOverflow!"
printf "%s\n" "${#myString}"
20
But for arrays, this expansion type has two meanings:
For individual elements, it reports the string length of the element
(as for every "normal" parameter)
For the mass subscripts # and * it
reports the number of set elements in the array
Consider an example over arrays,
myArray=(1 2 3 4 15)
printf "%s\n" "${myArray[#]}" # <-- Gives me list of elements
1
2
3
4
15
printf "%s\n" "${#myArray[#]}" # <-- Gives me number of elements
5
It gets interesting now, the length of the last element 2 can be obtained by doing
printf "%s\n" "${#myArray[4]}"
2
The '#' acts the same way as '*'. Instead of providing a specific index this references the full thing.
The '#' is telling bash you want the length
https://www.cyberciti.biz/faq/finding-bash-shell-array-length-elements/
I want to filter my sequences which has more than 8 same consecutive nucleotides like "GGGGGGGG", "CCCCCCCC", etc in my fastq files.
How should I do that?
The quick and incorrect way, which might be close enough: grep -E -B1 -A2 'A{8}|C{8}|G{8}|T{8}' yourfile.fastq.
This will miss blocks where the 8-mer is split across two lines (e.g. the first line ends with AAAA and the second starts with AAAA). It also assumes the output has blocks of 4 lines each.
The proper way: write a little program (in Python, or a language of your choice) which buffers one FASTQ block (e.g. 4 lines) and checks that the concatenation of the previous (buffered) block's sequence and the current block's sequence do not have an 8-mer as above. If that's the case, then output the buffered block.
I ended up to use below codes in R and solved my problem.
library(ShortRead)
fq <- FastqFile("/Users/path/to/file")
reads_fq <- readFastq(fq)
trimmed_fq <- reads_fq[grep("GGGGGGGG|TTTTTTTTT|AAAAAAAAA|CCCCCCCCC",
sread(reads_fq), invert = TRUE)]
writeFastq(trimmed_fq, "new_name_for_fq.fastq", compress = FALSE)
You can use the Python package biotite for it (https://www.biotite-python.org).
Let's say you have the following FASTQ file:
#Read:01
CCCAAGGGCCCCCCCCCACTGCGATCACCTGGTTGCTGCCGGGAAAGGAGACCCAGGAGGTGAAACGGACTGGTGAATTG
CGGGGGTAGATATGGCGGGTGACACAAAAACATATAATCGGGCC
+
.+.+:'-FEAC-4'4CA-3-5#/4+?*G#?,<)<E&5(*82C9FH4G315F*DF8-4%F"9?H5535F7%?7#+6!FDC&
+4=4+,#2A)8!1B#,HA18)1*D1A-.HGAED%?-G10'6>:2
#Read:02
AACACTACTTCGCTGTCGCCAAAGGTTGGTGTAGGTCGGACTTCGAATTATCGATACTAGTTAGTAGTACGTCGCGTGGC
GTCAGCTCGTATGCTCTCAGAACAGGGAGAACTAGCACCGTAAGTAACCTAGCTCCCAAC
+
6%9,#'4A0&%.19,1E)E?!9/$.#?(!H2?+E"")?6:=F&FE91-*&',,;;$&?#2A"F.$1)%'"CB?5$<.F/$
7055E>#+/650B6H<8+A%$!A=0>?'#",8:#5%18&+3>'8:28+:5F0);E9<=,+
This is a script, that should do the work:
import biotite.sequence.io.fastq as fastq
import biotite.sequence as seq
# 'GGGGGGGG', 'CCCCCCCC', etc.
consecutive_nucs = [seq.NucleotideSequence(nuc * 8) for nuc in "ACGT"]
fastq_file = fastq.FastqFile("Sanger")
fastq_file.read("example.fastq")
# Iterate over sequence entries in file
for header in fastq_file:
sequence = fastq_file.get_sequence(header)
# Iterative over each of the consecutive sequences
for consecutive_nuc in consecutive_nucs:
# Find all indices, where a match was found
matches = seq.find_subsequence(sequence, consecutive_nuc)
if len(matches) > 0:
# If any match was found report it
print(
f"Found '{consecutive_nuc}' "
f"in sequence '{header}' at position {matches[0]}"
)
This is the output:
Found 'CCCCCCCC' in sequence 'Read:01' at pos 8
I'm using the command line tool from Temperature Monitor, the mac software, which looks like this:
$ /Applications/TemperatureMonitor.app/Contents/MacOS/tempmonitor -c -l -a
SMART Disk Hitachi HTS547550A9E384 (J2250050GMBY3C): 30 C
SMART Disk TOSHIBA MK5065GSXF (71SNCPW4T): 28 C
SMC BATTERY: 30 C
SMC BATTERY POSITION 2: 31 C
SMC BATTERY POSITION 3: 28 C
SMC CPU A DIODE: 47 C
SMC CPU A PROXIMITY: 45 C
SMC GPU 1 CHIP: 40 C
SMC LEFT PALM REST: 28 C
SMC MAIN HEAT SINK 2: 38 C
SMC MAIN HEAT SINK 3: 37 C
SMC MAIN LOGIC BOARD: 36 C
SMC PLATFORM CONTROLLER HUB: 49 C
SMC SSD BAY: 36 C
I want to clean this up a bit. So for example, let's say I want to get the average of the three Battery temperature readings. I thought of piping into grep for Battery, then awking through all the fields for the correct data, but that seems really messy.
So I want the three variables $BATTERY_1, $BATTERY_2, and $BATTERY_3 to have the content 30, 31, and 28 respectively.
Any suggestions on the cleanest way to do so?
It will be easier to create an array and then move the values from the array into the plain variables. It is trivial to do the extraction with awk:
TEMPMON="/Applications/TemperatureMonitor.app/Contents/MacOS/tempmonitor"
battery=( $("$TEMPMON" -c -l -a | awk '/BATTERY/ { print $(NF-1) }') )
BATTERY_1=${battery[0]}
BATTERY_2=${battery[1]}
BATTERY_3=${battery[2]}
To complement #Jonathan Leffler's helpful answer:
If you don't actually need the individual values and instead want the average only, try:
... | awk '/ BATTERY/ {sum+=$(NF-1); ++i} END {OFMT="%.2f"; print sum / i}'
OFMT="%.2f" sets the (printf-style) output number format to 2 decimal places, resulting in 29.67.
Update: The OP, in a comment, asks for output in the format <Item name>: <avg temp> (<temp 1>, <temp 2>, <temp 3>):
... | awk -v itm='BATTERY' '
$0 ~ itm {
vals = vals (i ? " " : "") $(NF-1)
sum += $(NF-1); ++i
}
END {
printf "%s: %.2f (%s)\n", itm, sum / i, vals
}'
-v itm='BATTERY' passes the name of the items to match as awk variable itm.
$0 ~ itm matches (~) the current input line ($0) against itm (interpreted as a regular expression, which in this simple case performs substring matching).
awk splits input lines into fields $1, $2, ... by whitespace by default, and stores the number of fields in special variable NF. Since the values in the input data are in the next-to-last field, $(NF-1) references each line's value.
vals = ... builds up a string list of matching values; note how merely placing strings and variables next to each other causes them to be concatenated (as strings).
(i ? " " : "") is a C-style ternary conditional that returns a single space if condition i is true (i.e., if variable i has a nonzero value), and an empty string otherwise. In other words: if the value is not the first one, append a space before appending the value to the list of values built up so far. Note that uninitialized variables in awk default to an empty string in a string context, and 0 (false) in a numeric/Boolean context.
sum += ... sums up the values; ++i keeps the count of values.
END is a special pattern whose associated action (block) is processed after all input lines.
printf, for output based on a format (template) string, works like its C counterpart, and in this case outputs the item name (1st %s, instantiated with itm), the average with 2 decimal places (%.2f, instantiated with sum / i) and the list of values (last %s, instantiated with vals).
Locked. This question and its answers are locked because the question is off-topic but has historical significance. It is not currently accepting new answers or interactions.
Introduction
A valid Sudoku grid is filled with numbers 1 to 9, with no number occurring more than once in each sub-block of 9, row or column. Read this article for further details if you're unfamiliar with this popular puzzle.
Challenge
The challenge is to write the shortest program that validates a Sudoku grid that might not be full.
Input will be a string of 9 lines of 9 characters each, representing the grid. An empty cell will be represented by a .. Your output should be Valid if the grid is valid, otherwise output Invalid.
Example
Input
123...789
...456...
456...123
789...456
...123...
564...897
...231...
897...564
...564...
Output
Valid
Input
123456789
987654321
123456789
123456789
987654321
123456789
123456789
987654321
123456789
Output
Invalid
Code Golf Rules
Please post your shortest code in any language that solves this problem. Input and output may be handled via stdin and stdout or by other files of your choice.
Winner will be the shortest solution (by byte count) in a language with an implementation existing prior to the posting of this question. So while you are free to use a language you've just made up in order to submit a 0-byte solution, it won't count, and you'll probably get downvotes.
Golfscript: 56
n%{zip''+9/.{'.'-..&=}%$0=\}:|2*{3/}%|;**"InvV"3/="alid"
C: 165 162 161 160 159
int v[1566],x,y=9,c,b;main(){while(y--)for(x=9;x--+1;)if((c
=getchar()*27)>1242)b|=v[x+c]++|v[y+9+c]++|v[x-x%3+y/3+18+c]
++;puts(b?"Invalid":"Valid");return 0;}
The two newlines are not needed. One char saved by josefx :-) ...
Haskell: 207 230 218 195 172
import List
t=take 3
h=[t,t.drop 3,drop 6]
v[]="V"
v _="Inv"
f s=v[1|v<-[s,transpose s,[g=<<f s|f<-h,g<-h]],g<-map(filter(/='.'))v,g/=nub g]++"alid\n"
main=interact$f.lines
Perl: 168 128
$_=join'',<>;#a=/.../g;print+(/(\d)([^\n]{0,8}|(.{10})*.{9})\1/s
+map"#a[$_,$_+3,$_+6]"=~/(\d).*\1/,0..2,9..11,18..20)?Inv:V,alid
The first regex checks for duplicates that are in the same row and column; the second regex handles duplicates in the "same box".
Further improvement is possible by replacing the \n in the first regex with a literal newline (1 char), or with >= Perl 5.12, replacing [^\n] with \N (3 char)
Earlier, 168 char solution:
Input is from stdin, output is to stderr because it makes things so easy. Linebreaks are optional and not counted.
$_=join'',<>;$m=alid.$/;$n=Inv.$m;/(\d)(\N{0,8}|(.{10})*.{9})\1/s&&
die$n;#a=/.../g;for$i(0,8,17){for$j($i..$i+2){
$_=$a[$j].$a[$j+3].$a[$j+6];/(\d).*\1/&&die$n}}die"V$m"
Python: 230 221 200 185
First the readable version at len=199:
import sys
r=range(9)
g=[raw_input()for _ in r]
s=[[]for _ in r*3]
for i in r:
for j in r:
n=g[i][j]
for x in i,9+j,18+i/3*3+j/3:
<T>if n in s[x]:sys.exit('Invalid')
<T>if n>'.':s[x]+=n
print'Valid'
Since SO doesn't display tab characters, I've used <T> to represent a single tab character.
PS. the same approach minEvilized down to 185 chars:
r=range(9)
g=[raw_input()for _ in r]
s=['']*27
for i in r:
for j in r:
for x in i,9+j,18+i/3*3+j/3:n=g[i][j];s[x]+=n[:n>'.']
print['V','Inv'][any(len(e)>len(set(e))for e in s)]+'alid'
Perl, 153 char
#B contains the 81 elements of the board.
&E tests whether a subset of #B contains any duplicate digits
main loop validates each column, "block", and row of the puzzle
sub E{$V+="#B[#_]"=~/(\d).*\1/}
#B=map/\S/g,<>;
for$d(#b=0..80){
E grep$d==$_%9,#b;
E grep$d==int(($_%9)/3)+3*int$_/27,#b;
E$d*9..$d*9+8}
print$V?Inv:V,alid,$/
Python: 159 158
v=[0]*244
for y in range(9):
for x,c in enumerate(raw_input()):
if c>".":
<T>for k in x,y+9,x-x%3+y//3+18:v[k*9+int(c)]+=1
print["Inv","V"][max(v)<2]+"alid"
<T> is a single tab character
Common Lisp: 266 252
(princ(let((v(make-hash-table))(r "Valid"))(dotimes(y 9)(dotimes(x
10)(let((c(read-char)))(when(>(char-code c)46)(dolist(k(list x(+ 9
y)(+ 18(floor(/ y 3))(- x(mod x 3)))))(when(>(incf(gethash(+(* k
9)(char-code c)-49)v 0))1)(setf r "Invalid")))))))r))
Perl: 186
Input is from stdin, output to stdout, linebreaks in input optional.
#y=map/\S/g,<>;
sub c{(join'',map$y[$_],#$h)=~/(\d).*\1/|c(#_)if$h=pop}
print(('V','Inv')[c map{$x=$_;[$_*9..$_*9+8],[grep$_%9==$x,0..80],[map$_+3*$b[$x],#b=grep$_%9<3,0..20]}0..8],'alid')
(Linebreaks added for "clarity".)
c() is a function that checks the input in #y against a list of lists of position numbers passed as an argument. It returns 0 if all position lists are valid (contain no number more than once) and 1 otherwise, using recursion to check each list. The bottom line builds this list of lists, passes it to c() and uses the result to select the right prefix to output.
One thing that I quite like is that this solution takes advantage of "self-similarity" in the "block" position list in #b (which is redundantly rebuilt many times to avoid having #b=... in a separate statement): the top-left position of the ith block within the entire puzzle can be found by multiplying the ith element in #b by 3.
More spread out:
# Grab input into an array of individual characters, discarding whitespace
#y = map /\S/g, <>;
# Takes a list of position lists.
# Returns 0 if all position lists are valid, 1 otherwise.
sub c {
# Pop the last list into $h, extract the characters at these positions with
# map, and check the result for multiple occurences of
# any digit using a regex. Note | behaves like || here but is shorter ;)
# If the match fails, try again with the remaining list of position lists.
# Because Perl returns the last expression evaluated, if we are at the
# end of the list, the pop will return undef, and this will be passed back
# which is what we want as it evaluates to false.
(join '', map $y[$_], #$h) =~ /(\d).*\1/ | c(#_) if $h = pop
}
# Make a list of position lists with map and pass it to c().
print(('V','Inv')[c map {
$x=$_; # Save the outer "loop" variable
[$_*9..$_*9+8], # Columns
[grep$_%9==$x,0..80], # Rows
[map$_+3*$b[$x],#b=grep$_%9<3,0..20] # Blocks
} 0..8], # Generates 1 column, row and block each time
'alid')
Perl: 202
I'm reading Modern Perl and felt like coding something... (quite a cool book by the way:)
while(<>){$i++;$j=0;for$s(split//){$j++;$l{$i}{$s}++;$c{$j}{$s}++;
$q{(int(($i+2)/3)-1)*3+int(($j+2)/3)}{$s}++}}
$e=V;for$i(1..9){for(1..9){$e=Inv if$l{$i}{$_}>1or$c{$i}{$_}>1or$q{$i}{$_}>1}}
print $e.alid
Count is excluding unnecessary newlines.
This may require Perl 5.12.2.
A bit more readable:
#use feature qw(say);
#use JSON;
#$json = JSON->new->allow_nonref;
while(<>)
{
$i++;
$j=0;
for $s (split //)
{
$j++;
$l{$i}{$s}++;
$c{$j}{$s}++;
$q{(int(($i+2)/3)-1)*3+int(($j+2)/3)}{$s}++;
}
}
#say "lines: ", $json->pretty->encode( \%l );
#say "columns: ", $json->pretty->encode( \%c );
#say "squares: ", $json->pretty->encode( \%q );
$e = V;
for $i (1..9)
{
for (1..9)
{
#say "checking {$i}{$_}: " . $l{$i}{$_} . " / " . $c{$i}{$_} . " / " . $q{$i}{$_};
$e = Inv if $l{$i}{$_} > 1 or $c{$i}{$_} > 1 or $q{$i}{$_} > 1;
}
}
print $e.alid;
Ruby — 176
f=->x{x.any?{|i|(i-[?.]).uniq!}}
a=[*$<].map{|i|i.scan /./}
puts f[a]||f[a.transpose]||f[a.each_slice(3).flat_map{|b|b.transpose.each_slice(3).map &:flatten}]?'Invalid':'Valid'
Lua, 341 bytes
Although I know that Lua isn't the best golfing language, however, considering it's size, I think it's worth posting it ;).
Non-golfed, commented and error-printing version, for extra fun :)
i=io.read("*a"):gsub("\n","") -- Get input, and strip newlines
a={{},{},{}} -- checking array, 1=row, 2=columns, 3=squares
for k=1,3 do for l=1,9 do a[k][l]={0,0,0,0,0,0,0,0,0}end end -- fillup array with 0's (just to have non-nils)
for k=1,81 do -- loop over all numbers
n=tonumber(i:sub(k,k):match'%d') -- get current character, check if it's a digit, and convert to a number
if n then
r={math.floor((k-1)/9)+1,(k-1)%9+1} -- Get row and column number
r[3]=math.floor((r[1]-1)/3)+3*math.floor((r[2]-1)/3)+1 -- Get square number
for l=1,3 do v=a[l][r[l]] -- 1 = row, 2 = column, 3 = square
if v[n] then -- not yet eliminated in this row/column/square
v[n]=nil
else
print("Double "..n.." in "..({"row","column","square"}) [l].." "..r[l]) --error reporting, just for the extra credit :)
q=1 -- Flag indicating invalidity
end
end
end
end
io.write(q and"In"or"","Valid\n")
Golfed version, 341 bytes
f=math.floor p=io.write i=io.read("*a"):gsub("\n","")a={{},{},{}}for k=1,3 do for l=1,9 do a[k][l]={0,0,0,0,0,0,0,0,0}end end for k=1,81 do n=tonumber(i:sub(k,k):match'%d')if n then r={f((k-1)/9)+1,(k-1)%9+1}r[3]=f((r[1]-1)/3)+1+3*f((r[2]-1)/3)for l=1,3 do v=a[l][r[l]]if v[n]then v[n]=nil else q=1 end end end end p(q and"In"or"","Valid\n")
Python: 140
v=[(k,c) for y in range(9) for x,c in enumerate(raw_input()) for k in x,y+9,(x/3,y/3) if c>'.']
print["V","Inv"][len(v)>len(set(v))]+"alid"
ASL: 108
args1["\n"x2I3*x;{;{:=T(T'{:i~{^0}?})}}
{;{;{{,0:e}:;{0:^},u eq}}/`/=}:-C
dc C#;{:|}C&{"Valid"}{"Invalid"}?P
ASL is a Golfscript inspired scripting language I made.