I have a question about a regex. Given this part of a regex:
(.[^\\.]+)
The part [^\.]+ Does this mean get everything until the first dot? So with this text:
Hello my name is Martijn. I live in Holland.
I get 2 results: both sentences. But when I leave the + sign, I get 2 two characters: he, ll, o<space>, my, etc. Why is that?
Your regex .[^\\.]+ means:
Match any character
Match any character until you get slash or a dot ".". Note that [^\\.] means NOT slash or NOT dot, which means either a dot or a slash is not a match. It will keep on matching characters until it founds a dot or slash because of the "+" at the end. It is called a greedy quantifier because of that.
When you input (quotes not included): "Hello my name is Martijn. I live in Holland."
The matches are:
Hello my name is Martijn
. I live in Holland
Note that the dot is not included in the first match since it stops at n in Martijn and the second match starts with the dot.
When you remove the +: (.[^\\.])
It just means:
Match any character
Match any character except a dot or a slash.
Because a dot outside a character class (ie, not between []) means (almost) any character.
So, .[^\\.] means match (almost) any character followed by something which is not a dot nor a backslash (dots don't need to be escaped in a character class to mean just a dot, but backslashes do),
This, in your example, is h (any character) e (not a dot nor a backslash) and so on and so forth.
Whereas with a + (one or more of not a dot nor a backslash) you will match all characters which are not dots until a dot.
The regex means:
any one character followed by more than zero characters that are not a backslash or a period.
Related
I'm creating a small example to be put into mutate(). Not sure why this doesn't work.
> str_extract("rs1234-<b>C</b>","^rs*\\d$")
[1] NA
I'd be great if you can point to my misunderstanding of the language instead of merely providing a solution. I expect to get "rs1234".
The ^rs*\d$ regex matches
^ - start of string
rs* - r and zero or more occurrences of s char
\d - a digit
$ - end of string.
So, your pattern matches strings like rsssss1, r3, etc.
You need
str_extract("rs1234-<b>C</b>", "^rs\\d+")
where ^rs\d+ matches rs at the start of string and then one or more digits. See this regex demo.
But if I just want the substring in between "rs" and the last number. What should I do?
You would use rs.*\d:
str_extract("rs1234-<b>C</b>", "rs.*\\d")
where rs.*\d matches rs, then any zero or more chars other than line break chars as many as possible and then a digit.
NOTE: If you need to match line endings, too, you need to prepend the last pattern with (?s) inline DOTALL modifier.
See this regex demo.
I need to identify matching course number that have xx.3xxxxxx.
These are some examples of the course numbers.
26.3730004
27.0210000
26.3730009
26.7114001
23.9610071
26.0A34430
23.3670005
26.0B05430
I tried many patterns one example I used is the pattern below. It did not get any match.
"[^0-9]{2}\Q.\E3[^0-9]+$"
I tried using grep and grepl. I actually need the code to return indexes.
This code shows my attempt to tag the rows that have matches.
Teacher$virtual[
which(
grepl("[^0-9]{2}\\Q.\\E3[^0-9]+$",Teacher$CourseNumber))]
<- "1"
I need to remove any row from my dataframe that have the course number with that pattern. XX.3XXXXXX
But, my code did not find any match. Can you please help me?
You should use
grepl("^[0-9]{2}\\.3", Teacher$CourseNumber)
See the regex graph:
Details:
^ - start of a string
[0-9]{2} - two digits
\\. - a dot (note that a regex escape is a literal backslash, but inside a string literal, "...", a single backslash is used to form string escape sequences, hence the backslash must be double to obtain a literal backslash char necessary for a regex escape)
3 - a 3 char.
NOTE: If you want to use in-pattern quoting with \Q and \E (in between which all chars are treated literally) you need to use PCRE regex, add perl=TRUE and use
grepl("^[0-9]{2}\\Q.\\E3", Teacher$CourseNumber, perl=TRUE)
Now, the dot is treated as a literal dot, not a . metacharacter that matches any char but a line break char (in a PCRE regex, . does not match line break chars by default).
Here, this simple expression would likely cover that:
^[0-9]{2}\.[3].+$
which has a [3] boundary right after the .. It would probably work without start and end anchors:
[0-9]{2}\.[3].+
Demo
We can add or reduce the boundaries, if it'd be necessary.
I have the following QRegExpValidator
QRegExpValidator doubleValidator = new QRegExpValidator(QRegExp("[-+]?[0-9]*[\\.,]?[0-9]+([eE][-+]?[0-9]+)?"));
It's supposed to be a Double numbers validator that accepts numbers, only one "e" sign, one comma OR dot and one + or - sign at the beggining of the string or after the "e" sign. It works for every case, except that it allows the string to start with a comma or dot. I tried to use [^\\.,] and variations and they did in fact work, but in this case, it would also allow to put two +- signs.
How can I make this to work?
The [-+]?[0-9]*[.,]?[0-9]+([eE][-+]?[0-9]+)? pattern allows the , or . at the start because [-+]? and [0-9]* can match empty strings due to the ? (one or zero occurrences) and * (zero or more occurrences) quantifiers, and then [.,] matches a single occurrence of . or ,. Besides, if the method you are using does not anchor the pattern by default, you also need ^ and $ anchors around the pattern.
I suggest fixing that with
"^[-+]?[0-9]+([.,][0-9]+)?([eE][-+]?[0-9]+)?$"
^ ^^^^^^^^^^^^^^ ^
Note you do not need to escape the dot inside a character class, [.] always matches a dot char only.
The [0-9]+([.,][0-9]+)? matches 1+ digits and then an optional sequence of a . or , followed with 1+ digits.
I would like to split strings like the following:
x <- "abc-1230-xyz-[def-ghu-jkl---]-[adsasa7asda12]-s-[klas-bst-asdas foo]"
by dash (-) on the condition that those dashes must not be contained inside a pair of []. The expected result would be
c("abc", "1230", "xyz", "[def-ghu-jkl---]", "[adsasa7asda12]", "s",
"[klas-bst-asdas foo]")
Notes:
There is no nesting of square brackets inside each other.
The square brackets can contain any characters / numbers / symbols except square brackets.
The other parts of the string are also variable so that we can only assume that we split by - whenever it's not inside [].
There's a similar question for python (How to split a string by commas positioned outside of parenthesis?) but I haven't yet been able to accurately adjust that to my scenario.
You could use look ahead to verify that there is no ] following sooner than a [:
-(?![^[]*\])
So in R:
strsplit(x, "-(?![^[]*\\])", perl=TRUE)
Explanation:
-: match the hyphen
(?! ): negative look ahead: if that part is found after the previously matched hyphen, it invalidates the match of the hyphen.
[^[]: match any character that is not a [
*: match any number of the previous
\]: match a literal ]. If this matches, it means we found a ] before finding a [. As all this happens in a negative look ahead, a match here means the hyphen is not a match. Note that a ] is a special character in regular expressions, so it must be escaped with a backslash (although it does work without escape, as the engine knows there is no matching [ preceding it -- but I prefer to be clear about it being a literal). And as backslashes have a special meaning in string literals (they also denote an escape), that backslash itself must be escaped again in this string, so it appears as \\].
Instead of splitting, extract the parts:
library(stringr)
str_extract_all(x, "(\\[[^\\[]*\\]|[^-])+")
I am not familiar with r language, but I believe it can do regex based search and replace. Instead of struggling with one single regex split function, I would go in 3 steps:
replace - in all [....] parts by a invisible char, like \x99
split by -
for each element in the above split result(array/list), replace \x99 back to -
For the first step, you can find the parts by \[[^]]
I have a regular expression
^[a-zA-Z+#-.0-9]{1,5}$
which validates that the word contains alpha-numeric characters and few special characters and length should not be more than 5 characters.
How do I make this regular expression to accept a maximum of five words matching the above regular expression.
^[a-zA-Z+#\-.0-9]{1,5}(\s[a-zA-Z+#\-.0-9]{1,5}){0,4}$
Also, you could use for example [ ] instead of \s if you just want to accept space, not tab and newline. And you could write [ ]+ (or \s+) for any number of spaces (or whitespaces), not just one.
Edit: Removed the invalid solution and fixed the bug mentioned by unicornaddict.
I believe this may be what you're looking for. It forces at least one word of your desired pattern, then zero to four of the same, each preceded by one or more white-space characters:
^XX(\s+XX){0,4}$
where XX is your actual one-word regex.
It's separated into two distinct sections so that you're not required to have white-space at the end of the string. If you want to allow for such white-space, simply add \s* at that point. For example, allowing white-space both at start and end would be:
^\s*XX(\s+XX){0,4}\s*$
You regex has a small bug. It matches letters, digits, +, #, period but not hyphen and also all char between # and period. This is because hyphen in a char class when surrounded on both sides acts as a range meta char. To avoid this you'll have to escape the hyphen:
^[a-zA-Z+#\-.0-9]{1,5}$
Or put it at the beg/end of the char class, so that its treated literally:
^[-a-zA-Z+#-.0-9]{1,5}$
^[a-zA-Z+#.0-9-]{1,5}$
Now to match a max of 5 such words you can use:
^(?:[a-zA-Z+#\-.0-9]{1,5}\s+){1,5}$
EDIT: This solution has a severe limitation of matching only those input that end in white space!!! To overcome this limitation you can see the ans by Jakob.