introduce pagination for all paged api calls and introduce cache invalidation

This commit is contained in:
t865095
2014-07-08 13:52:53 +02:00
parent 7e35716ff0
commit b9764c0a44
2 changed files with 34 additions and 2 deletions

View File

@@ -32,6 +32,7 @@ import java.io.InputStreamReader;
import java.io.InterruptedIOException;
import java.io.Reader;
import java.io.UnsupportedEncodingException;
import java.lang.reflect.Array;
import java.lang.reflect.Field;
import java.net.HttpURLConnection;
import java.net.MalformedURLException;
@@ -47,10 +48,14 @@ import java.util.List;
import java.util.Map;
import java.util.NoSuchElementException;
import java.util.Set;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
import java.util.zip.GZIPInputStream;
import org.apache.commons.io.IOUtils;
import javax.net.ssl.HttpsURLConnection;
/**
* A builder pattern for making HTTP call and parsing its output.
*
@@ -177,7 +182,26 @@ class Requester {
buildRequest(uc);
try {
return parse(uc,type,instance);
T result = parse(uc, type, instance);
if (type != null && type.isArray()) { // we might have to loop for pagination - done through recursion
final String links = uc.getHeaderField("link");
if (links != null && links.contains("rel=\"next\"")) {
Pattern nextLinkPattern = Pattern.compile(".*<(.*)>; rel=\"next\"");
Matcher nextLinkMatcher = nextLinkPattern.matcher(links);
if (nextLinkMatcher.find()) {
final String link = nextLinkMatcher.group(1);
T nextResult = _to(link, type, instance);
final int resultLength = Array.getLength(result);
final int nextResultLength = Array.getLength(nextResult);
T concatResult = (T) Array.newInstance(type.getComponentType(), resultLength + nextResultLength);
System.arraycopy(result, 0, concatResult, 0, resultLength);
System.arraycopy(nextResult, 0, concatResult, resultLength, nextResultLength);
result = concatResult;
}
}
}
return result;
} catch (IOException e) {
handleApiError(e,uc);
}