disclaimer - absolutely new to regexes....
I have a string like this:
subject=something||x-access-token=something
For this I need to extract two values. Subject and x-access-token.
As a starting point, I wanted to collect two strings: subject= and x-access-token=. For this here is what I did:
/[a-z,-]+=/g.exec(mystring)
It returns only one element subject=. I expected both of them. Where i am doing wrong?
The g modifier does not affect exec, because exec only returns the first match by specification. What you want is the match method:
mystring.match(/[a-z,-]+=/g)
No regex necessary. Write a tiny parser, it's easy.
function parseValues(str) {
var result = {};
str.split("||").forEach(function (item) {
var parts = item.split("=");
result[ parts[0] /* key */ ] = parts[1]; /* value */
});
return result;
}
usage
var obj = parseValues("subject=something||x-access-token=something-else");
// -> {subject: "something", x-access-token: "something-else"}
var subj = obj.subject;
// -> "something"
var token = obj["x-access-token"];
// -> "something-else"
Additional complications my arise when there is an escaping schema involved that allows you to have || inside a value, or when a value can contain an =.
You will hit these complications with regex approach as well, but with a parser-based approach they will be much easier to solve.
You have to execute exec twice to get 2 extracted strings.
According to MDN: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/RegExp/exec
If your regular expression uses the "g" flag, you can use the exec() method multiple times to find successive matches in the same string.
Usually, people extract all strings matching the pattern one by one with a while loop. Please execute following code in browser console to see how it works.
var regex = /[a-z,-]+=/g;
var string = "subject=something||x-access-token=something";
while(matched = regex.exec(string)) console.log(matched);
You can convert the string into a valid JSON string, then parse it to retrieve an object containing the expected data.
var str = 'subject=something||x-access-token=something';
var obj = JSON.parse('{"' + str.replace(/=/g, '":"').replace(/\|\|/g, '","') + '"}');
console.log(obj);
I don't think you need regexp here, just use the javascript builtin function "split".
var s = "subject=something1||x-access-token=something2";
var r = s.split('||'); // r now is an array: ["subject=something1", "x-access-token=something2"]
var i;
for(i=0; i<r.length; i++){
// for each array's item, split again
r[i] = r[i].split('=');
}
At the end you have a matrix like the following:
y x 0 1
0 subject something1
1 x-access-token something2
And you can access the elements using x and y:
"subject" == r[0][0]
"x-access-token" == r[1][0]
"something2" == r[1][1]
If you really want to do it with a pure regexp:
var input = 'subject=something1||x-access-token=something2'
var m = /subject=(.*)\|\|x-access-token=(.*)/.exec(input)
var subject = m[1]
var xAccessToken = m[2]
console.log(subject);
console.log(xAccessToken);
However, it would probably be cleaner to split it instead:
console.log('subject=something||x-access-token=something'
.split(/\|\|/)
.map(function(a) {
a = a.split(/=/);
return { key: a[0], val: a[1] }
}));
Related
I have a calculation string from a database like:
var calc = "{a}+{b}==2"
and I want to pull all the elements with "{}" so that I can look up their values from a database. What's the fastest way of doing this, so I end up with an ordered array that I can look up, and replace the values back in the string.
I've considered:
- For loop, looking for { then finding the next }
- Split with a map
- IndexOf
Using regex
var exp = /{([^}]+)}/g ,index;
while(index = exp.exec("{a}+{b}==2")) {
console.log(index[1]);
}
.
Demo
Not sure if it's the "fastest" way, but you should consider using a regex.
Something like:
var calc = "{a}+{b}==2";
var re = /{([^}]+)}/g;
var res;
var result = [];
while (res = re.exec(calc))
{
result.push(res[1]);
}
console.log(result);
Your regex may need to be refined based on the actual definition of the {} expressions (based on allowed characters, quoting, etc.).
Once you have received the values back, you can then use replace to replace the values.
var values = {a: 1, b: 3};
var replaced = calc.replace(re,function(match,name) { return values[name]; });
console.log(replaced);
NB: be very careful if you plan to then send this to eval or the like.
Regex comes to the mind first but one other way of implementing this job in O(n) time could be;
function getDatas(s){
var dataOn = false;
return Array.prototype.reduce.call(s,(d,c) => dataOn ? c !== "}" ? (d[d.length-1] += c,d)
: (dataOn = false, d)
: c === "{" ? (dataOn = true, d.push(""),d)
: d, []);
}
var calc = "{a}+{b}+{colorSpace}==2",
result = getDatas(calc);
console.log(result);
So out of curiosity i have done some tests on JSBen and it seems that #jcaron's regex is indeed much efficient. You may extend those tests with any of your other ideas like indexOf or for loop.
Is there a method in E4X to convert an XMLList of primitive data types into an Array?
For example, the Array class has a join method that converts an Array into a string and delimits each item with a comma using join(","). Is there a way to do something similar with an XMLList?
Example:
var xmlList:XMLList = new XMLList();
xmlList[0] = "Hello";
xmlList[1] = "World";
var myItems:String = xmlList.join(","); // there is no join method in E4X
var anArray:Array = myItems.split(",");
I think that the easiest way to do what you are looking for, is to convert your XMLList object to an array an then use Array.join() like this :
var xml:XMLList = new XMLList();
xml[0] = 'hello';
xml[1] = 'world';
trace(xml.toXMLString().split('\n').join(',')); // gives : hello,world
Here of course you have to put the right line break (\n, \r or \r\n) according to your OS.
You can also use use XMLList.prototype to add the join()function :
XMLList.prototype.join = function(sep:String = ','):String {
var str:String = '';
for(var i:int = 0; i < this.length(); i++){
str += this[i] + sep;
}
return str.slice(0, -1); // the String.slice() is used to remove the last ','
}
trace(xml.join(',')); // gives : hello,world
I tried just to give you some manners to do what you want.
Hope that can help.
My array:
var str=['data1,data2 '];
I have used:
var arr = str.split(",");
But one error is showed. TypeError: Object data1,data2 has no method 'split'. How can I solve this problem.
My output will be:
arr= data1,data2
// or
arr[0]=data1;
arr[1]=data2;
How can I solve this problem ?
You should do this :
var arr = str.toString().split(",");
"TypeError: Object data1,data2 has no method 'split'" indicates the variable is not considered as a string. Therefore, you must typecast it.
update 08.10.2015 I have noticed someone think the above answer is a "dirty workaround" and surprisingly this comment is upvoted. In fact it is the exact opposite - using str[0].split(",") as 3 (!) other suggests is the real "dirty workaround". Why? Consider what would happen in these cases :
var str = [];
var str = ['data1,data2','data3,data4'];
var str = [someVariable];
str[0].split(",") will fail utterly as soon str holds an empty array, for some reason not is holding a String.prototype or will give an unsatisfactory result if str holds more than one string. Using str[0].split(",") blindly trusting that str always will hold 1 string exactly and never something else is bad practice. toString() is supported by numbers, arrays, objects, booleans, dates and even functions; str[0].split() has a huge potential of raising errors and stop further execution in the scope, and by that crashing the entire application.
If you really, really want to use str[0].split() then at least do some minimal type checking :
var arr;
if (typeof str[0] == 'string') {
arr = str[0].split(',')
} else {
arr = [];
}
If your starting point is a array with a string inside. This should work:
var arr = str[0].split(",");
Otherwise you should have a string as starting point for your code to work as you expected:
var str = 'data1,data2';
If you have more elements in the array you will need to iterate them with a for loop.
Edit to add other cases:
If you have several strings in that array, then you should be more carefull and do something like this:
var str = ['data1,data2 ', ' data3, data4 ']; // notice these strings have many spaces in different places
var longString = str.join(',');
var array = longString.split(',').map(s => s.trim()).filter(Boolean); // removing spaces from start and end of strings, and eventually removing empty positions
console.log(array);
As you said, str is an array (with one element). If you want to split the string contained in the array, you have to access the array first:
var arr = str[0].split(",");
let's say we have two date :
date1: 17/01/1989
date2: 20/02/2000
if we want to compare them just split the string and compare like this
var date1= date1.toString().split("/");
var date2= date2.toString().split("/");
var a = parseInt(date1[2] + date1[1] + date1[0]);
var b = parseInt(date2[2] + date2[1] + date2[0]);
if(a < b){
alert("date2 bigger than date1")}
}
else if(a > b){
alert("date1 bigger than date2")
}
else{
alert("date 1 and date2 are equals ");
}
I am parsing some key value pairs that are separated by colons. The problem I am having is that in the value section there are colons that I want to ignore but the split function is picking them up anyway.
sample:
Name: my name
description: this string is not escaped: i hate these colons
date: a date
On the individual lines I tried this line.split(/:/, 1) but it only matched the value part of the data. Next I tried line.split(/:/, 2) but that gave me ['description', 'this string is not escaped'] and I need the whole string.
Thanks for the help!
a = line.split(/:/);
key = a.shift();
val = a.join(':');
Use the greedy operator (?) to only split the first instance.
line.split(/: (.+)?/, 2);
If you prefer an alternative to regexp consider this:
var split = line.split(':');
var key = split[0];
var val = split.slice(1).join(":");
Reference: split, slice, join.
Slightly more elegant:
a = line.match(/(.*?):(.*)/);
key = a[1];
val = a[2];
May be this approach will be the best for such purpose:
var a = line.match(/([^:\s]+)\s*:\s*(.*)/);
var key = a[1];
var val = a[2];
So, you can use tabulations in your config/data files of such structure and also not worry about spaces before or after your name-value delimiter ':'.
Or you can use primitive and fast string functions indexOf and substr to reach your goal in, I think, the fastest way (by CPU and RAM)
for ( ... line ... ) {
var delimPos = line.indexOf(':');
if (delimPos <= 0) {
continue; // Something wrong with this "line"
}
var key = line.substr(0, delimPos).trim();
var val = line.substr(delimPos + 1).trim();
// Do all you need with this key: val
}
Split string in two at first occurrence
To split a string with multiple i.e. columns : only at the first column occurrence
use Positive Lookbehind (?<=)
const a = "Description: this: is: nice";
const b = "Name: My Name";
console.log(a.split(/(?<=^[^:]*):/)); // ["Description", " this: is: nice"]
console.log(b.split(/(?<=^[^:]*):/)); // ["Name", " My Name"]
it basically consumes from Start of string ^ everything that is not a column [^:] zero or more times *. Once the positive lookbehind is done, finally matches the column :.
If you additionally want to remove one or more whitespaces following the column,
use /(?<=^[^:]*): */
Explanation on Regex101.com
function splitOnce(str, sep) {
const idx = str.indexOf(sep);
return [str.slice(0, idx), str.slice(idx+1)];
}
splitOnce("description: this string is not escaped: i hate these colons", ":")
What would be the cleanest way of doing this that would work in both IE and Firefox?
My string looks like this sometext-20202
Now the sometext and the integer after the dash can be of varying length.
Should I just use substring and index of or are there other ways?
How I would do this:
// function you can use:
function getSecondPart(str) {
return str.split('-')[1];
}
// use the function:
alert(getSecondPart("sometext-20202"));
A solution I prefer would be:
const str = 'sometext-20202';
const slug = str.split('-').pop();
Where slug would be your result
var testStr = "sometext-20202"
var splitStr = testStr.substring(testStr.indexOf('-') + 1);
var the_string = "sometext-20202";
var parts = the_string.split('-', 2);
// After calling split(), 'parts' is an array with two elements:
// parts[0] is 'sometext'
// parts[1] is '20202'
var the_text = parts[0];
var the_num = parts[1];
With built-in javascript replace() function and using of regex (/(.*)-/), you can replace the substring before the dash character with empty string (""):
"sometext-20202".replace(/(.*)-/,""); // result --> "20202"
AFAIK, both substring() and indexOf() are supported by both Mozilla and IE. However, note that substr() might not be supported on earlier versions of some browsers (esp. Netscape/Opera).
Your post indicates that you already know how to do it using substring() and indexOf(), so I'm not posting a code sample.
myString.split('-').splice(1).join('-')
I came to this question because I needed what OP was asking but more than what other answers offered (they're technically correct, but too minimal for my purposes). I've made my own solution; maybe it'll help someone else.
Let's say your string is 'Version 12.34.56'. If you use '.' to split, the other answers will tend to give you '56', when maybe what you actually want is '.34.56' (i.e. everything from the first occurrence instead of the last, but OP's specific case just so happened to only have one occurrence). Perhaps you might even want 'Version 12'.
I've also written this to handle certain failures (like if null gets passed or an empty string, etc.). In those cases, the following function will return false.
Use
splitAtSearch('Version 12.34.56', '.') // Returns ['Version 12', '.34.56']
Function
/**
* Splits string based on first result in search
* #param {string} string - String to split
* #param {string} search - Characters to split at
* #return {array|false} - Strings, split at search
* False on blank string or invalid type
*/
function splitAtSearch( string, search ) {
let isValid = string !== '' // Disallow Empty
&& typeof string === 'string' // Allow strings
|| typeof string === 'number' // Allow numbers
if (!isValid) { return false } // Failed
else { string += '' } // Ensure string type
// Search
let searchIndex = string.indexOf(search)
let isBlank = (''+search) === ''
let isFound = searchIndex !== -1
let noSplit = searchIndex === 0
let parts = []
// Remains whole
if (!isFound || noSplit || isBlank) {
parts[0] = string
}
// Requires splitting
else {
parts[0] = string.substring(0, searchIndex)
parts[1] = string.substring(searchIndex)
}
return parts
}
Examples
splitAtSearch('') // false
splitAtSearch(true) // false
splitAtSearch(false) // false
splitAtSearch(null) // false
splitAtSearch(undefined) // false
splitAtSearch(NaN) // ['NaN']
splitAtSearch('foobar', 'ba') // ['foo', 'bar']
splitAtSearch('foobar', '') // ['foobar']
splitAtSearch('foobar', 'z') // ['foobar']
splitAtSearch('foobar', 'foo') // ['foobar'] not ['', 'foobar']
splitAtSearch('blah bleh bluh', 'bl') // ['blah bleh bluh']
splitAtSearch('blah bleh bluh', 'ble') // ['blah ', 'bleh bluh']
splitAtSearch('$10.99', '.') // ['$10', '.99']
splitAtSearch(3.14159, '.') // ['3', '.14159']
For those trying to get everything after the first occurrence:
Something like "Nic K Cage" to "K Cage".
You can use slice to get everything from a certain character. In this case from the first space:
const delim = " "
const name = "Nic K Cage"
const result = name.split(delim).slice(1).join(delim) // prints: "K Cage"
Or if OP's string had two hyphens:
const text = "sometext-20202-03"
// Option 1
const opt1 = text.slice(text.indexOf('-')).slice(1) // prints: 20202-03
// Option 2
const opt2 = text.split('-').slice(1).join("-") // prints: 20202-03
Efficient, compact and works in the general case:
s='sometext-20202'
s.slice(s.lastIndexOf('-')+1)
Use a regular expression of the form: \w-\d+ where a \w represents a word and \d represents a digit. They won't work out of the box, so play around. Try this.
You can use split method for it. And if you should take string from specific pattern you can use split with req. exp.:
var string = "sometext-20202";
console.log(string.split(/-(.*)/)[1])
Everyone else has posted some perfectly reasonable answers. I took a different direction. Without using split, substring, or indexOf. Works great on i.e. and firefox. Probably works on Netscape too.
Just a loop and two ifs.
function getAfterDash(str) {
var dashed = false;
var result = "";
for (var i = 0, len = str.length; i < len; i++) {
if (dashed) {
result = result + str[i];
}
if (str[i] === '-') {
dashed = true;
}
}
return result;
};
console.log(getAfterDash("adfjkl-o812347"));
My solution is performant and handles edge cases.
The point of the above code was to procrastinate work, please don't actually use it.
To use any delimiter and get first or second part
//To divide string using deimeter - here #
//str: full string that is to be splitted
//delimeter: like '-'
//part number: 0 - for string befor delimiter , 1 - string after delimiter
getPartString(str, delimter, partNumber) {
return str.split(delimter)[partNumber];
}