Match a string with regular expression - regex

I have a problem to create a regular expression. I have a string like this (a simple json response froma webserver):
{
"key1":valueInt1,
"key2":valueInt2,
[...]
"specialKey":"",
"subtitle":"Every kind of character, like char, num or punct",
"key3":"Useful line",
}
What I want is to delete the keys "specialKey" and its value and "subtitle" with its value. Both of them can be empty (like specialKey in my example).
I have tried something like that:
(\"subtitle\"\:\")([:punct:]*[:space:]*[:word:]*)*(\",)
Without success. The error is the part ([:punct:][:space:][:word:]) to match my random sentence.
Thanks for your help!


JSON is not a regular language so regular expressions are not the best tool for this kind of manipulation. You better use a JSON parser for achieving your goal.


Character classes are usually only used within bracket expressions. If you change what you have into:
(\"subtitle\"\:\")([[:punct:][:space:][:word:]]*)*(\",)
It should work.

Related

Regex HTTP Response Body Message

I use a jmeter for REST testing.
I have made a HTTP Request, and this is the response data:
{"id":11,"name":"value","password":null,"status":"ACTIVE","lastIp":"0.0.0.0","lastLogin":null,"addedDate":1429090984000}
I need just the ID (which is 11) in
{"id":11,....
I use the REGEX below :
([0-9].+?)
It works perfectly but it will be a problem if my ID more than 2 digits. I need to change the REGEX to :
([0-9][0-9].+?)
Is there any dynamic REGEX for my problem. Thank you for your attention.
Regards,
Stefio

If you want any integer between {"id": and , use the following Regular Expression:
{"id":(\d+),
However the smarter way of dealing with JSON data could be JSON Path Extractor (available via JMeter Plugins), going forward this option can be much easier to use against complex JSON.
See Using the XPath Extractor in JMeter guide (scroll down to "Parsing JSON") to learn more on syntax and use cases.

I suggest using the following regular expression:
"id":([^,]*),
This will first find "id": and then look for anything that is not a comma until it finds a comma. Note the character grouping is only around the value of the ID.
This will work for ANY length ID.
Edit:
The same concept works for almost any JSON data, for example where the value is quoted:
"key":"([^"]*)"
That regular expression will extract the value from given key, as long as value is quoted and does not contain quotes. It first finds "key": and then matches anything that is not a quote until the next quote.

You can use the quantifier like this:
([0-9]{2,}.+?)
It will catch 2 or more digits, and then any symbol, 1 or more times. If you want to allow no other characters after the digits, use * instead of +:
([0-9]{2,}.*?)
Regex demo

regex to match everything except character

I have a payload that contains the following:
����\�p�a�t�r�i�c�k�-�t�e�s�t�-�f�i�l�e�.�t�x�t������x�SMB2
I'm looking to extract the file name of patrick-test-file.txt
I can get close by using this, but it continues to include everything (including ascii characters)
[\\\\](.*?)x�SMB2
With a result of this: �p�a�t�r�i�c�k�-�t�e�s�t�-�f�i�l�e�.�t�x�t������ for the capture group.
How would I just match the characters of the file name, which could be anything of variable length, and could contain alphanumeric characters? Is this possible with pure regex?
Any help is much appreciated.

Sometimes you just can't do a single language-agnostic Regular Expression to accomplish something. And sometimes (usually) it is more performant to do a series of string functions.
I wouldn't personally accept any solution which has hard-coded values, such as x�SMB2.
If you want to use Regular Expressions only, you can first select the File-Name portion like so: (([-\w\d.\\]+)[^-\w\d.\\]?)+, then go ahead and replace [^-\w\d.\\] with nothing "".
Honestly, given the limited detail, the best function is like so:
var fileName = "\patrick-test-file.txt";
But half-joking aside, and with that limited detail, your best bet is to do a couple string functions:
var yuckyString = #"����\�p�a�t�r�i�c�k�-�t�e�s�t�-�f�i�l�e�.�t�x�t������x�SMB2";
var fileNameArea = yuckyString.Split(new[] { "��" }, StringSplitOptions.RemoveEmptyEntries)[0];
var fileName = fileNameArea.Replace("�", "");
Granted, there was no language listed, so I'm using C#. Also, the answer would change if there were irregularities with those special characters. With the limited info, the pattern seems clear.

Extract all tokens from string using regex in Scala

I have a string like "httpx://__URL__/__STUFF__?param=value"
This sample is a url by convention...it could be anything with zero or more __X__ tokens in it.
I want to use a regex to extract a list of all the tokens, so output here would be List("__URL__","__STUFF__"). Remember, I don't know beforehand how many (if any) tokens may be in the input string.
I've been struggling but unable to come up with a regex expression that will do the trick.
Something like this did not work:
(?:.?(__[a-zA-Z0-9]+__).?)+

Scala Regex, which is just a wrapper around Java Regex, will never return multiple subgroups for repetitions.
The only way about it is to have a regex for the token, and then find it multiple times. You pretty much already have everything you want:
"__[a-zA-Z0-9]+__".r findAllIn "httpx://__URL__/__STUFF__?param=value"
That returns an Iterator. Use .toSeq or similar to convert into a collection.

Greg, have you tried a simple
_+[^_]+_+
This will match all the __TOKENS__
It doesn't do any check for any __TOKENLIKE__ string after the ?params, but you have mentioned you are not only using that for urls. If you need some refinement, please let us know.

Combine a regex with split:
def urlPathComponents(s: String): Option[Array[String]] =
"""(?<=http(s?)://)[^?]+""".r findFirstIn s map (_.split("/"))

Regex Capturing Groups in Vala

Is there such a thing? I've been looking around the Vala API and the Regex object seems to have no support for capturing groups so that I can reference them later. Is there currently any way to get around this apparent limitation? Say I'm parsing a string out of a group of strings (the contents of a file) for a given pattern like:
parameter = value
But I want the syntax to be lax so that it could also say parameter=value or parameter = value etc... The first idea that springs to mind is using regular expressions with capturing groups but there seems to be no support for this feature as a part of Vala right now, as far as I can see.
The only alternative I can come up with is splitting the string with a regular expression that matches whitespace so that I end up with an array I can analyze, but then again the file might not contain only "parameter = value"-like formatted lines.

It goes something like this. Disclaimer, this is off the top of my head:
Regex r = /^\s*(?P<parameter>.*)\s*=\s*(?P<value>.*)\s*$/;
MatchInfo info;
if(r.match(the_string, 0, out info)) {
var parameter = info.fetch_named("parameter");
var value = info.fetch_named("value");
}

regular expression: how to ignore rest of the line

I have an input like this (a JSON format)
{"location":[{"id":"1BCDEFGHIJKLM","somename":"abcd","fewname":"xyzland","sid":"","sname":"","regionname":"Zee-Whole","type":"some","siteCode":"","someCode":"ROTXY","fewCode":"NL","pCode":"ROTXY","someid":"1BCDEFGHIJKLM","fewid":"GIC8"},{"id":"7823XYZHMOPRE","somename":"abcd Junction","fewname":"United States","sid":"","sname":"","regionname":"New York","type":"some","siteCode":"","someCode":"USRTJ","fewCode":"US","pCode":"USNWK","someid":"7823XYZHMOPRE","fewid":"7823XYZLMOPRE"},{"id":"799XYZHMOPRE","somename":"abcd-Maasvlakte","fewname":"xyzland","sid":"","sname":"","regionname":"Zee-Whole","type":"some","siteCode":"","someCode":"XYROT","fewCode":"NL","pCode":"","someid":"799XYZHMOPRE","fewid":"OIUOWER348534"}]}
Now, I want to pick up the first "id" value which is 1BCDEFGHIJKLM using regular expressions. I have managed upto the point using
[^({"location":[?{"id":")].{0,12} but this is incomplete. Could some one help how do I ignore the rest of the line after the value 1BCDEFGHIJKLM

Regex isn't the way to do this. Whatever platform you are using, it must have a JSON parser.
That will be your best error-free solution.
Assuming you must use regex, you can grab all the id's using "id":"(.*?)", and take the first match.
I found the following article, which might help you.

While messy, how is your regex incomplete?
It could be shortened to ("id":"([^"]+)") which is more readable, and doesn't limit the ID to twelve characters. If that is beneficial.
If you problem is getting more than one result, most languages have a "g" global switch.
In javascript, the following would return "1BCDEFGHIJKLM":
var firstID = str.match(/"id":"([^"]+)"/)[1]
As match()returns an array, in which [0] is the entire returned string, and [1] the first parenthasis.

Don't have to use regex. In your favourite language, split on commas. Then go through each item, check for "id" and split on colon (:). Get the last element. Eg Python
>>> s
'{"location":[{"id":"1BCDEFGHIJKLM","somename":"abcd","fewname":"xyzland","sid":"","sname":"","regionname":"Zee-Whole","type":"some","siteCode":"","someCode":"ROTXY","fewCode":"NL","pCode":"ROTXY","someid":"1BCDEFGHIJKLM","fewid":"GIC8"},{"id":"7823XYZHMOPRE","somename":"abcd Junction","fewname":"United States","sid":"","sname":"","regionname":"New York","type":"some","siteCode":"","someCode":"USRTJ","fewCode":"US","pCode":"USNWK","someid":"7823XYZHMOPRE","fewid":"7823XYZLMOPRE"},{"id":"799XYZHMOPRE","somename":"abcd-Maasvlakte","fewname":"xyzland","sid":"","sname":"","regionname":"Zee-Whole","type":"some","siteCode":"","someCode":"XYROT","fewCode":"NL","pCode":"","someid":"799XYZHMOPRE","fewid":"OIUOWER348534"}]}'
>>> for i in s.split(","):
... if '"id"' in i:
... print i.split(":")[-1]
... break
...
"1BCDEFGHIJKLM"
Of course, ideally, you should use a dedicated JSON parser.