NSRegularExpression not getting exact text - iphone

I have a string like:
<book>MyBook</book><value>myValue</value>
Now I want to get the text "myValue" out of this string. I want to use NSRegularExpression to do this. I tried this:
NSRegularExpression *regex = [NSRegularExpression regularExpressionWithPattern:#"(<book>MyBook</book>\\s*<value>).*?(</value>)"
options:NSRegularExpressionCaseInsensitive
error:&error];
NSArray *textArray = [regex matchesInString:myData options:0 range:NSMakeRange(0, [myData length])];
NSTextCheckingResult * result = [rege firstMatchInString:myData
options:0
range:NSMakeRange(0, [myData length])];
The result is:
<book>MyBook</book><value>myValue</value>
So I get the whole string, but I only want "myValue". How can I do this? What am I missing here?
Thanks in advance!

That happens because you wrote a regex that matches the entire string. I'd reckon that writing a regex that will only match the myValue part of the string is way too complicated to be bothered with (due to the fact that you've got MyBook string that will probably match anything myValue does).
I'd recommend not using regex for this, as they are not intended for the use you've described here. If you don't want to use any XML deserialization, you could use a NSScanner or any of the NSString class methods which will yield a simpler, and easier code to maintain.
For example, using an NSScanner and a few other methods:
NSString *stringToBeScanned = #"<book>MyBook</book><value>myValue</value>";
NSString *myValue;
NSScanner *scanner = [NSScanner scannerWithString:stringToBeScanned];
[scanner scanUpToString:#"<value>" intoString:nil];
// After the above, we've got "<value>myValue</value>" left to scan
[scanner scanUpToString:#"</value>" intoString:&myValue];
// We ended up with a "<value>myValue" type of a string
// This will trim the remaining of the string we don't need
myValue = [myValue stringByReplacingOccurrencesOfString:#"<value>" withString:#""];
The above could probably be written better and I might have made a mistake or two writing it out my head, but the principle should work.

Related

How can I find a dynamic number in a long NSString?

I have a very big NSString, which holds around 1500 characters in it. In this string I need to extract a phone number, which may change frequently, as it is a dynamic data. The phone number will be in the format of 251-221-2000, how can I extract this?
Check out this previous question on regular expressions and NSString.
Search through NSString using Regular Expression
In your case an appropriate regular expression would be #"\\d{3}-\\d{3}-\\d{4}".
This sounds like a perfect candidate for a regular expression. You can use the NSRegularExpression class to achieve this. You can test your regular expression at http://www.regextester.com
NSString *yourString = #"Your 1500 characters string ";
NSError *error = NULL;
NSRegularExpression *regex = [NSRegularExpression
regularExpressionWithPattern:#"\d{3}-\d{3}-\d{4}"
options:NSRegularExpressionCaseInsensitive
error:&error];
[regex enumerateMatchesInString:yourString options:0 range:NSMakeRange(0, [yourString length]) usingBlock:^(NSTextCheckingResult *match, NSMatchingFlags flags, BOOL *stop){
// your code to handle matches here
}];
Let me know it is working or not.

Parse NSString from right hand side?

> (2009 RX7)</font></td>
>monospace" size="-1">214869 (2007 PAZ)</font></td>
>monospace" size="-1"> 4155 Accord</font></td>
I wonder if someone could offer me a little help, I have a list of NSString items (See Above) that I want to parse some data from. My problem is that there are no tags that I can use within the strings nor do the items I want have fixed positions. The data I want to extract is:
2009 RX7
2007 PAZ
4155 Accord
My thinking is that its going to be easier to parse from the right hand end, remove the </font></td> and then use ";" to separate the data items:
(2009&nbsp RX7)
(2007&nbsp PAZ)
4155&nbsp Accord
which can them be cleaned up to match the example given. Any pointers on doing this or working through from the right would be very much appreciated.
Personally I think you are better off with a regex. So my solution would be:
Regex of: ([0-9]+)[^;]+;([A-Za-z0-9]+)
Which for all the example text provides 3 matches. ie for:
(2009 RX7)</font></td>
0: 2009 RX7)<
1: 2009
2: RX7
I haven't coded this up, but did test the Regex at www.regextester.com
Regex's are implemented via NSRegularExpression and are available in iOS 4.0 and later.
Edit
Given that this appears to be a web scraping application, you never know when those pesky HTML code monkeys will change their output and break your carefully crafted matching methodology. As such I would change my regex to:
([0-9]+)([^;]+;)+([A-Za-z0-9]+)
Which adds an extra group, but allows for any number of elements between the number and the string.
Try this code:
NSString *str = #"> (2009 RX7)</font></td>";
NSRange fontRange = [str rangeOfString:#"</Font>" options:NSBackwardsSearch];
NSRange lastSemi = [str rangeOfString:#";" options:NSBackwardsSearch range:NSMakeRange(0, fontRange.location-1)];
NSRange priorSemi = [str rangeOfString:#";" options:NSBackwardsSearch range:NSMakeRange(0, lastSemi.location-1)];
NSString *yourString = [str substringWithRange:NSMakeRange(priorSemi.location+1, fontRange.location-1)];
The key element here is the NSBackwardsSearch search option.
This should do the trick:
NSString *s = #">monospace\" size=\"-1\"> 4155 Accord</font></td>";
NSArray *strArray = [s componentsSeparatedByString:#";"];
// you're interested in last two objects
NSArray *tmp = [strArray subarrayWithRange:NSMakeRange(strArray.count - 2, 2)];
In tmp you'll have something like:
"4155&nbsp",
"Accord</font></td>"
strip unneeded chars and you're all set.
Using NSRegularExpression:
NSRegularExpression *regex;
NSTextCheckingResult *match;
NSString *pattern = #"([0-9]+) ([A-Za-z0-9]+)[)]?</font></td>";
NSString *string = #"> (2009 RX7)</font></td>";
regex = [NSRegularExpression
regularExpressionWithPattern:pattern
options:NSRegularExpressionCaseInsensitive
error:nil];
match = [regex firstMatchInString:string options:0 range:NSMakeRange(0, [string length])];
NSLog(#"'%#'", [string substringWithRange:[match rangeAtIndex:1]]);
NSLog(#"'%#'", [string substringWithRange:[match rangeAtIndex:2]]);
NSLog output:
'2009'
'RX7'

How to check if a string contains English letters (A-Z)?

How can I check whether a string contains the English Letters (A through Z) in Objective-C?
In PHP, there is preg_match method for that.
One approach would be to use regular expressions — the NSRegularExpression class. The following demonstrates how you could detect any English letters, but the pattern could be modified to match only if the entire string consists of such letters. Something like ^[a-zA-Z]*$.
NSRegularExpression *regex = [[[NSRegularExpression alloc]
initWithPattern:#"[a-zA-Z]" options:0 error:NULL] autorelease];
// Assuming you have some NSString `myString`.
NSUInteger matches = [regex numberOfMatchesInString:myString options:0
range:NSMakeRange(0, [myString length])];
if (matches > 0) {
// `myString` contains at least one English letter.
}
Alternatively, you could construct an NSCharacterSet containing the characters you're interested in and use NSString's rangeOfCharacterFromSet: to find the first occurrence of any one. I should note that this method only finds the first such character in the string. Maybe not what you're after.
Finally, I feel like you could do something with encodings, but haven't given this much thought. Perhaps determine if the string could be represented using ASCII (using canBeConvertedToEncoding:) and then check for numbers/symbols?
Oh, and you could always iterate over the string and check each character! :)
You can use simple NSPredicate test.
NSString *str = #"APPLE";
NSString *regex = #"[A-Z]+";
NSPredicate *test = [NSPredicate predicateWithFormat:#"SELF MATCHES %#", regex];
BOOL result = [test evaluateWithObject:str];
You could also use NSCharacterSet and use the rangeOfCharacterFromSet: method to see if the returned NSRange is the entire range of the string. characterSetWithRange would be a good place to start to create your characterSet.
You can use the NSRegularExpression Class (Apple's documentation on the class can be viewed here)

what's the best way to detect if a word in NSString has a number?

example: word with number in string
NSString *str = [NSString stringWithFormat:#"this is an 101 example1 string"]
Since example1 has a number in the end and i want to remove it. I can break it into an array and filter it out using predicate, but that seems slow to me since I need to do like a million of these.
What would be a more efficient way?
Thanks!
Probably NSRegularExpression. I think ([^0-9 ]+)\d+|\d+([^0-9 ]+) should do it. Just replace it with $1.
Based on Chuck's response, here is the complete code in case someone might find it useful:
NSError *error = NULL;
NSRegularExpression *regex = [NSRegularExpression regularExpressionWithPattern:#"([^0-9 ]+)\\d+|\\d+([^0-9 ]+)"
options:NSRegularExpressionCaseInsensitive
error:&error];
NSString *modifiedString = [regex stringByReplacingMatchesInString:str2
options:0
range:NSMakeRange(0, [str2 length])
withTemplate:#"$1"];

URL-encoding and HTML-encoding NSStrings

Is their a method to encode/decode HTML and URL (in Xcode, using Objective-C)?
[NSString stringWithContentsOfFile:<#(NSString *)path#> encoding:<#(NSStringEncoding)enc#> error:<#(NSError **)error#>]
This doesn't seem to work how i expected. I thought it will convert special characters like "<" to equivalent HTML entities i.e. "<" in this case.
Here's a reference to the w3school link related to this topic (general):
HTML URL Encoding Reference
HTML Entities Reference
Thanking in anticipation.
Returns a representation of the receiver using a given encoding to determine the percent escapes necessary to convert the receiver into a legal URL string.
- (NSString *)stringByAddingPercentEscapesUsingEncoding:(NSStringEncoding)encoding
and
Returns a new string made by replacing in the receiver all percent escapes with the matching characters as determined by a given encoding.
- (NSString *)stringByReplacingPercentEscapesUsingEncoding:(NSStringEncoding)encoding
The method you cite reads a file from disk with a given character encoding (such as UTF-8 or ASCII). It has nothing to do with URL or HTML escaping.
If you want to add URL percent escapes, you want this method:
[myString stringByAddingPercentEscapesUsingEncoding:NSUTF8StringEncoding]
Make sure you read the documentation about this method, because there are certain subtleties about what it escapes and what it leaves alone. In some cases, you may have to use the more complex, but more flexible, CFURLCreateStringByAddingPercentEscapes(). (If you do, note that you can cast CFStringRef to NSString * and vice versa.)
There's nothing built in that I know of to do XML/HTML-style entity escaping, but this function ought to handle the basics:
NSString * convertToXMLEntities(NSString * myString) {
NSMutableString * temp = [myString mutableCopy];
[temp replaceOccurrencesOfString:#"&"
withString:#"&"
options:0
range:NSMakeRange(0, [temp length])];
[temp replaceOccurrencesOfString:#"<"
withString:#"<"
options:0
range:NSMakeRange(0, [temp length])];
[temp replaceOccurrencesOfString:#">"
withString:#">"
options:0
range:NSMakeRange(0, [temp length])];
[temp replaceOccurrencesOfString:#"\""
withString:#"""
options:0
range:NSMakeRange(0, [temp length])];
[temp replaceOccurrencesOfString:#"'"
withString:#"&apos;"
options:0
range:NSMakeRange(0, [temp length])];
return [temp autorelease];
}
To do HTML/XML entity encoding, you can use a CFMutableString function:
NSString *result = .....;
CFStringTransform((CFMutableStringRef)result, NULL, kCFStringTransformToXMLHex, false);
By setting the last parameter of CFStringTransform to true, it should work for decoding (hex) entities as well.
Use CFStringTransform for HTML entity encoding/decoding:
CFStringTransform((CFTypeRef)yourMutableString, NULL, CFSTR("Any-Hex/XML"), FALSE );
You need to use the ICU transform "Any-Hex/XML". kCFStringTransformToXMLHex isn't aggressive enough.