Originally by: Xeross155

---

If he needs just 1 specific value, and he needs to get it from a lot of xml strings regex might be faster.

---

Originally by: darius mclever

---

Originally by: Xeross155

---

If he needs just 1 specific value, and he needs to get it from a lot of xml strings regex might be faster.

---

with all the cleaning up (removing whitespaces/newlines) of the xml to get some predictable format?
I highly doubt that you are faster than using a native xml parser and putting the small xpath on it.

---

Originally by: Zeke Mobius

---

I swear the catholic church was faster at admitting the earth was round than CCP at fixing stuff.

---

Originally by: Miss Direction

---

Hi Fovo,

A regex that would select this value for you using a backreference would be

<avg\b[^>]*>(.*?)</avg>

It would carry over any whitespace however if it was present.

---

<avg\b[^>]*>\s*(\d+(?:\.\d+))?\s*<\/avg>

Originally by: Tonto Auri

---

What language you are using?
It would be easier to provide complete solution rather than trying to pin the parts, especially when you are not very confident in your skills.

The general answer would be to match regexp against all entries rather than stop at first one.
Then you would get all 3 results at once.
Also, referring to the "any spaces" (although EVEc API is rather strict in this part), you can actually defer their appearance in the result using exact match.

<avg\b[^>]*>\s*(\d+(?:\.\d+))?\s*<\/avg>

However, I should warn you that in case of empty <avg/> tag your regexp will fail and you will get unreliable results.
That's why it is generally advisable to use XML parsers for XML data.
You could also request up to 25(?) type ids in single transaction, instead of hammering EVEc servers with multiple requests.

---

Originally by: Tonto Auri

---

<avg\b[^>]*>\s*(\d+(?:\.\d+))?\s*<\/avg>

---

Originally by: Fovo

---

It is a web page that uses a java back end. The sum of it is I point it to the page, I tell it the url to go fetch and I provide it the regular expression to parse the page content with. The back end then saves the result (if there is one) to a database. Using a 3rd party parser or "Re-writing server code" isn't really an option.
I either find a regular expression to give the server so it can do the work or I drop teh entire thing and call it a fail.

---

import java.net.URL;
import javax.xml.parsers.DocumentBuilderFactory;
import javax.xml.xpath.*;
import org.w3c.dom.Document;


URL url = new URL("http://api.eve-central.com/api/marketstat?typeid=34®ionlimit=10000002");
Document document = DocumentBuilderFactory.newInstance().newDocumentBuilder().parse(url.openStream());
XPath xpath = XPathFactory.newInstance().newXPath();

XPathExpression expr = xpath.compile("//marketstat/type/sell/avg/text()");
Object result = expr.evaluate(document);

Originally by: Lutz Major

---

Do I get this right? You fetch and parse the file in Java?

import java.net.URL;
import javax.xml.xpath.*;
import org.xml.sax.InputSource;


URL url = new URL("http://api.eve-central.com/api/marketstat?typeid=34®ionlimit=10000002");
XPathExpression expr = XPathFactory.newInstance().newXPath().compile("//marketstat/type/sell/avg/text()");
String result = expr.evaluate(new InputSource(url.openStream()));

If you want it even easier use an unmarshaller and annotations.
Edit: made it even simplier

---

.*<sell.*<avg>(\d*\.\d+)</.*

<sell[\s\S]+?<avg>(.+?)<

Originally by: Zeke Mobius

---

I swear the catholic church was faster at admitting the earth was round than CCP at fixing stuff.

---