This question already has an answer here:
How do I avoid trailing empty items being removed when splitting strings?
(1 answer)
Closed 8 years ago.
In both Ruby and JavaScript I can write expression " x ".split(/[ ]+/)
. In JavaScript I get somehow reasonable result ["", "x", ""], but in Ruby (2.0.0) i get ["", "x"], which is for me quite counterintuitive. I have problems to understand how regular expressions works in Ruby. Why don't I get the same result as in JavaScript or just ["x"]?
From string#split documentation, emphasis my own:
split(pattern=$;, [limit])
If pattern is a String, then its contents are used as the delimiter when splitting str. If pattern is a single space, str is split on whitespace, with leading whitespace and runs of contiguous whitespace characters ignored.
If pattern is a Regexp, str is divided where the pattern matches. Whenever the pattern matches a zero-length string, str is split into individual characters. If pattern contains groups, the respective matches will be returned in the array as well.
If pattern is omitted, the value of $; is used. If $; is nil (which is the default), str is split on whitespace as if ` ' were specified.
If the limit parameter is omitted, trailing null fields are suppressed. If limit is a positive number, at most that number of fields will be returned (if limit is 1, the entire string is returned as the only entry in an array). If negative, there is no limit to the number of fields returned, and trailing null fields are not suppressed.
So if you were to use " x ".split(/[ ]+/, -1) you would get your expected result of ["", "x", ""]
*edited to reflect Wayne's comment
I found this in the C code for String#split, almost right at the end:
if (NIL_P(limit) && lim == 0) {
long len;
while ((len = RARRAY_LEN(result)) > 0 &&
(tmp = RARRAY_AREF(result, len-1), RSTRING_LEN(tmp) == 0))
rb_ary_pop(result);
}
So it actually pops empty strings off the end of the result array before returning! It looks like the creators of Ruby didn't want String#split to return a bunch of empty strings.
Notice the check for NIL_P(limit) -- this accords exactly with what the documentation says, as #dax pointed out.
Related
Trying to extract out the individual elements of a server name in the format "ubuntu-prod-sfo1-01" which would give me the result of ["ubuntu","prod", "sfo1", "01"]. So everything between "-" AND the beginning element and ending element when end and start with "-", respectfully.
"ubuntu-prod-sfo1-01" ==> ["ubuntu","prod", "sfo1", "01"]
My attempts at it have failed as the best solution I could find would give me the first "prod", but would fail to get each of the remaining elements. The problem seems to be the reusing of the '-' between the element.
You can split a string into substrings at a delimiter, in your case -.
string = "ubuntu-prod-sfo1-01"
string.split('-')
=> ["ubuntu", "prod", "sf01", "01"]
From the official documentation
Divides str into substrings based on a delimiter, returning an array of these substrings.
If pattern is a String, then its contents are used as the delimiter when splitting str. If pattern is a single space, str is split on whitespace, with leading whitespace and runs of contiguous whitespace characters ignored.
You don't need a regex for that, you can just do...
"ubuntu-prod-sfo1-01".split('-')
=> ["ubuntu","prod", "sfo1", "01"]
I am currently working on a ruby program to calculate terms. It works perfectly fine except for one thing: brackets. I need to filter the content or at least, to put the content into an array, but I have tried for an hour to come up with a solution. Here is my code:
splitted = term.split(/\(+|\)+/)
I need an array instead of the brackets, for example:
"1-(2+3)" #=>["1", "-", ["2", "+", "3"]]
I already tried this:
/(\((?<=.*)\))/
but it returned:
Invalid pattern in look-behind.
Can someone help me with this?
UPDATE
I forgot to mention, that my program will split the term, I only need the content of the brackets to be an array.
If you need to keep track of the hierarchy of parentheses with arrays, you won't manage it just with regular expressions. You'll need to parse the string word by word, and keep a stack of expressions.
Pseudocode:
Expressions = new stack
Add new array on stack
while word in string:
if word is "(": Add new array on stack
Else if word is ")": Remove the last array from the stack and add it to the (next) last array of the stack
Else: Add the word to the last array of the stack
When exiting the loop, there should be only one array in the stack (if not, you have inconsistent opening/closing parentheses).
Note: If your ultimate goal is to evaluate the expression, you could save time and parse the string in Postfix aka Reverse-Polish Notation.
Also consider using off-the-shelf libraries.
A solution depends on the pattern you expect between the parentheses, which you have not specified. (For example, for "(st12uv)" you might want ["st", "12", "uv"], ["st12", "uv"], ["st1", "2uv"] and so on). If, as in your example, it is a natural number followed by a +, followed by another natural number, you could do this:
str = "1-( 2+ 3)"
r = /
\(\s* # match a left parenthesis followed by >= 0 whitespace chars
(\d+) # match one or more digits in a capture group
\s* # match >= 0 whitespace chars
(\+) # match a plus sign in a capture group
\s* # match >= 0 whitespace chars
(\d+) # match one or more digits in a capture group
\s* # match >= 0 whitespace chars
\) # match a right parenthesis
/x
str.scan(r0).first
=> ["2", "+", "3"]
Suppose instead + could be +, -, * or /. Then you could change:
(\+)
to:
([-+*\/])
Note that, in a character class, + needn't be escaped and - needn't be escaped if it is the first or last character of the class (as in those cases it would not signify a range).
Incidentally, you received the error message, "Invalid pattern in look-behind" because Ruby's lookarounds cannot contain variable-length matches (i.e., .*). With positive lookbehinds you can get around that by using \K instead. For example,
r = /
\d+ # match one or more digits
\K # forget everything previously matched
[a-z]+ # match one or more lowercase letters
/x
"123abc"[r] #=> "abc"
TL;DR
I need some help making a regex that will match any commas in a string that are side by side with unlimited white space around them and between them. The commas and their surrounding white space cannot be within matching single quotes or double quotes. I then need to capture the non-whitespace values from around those commas and count how many of those commas there are.
The values captured from around the commas will become their own values in the final array, while the commas that were counted will become nil values that are added to the final array.
Explanation of the problem:
This is a pretty complex problem so any help is greatly appreciated. I'm adding functionality to a library I've been using for a while now. I have this string that contains an array
"['d,og,f:asdf,:hello,",,\",,alsee',,,'ho,la', "-123,4,5.3", true, :good, false,,, "gr\'\'\'true,\',\'ee\"n", ":::testme", true]"
I would like to split this string only around select commas so that I have an array containing the following values
'd,og,f:asdf,:hello,",,\",,alsee'
nil
nil
'ho,la'
"-123,4,5.3"
true
:good
false
nil
nil
"gr\'\'\'true,\',\'ee\"n"
":::testme"
true
Then nil values are coming from the side by side commas that are not contained in any string. I wrote the following regex to split the string above (I already got rid of the start and end brackets):
/(?<=(?:['\"]|false|true|^|,)),(?=(?:\s*(?:(?::[\w]+)|(?:(?::?(?:\"[\s\S]*\")|(?:'[\s\S]*'))|(?:false|true)))\s*(?:,|$)))/
This splits the string so I get these values:
(0) "'d,og,f:asdf,:hello,",,\",,alsee',,"
(1) "'ho,la'"
(2) " "-123,4,5.3""
(3) " true"
(4) " :good, false,,"
(5) " "gr\'\'\'true,\',\'ee\"n""
(6) " ":::testme""
(7) " true"
All the values are strings as can be seen by their surrounding double quotes. They will not all end up that way though. A true or false will be converted to a boolean. The values surrounded by internal quotes will end up as strings. Then a value preceded with a : will end up as a symbol.
There are problems with the values at index 0 and 4. Index 0 should be this:
(0.0) "'d,og,f:asdf,:hello,",,\",,alsee'"
(0.1) nil
(0.2) nil
As you can see, the two commas at the end are gone. They have become the two nil values you see above. Then the string starts at the first single quote and ends at the last single quote, signifying that this value in the array is a string.
Then index 4 (" :good, false,,") should be this:
(4.0) " :good"
(4.1) " false"
(4.2) nil
(4.3) nil
The two commas at the end have become nil. Then " false" is it's own value which will later be converted to a boolean, while " :good" is also it's own value and will later be converted to a symbol.
To fix the problem with index 4 I have all the values run through a second regex. Here it is:
/^(\s*:(?:(?:[\w]+|\"[\s\S]+\"|'[\s\S]+')\s*)),([\s\S]*)$/
Instead of splitting this one I get the capture groups. It ends up returning this array for the value at index 4:
(4.0) " :good"
(4.1) " false,,"
That's what I wanted except for one problem. The value at index 4.1 (" false,,") has the two trailing commas which should be nil values in the array.
I need some help making a regex that will match any commas in a string that are side by side with unlimited white space around them and between them. The commas and their surrounding white space cannot be within matching single quotes or double quotes. I then need to capture the non-whitespace values from around those commas and count how many of those commas there are.
The values captured from around the commas will become their own values in the final array, while the commas that were counted will become nil values that are added to the final array.
"['d,og,f:asdf,:hello,"
,,\
",,alsee',,,'ho,la', "
-123,4,5.3
", true, :good, false,,, "
gr\
'\'
I count 4 strings. 3 in double quotes and the last one in single quotes?
You say this is broken down into smaller strings by your regx. But what about the characters outside the 4 strings?
Sorry, it looks a bit of a mess.
Try putting it all in a here document string and then breaking it down by a regx.
I finally figured it out myself. You can see how it fits in with the rest if you look at the description of the question above.
/^(([\s]*,)*)[\s]*((?::[\w]+)|(?::?(?:\"[\s\S]*\")|(?:'[\s\S]*')|false|true))?(([\s]*,)*)$/
I just discovered the following odd behavior with String#split:
"a\tb c\nd".split
=> ["a", "b", "c", "d"]
"a\tb c\nd".split(' ')
=> ["a", "b", "c", "d"]
"a\tb c\nd".split(/ /)
=> ["a\tb", "c\nd"]
The source (string.c from 2.0.0) is over 200 lines long and contains a passage like this:
/* L 5909 */
else if (rb_enc_asciicompat(enc2) == 1) {
if (RSTRING_LEN(spat) == 1 && RSTRING_PTR(spat)[0] == ' '){
split_type = awk;
}
}
Later, in the code for the awk split type, the actual argument isn't even used any more and does the same as a plain split.
Does anyone else feel that this is somehow broken?
Are there good reasons for this?
Does “magic” like that happen more often than most people might think in Ruby?
It's consistent with Perl's split() behavior. Which in turn is based on Gnu awk's split(). So it's a long-standing tradition with origins in Unix.
From the perldoc on split:
As another special case, split emulates the default behavior of the
command line tool awk when the PATTERN is either omitted or a literal
string composed of a single space character (such as ' ' or "\x20" ,
but not e.g. / / ). In this case, any leading whitespace in EXPR is
removed before splitting occurs, and the PATTERN is instead treated as
if it were /\s+/ ; in particular, this means that any contiguous
whitespace (not just a single space character) is used as a separator.
However, this special treatment can be avoided by specifying the
pattern / / instead of the string " " , thereby allowing only a single
space character to be a separator.
Check out the documentation, this part in particular:
If pattern is a String, then its contents are used as the delimiter
when splitting str. If pattern is a single space, str is split on
whitespace, with leading whitespace and runs of contiguous whitespace
characters ignored.
If pattern is omitted, the value of $; is used. If $; is nil (which is
the default), str is split on whitespace as if ` ‘ were specified.
You can use a regexp to split the string.
Edit: I solved this by using strip! to remove leading and trailing whitespaces as I show in this video. Then, I followed up by restoring the white space at the end of each string the array by iterating through and adding whitespace. This problem varies from the "dupe" as my intent is to keep the whitespace at the end. However, strip! will remove both the leading and trailing whitespace if that is your intent. (I would have made this an answer, but as this is incorrectly marked as a dupe, I could only edit my original question to include this.)
I have an array of words where I am trying to remove any whitespace that may exist at the beginning of the word instead of at the end. rstrip! just takes care of the end of a string. I want whitespaces removed from the beginning of a string.
example_array = ['peanut', ' butter', 'sammiches']
desired_output = ['peanut', 'butter', 'sammiches']
As you can see, not all elements in the array have the whitespace problem, so I can't just delete the first character as I would if all elements started with a single whitespace char.
Full code:
words = params[:word].gsub("\n", ",").delete("\r").split(",")
words.delete_if {|x| x == ""}
words.each do |e|
e.lstrip!
end
Sample text that a user may enter on the form:
Corn on the cob,
Fibonacci,
StackOverflow
Chat, Meta, About
Badges
Tags,,
Unanswered
Ask Question
String#lstrip (or String#lstrip!) is what you're after.
desired_output = example_array.map(&:lstrip)
More comments about your code:
delete_if {|x| x == ""} can be replaced with delete_if(&:empty?)
Except you want reject! because delete_if will only return a different array, rather than modify the existing one.
words.each {|e| e.lstrip!} can be replaced with words.each(&:lstrip!)
delete("\r") should be redundant if you're reading a windows-style text document on a Windows machine, or a Unix-style document on a Unix machine
split(",") can be replaced with split(", ") or split(/, */) (or /, ?/ if there should be at most one space)
So now it looks like:
words = params[:word].gsub("\n", ",").split(/, ?/)
words.reject!(&:empty?)
words.each(&:lstrip!)
I'd be able to give more advice if you had the sample text available.
Edit: Ok, here goes:
temp_array = text.split("\n").map do |line|
fields = line.split(/, */)
non_empty_fields = fields.reject(&:empty?)
end
temp_array.flatten(1)
The methods used are String#split, Enumerable#map, Enumerable#reject and Array#flatten.
Ruby also has libraries for parsing comma seperated files, but I think they're a little different between 1.8 and 1.9.
> ' string '.lstrip.chop
=> "string"
Strips both white spaces...