page.title=Parsing XML Data parent.title=Performing Network Operations parent.link=index.html trainingnavtop=true previous.title=Managing Network Usage previous.link=managing.html @jd:body

This lesson teaches you to

Choose a Parser
Analyze the Feed
Instantiate the Parser
Read the Feed
Parse XML
Skip Tags You Don't Care About
Consume XML Data

You should also read

Web Apps Overview

Try it out

Download the sample

NetworkUsage.zip

Extensible Markup Language (XML) is a set of rules for encoding documents in machine-readable form. XML is a popular format for sharing data on the internet. Websites that frequently update their content, such as news sites or blogs, often provide an XML feed so that external programs can keep abreast of content changes. Uploading and parsing XML data is a common task for network-connected apps. This lesson explains how to parse XML documents and use their data.

Choose a Parser

We recommend {@link org.xmlpull.v1.XmlPullParser}, which is an efficient and maintainable way to parse XML on Android. Historically Android has had two implementations of this interface:

KXmlParser via {@link org.xmlpull.v1.XmlPullParserFactory#newPullParser XmlPullParserFactory.newPullParser()}.
ExpatPullParser, via {@link android.util.Xml#newPullParser Xml.newPullParser()}.

Either choice is fine. The example in this section uses ExpatPullParser, via {@link android.util.Xml#newPullParser Xml.newPullParser()}.

Analyze the Feed

The first step in parsing a feed is to decide which fields you're interested in. The parser extracts data for those fields and ignores the rest.

Here is an excerpt from the feed that's being parsed in the sample app. Each post to StackOverflow.com appears in the feed as an entry tag that contains several nested tags:

<?xml version="1.0" encoding="utf-8"?> 
<feed xmlns="http://www.w3.org/2005/Atom" xmlns:creativeCommons="http://backend.userland.com/creativeCommonsRssModule" ...">     
<title type="text">newest questions tagged android - Stack Overflow</title>
...
    <entry>
    ...
    </entry>
    <entry>
        <id>http://stackoverflow.com/q/9439999</id>
        <re:rank scheme="http://stackoverflow.com">0</re:rank>
        <title type="text">Where is my data file?</title>
        <category scheme="http://stackoverflow.com/feeds/tag?tagnames=android&sort=newest/tags" term="android"/>
        <category scheme="http://stackoverflow.com/feeds/tag?tagnames=android&sort=newest/tags" term="file"/>
        <author>
            <name>cliff2310</name>
            <uri>http://stackoverflow.com/users/1128925</uri>
        </author>
        <link rel="alternate" href="http://stackoverflow.com/questions/9439999/where-is-my-data-file" />
        <published>2012-02-25T00:30:54Z</published>
        <updated>2012-02-25T00:30:54Z</updated>
        <summary type="html">
            <p>I have an Application that requires a data file...</p>

        </summary>
    </entry>
    <entry>
    ...
    </entry>
...
</feed>

The sample app extracts data for the entry tag and its nested tags title, link, and summary.

Instantiate the Parser

The next step is to instantiate a parser and kick off the parsing process. In this snippet, a parser is initialized to not process namespaces, and to use the provided {@link java.io.InputStream} as its input. It starts the parsing process with a call to {@link org.xmlpull.v1.XmlPullParser#nextTag() nextTag()} and invokes the readFeed() method, which extracts and processes the data the app is interested in:

public class StackOverflowXmlParser {
    // We don't use namespaces
    private static final String ns = null;
   
    public List parse(InputStream in) throws XmlPullParserException, IOException {
        try {
            XmlPullParser parser = Xml.newPullParser();
            parser.setFeature(XmlPullParser.FEATURE_PROCESS_NAMESPACES, false);
            parser.setInput(in, null);
            parser.nextTag();
            return readFeed(parser);
        } finally {
            in.close();
        }
    }
 ... 
}

Read the Feed

The readFeed() method does the actual work of processing the feed. It looks for elements tagged "entry" as a starting point for recursively processing the feed. If a tag isn't an {@code entry} tag, it skips it. Once the whole feed has been recursively processed, readFeed() returns a {@link java.util.List} containing the entries (including nested data members) it extracted from the feed. This {@link java.util.List} is then returned by the parser.

private List readFeed(XmlPullParser parser) throws XmlPullParserException, IOException {
    List entries = new ArrayList();

    parser.require(XmlPullParser.START_TAG, ns, "feed");
    while (parser.next() != XmlPullParser.END_TAG) {
        if (parser.getEventType() != XmlPullParser.START_TAG) {
            continue;
        }
        String name = parser.getName();
        // Starts by looking for the entry tag
        if (name.equals("entry")) {
            entries.add(readEntry(parser));
        } else {
            skip(parser);
        }
    }  
    return entries;
}

Parse XML

The steps for parsing an XML feed are as follows:

As described in Analyze the Feed, identify the tags you want to include in your app. This example extracts data for the entry tag and its nested tags title, link, and summary.
Create the following methods:
- A "read" method for each tag you're interested in. For example, readEntry(), readTitle(), and so on. The parser reads tags from the input stream. When it encounters a tag named entry, title, link or summary, it calls the appropriate method for that tag. Otherwise, it skips the tag.
- Methods to extract data for each different type of tag and to advance the parser to the next tag. For example:
  - For the title and summary tags, the parser calls readText(). This method extracts data for these tags by calling parser.getText().
  - For the link tag, the parser extracts data for links by first determining if the link is the kind it's interested in. Then it uses parser.getAttributeValue() to extract the link's value.
  - For the entry tag, the parser calls readEntry(). This method parses the entry's nested tags and returns an Entry object with the data members title, link, and summary.
- A helper skip() method that's recursive. For more discussion of this topic, see Skip Tags You Don't Care About.

This snippet shows how the parser parses entries, titles, links, and summaries.

public static class Entry {
    public final String title;
    public final String link;
    public final String summary;

    private Entry(String title, String summary, String link) {
        this.title = title;
        this.summary = summary;
        this.link = link;
    }
}
  
// Parses the contents of an entry. If it encounters a title, summary, or link tag, hands them off
// to their respective "read" methods for processing. Otherwise, skips the tag.
private Entry readEntry(XmlPullParser parser) throws XmlPullParserException, IOException {
    parser.require(XmlPullParser.START_TAG, ns, "entry");
    String title = null;
    String summary = null;
    String link = null;
    while (parser.next() != XmlPullParser.END_TAG) {
        if (parser.getEventType() != XmlPullParser.START_TAG) {
            continue;
        }
        String name = parser.getName();
        if (name.equals("title")) {
            title = readTitle(parser);
        } else if (name.equals("summary")) {
            summary = readSummary(parser);
        } else if (name.equals("link")) {
            link = readLink(parser);
        } else {
            skip(parser);
        }
    }
    return new Entry(title, summary, link);
}

// Processes title tags in the feed.
private String readTitle(XmlPullParser parser) throws IOException, XmlPullParserException {
    parser.require(XmlPullParser.START_TAG, ns, "title");
    String title = readText(parser);
    parser.require(XmlPullParser.END_TAG, ns, "title");
    return title;
}
  
// Processes link tags in the feed.
private String readLink(XmlPullParser parser) throws IOException, XmlPullParserException {
    String link = "";
    parser.require(XmlPullParser.START_TAG, ns, "link");
    String tag = parser.getName();
    String relType = parser.getAttributeValue(null, "rel");  
    if (tag.equals("link")) {
        if (relType.equals("alternate")){
            link = parser.getAttributeValue(null, "href");
            parser.nextTag();
        } 
    }
    parser.require(XmlPullParser.END_TAG, ns, "link");
    return link;
}

// Processes summary tags in the feed.
private String readSummary(XmlPullParser parser) throws IOException, XmlPullParserException {
    parser.require(XmlPullParser.START_TAG, ns, "summary");
    String summary = readText(parser);
    parser.require(XmlPullParser.END_TAG, ns, "summary");
    return summary;
}

// For the tags title and summary, extracts their text values.
private String readText(XmlPullParser parser) throws IOException, XmlPullParserException {
    String result = "";
    if (parser.next() == XmlPullParser.TEXT) {
        result = parser.getText();
        parser.nextTag();
    }
    return result;
}
  ...
}

Skip Tags You Don't Care About

One of the steps in the XML parsing described above is for the parser to skip tags it's not interested in. Here is the parser's skip() method:

private void skip(XmlPullParser parser) throws XmlPullParserException, IOException {
    if (parser.getEventType() != XmlPullParser.START_TAG) {
        throw new IllegalStateException();
    }
    int depth = 1;
    while (depth != 0) {
        switch (parser.next()) {
        case XmlPullParser.END_TAG:
            depth--;
            break;
        case XmlPullParser.START_TAG:
            depth++;
            break;
        }
    }
 }

This is how it works:

It throws an exception if the current event isn't a START_TAG.
It consumes the START_TAG, and all events up to and including the matching END_TAG.
To make sure that it stops at the correct END_TAG and not at the first tag it encounters after the original START_TAG, it keeps track of the nesting depth.

Thus if the current element has nested elements, the value of depth won't be 0 until the parser has consumed all events between the original START_TAG and its matching END_TAG. For example, consider how the parser skips the <author> element, which has 2 nested elements, <name> and <uri>:

The first time through the while loop, the next tag the parser encounters after <author> is the START_TAG for <name>. The value for depth is incremented to 2.
The second time through the while loop, the next tag the parser encounters is the END_TAG </name>. The value for depth is decremented to 1.
The third time through the while loop, the next tag the parser encounters is the START_TAG <uri>. The value for depth is incremented to 2.
The fourth time through the while loop, the next tag the parser encounters is the END_TAG </uri>. The value for depth is decremented to 1.
The fifth time and final time through the while loop, the next tag the parser encounters is the END_TAG </author>. The value for depth is decremented to 0, indicating that the <author> element has been successfully skipped.

Consume XML Data

The example application fetches and parses the XML feed within an {@link android.os.AsyncTask}. This takes the processing off the main UI thread. When processing is complete, the app updates the UI in the main activity (NetworkActivity).

In the excerpt shown below, the loadPage() method does the following:

Initializes a string variable with the URL for the XML feed.
If the user's settings and the network connection allow it, invokes new DownloadXmlTask().execute(url). This instantiates a new DownloadXmlTask object ({@link android.os.AsyncTask} subclass) and runs its {@link android.os.AsyncTask#execute execute()} method, which downloads and parses the feed and returns a string result to be displayed in the UI.

public class NetworkActivity extends Activity {
    public static final String WIFI = "Wi-Fi";
    public static final String ANY = "Any";
    private static final String URL = "http://stackoverflow.com/feeds/tag?tagnames=android&sort=newest";
   
    // Whether there is a Wi-Fi connection.
    private static boolean wifiConnected = false; 
    // Whether there is a mobile connection.
    private static boolean mobileConnected = false;
    // Whether the display should be refreshed.
    public static boolean refreshDisplay = true; 
    public static String sPref = null;

    ...
      
    // Uses AsyncTask to download the XML feed from stackoverflow.com.
    public void loadPage() {  
      
        if((sPref.equals(ANY)) && (wifiConnected || mobileConnected)) {
            new DownloadXmlTask().execute(URL);
        }
        else if ((sPref.equals(WIFI)) && (wifiConnected)) {
            new DownloadXmlTask().execute(URL);
        } else {
            // show error
        }  
    }

The {@link android.os.AsyncTask} subclass shown below, DownloadXmlTask, implements the following {@link android.os.AsyncTask} methods:

{@link android.os.AsyncTask#doInBackground doInBackground()} executes the method loadXmlFromNetwork(). It passes the feed URL as a parameter. The method loadXmlFromNetwork() fetches and processes the feed. When it finishes, it passes back a result string.
{@link android.os.AsyncTask#onPostExecute onPostExecute()} takes the returned string and displays it in the UI.

// Implementation of AsyncTask used to download XML feed from stackoverflow.com.
private class DownloadXmlTask extends AsyncTask<String, Void, String> {
    @Override
    protected String doInBackground(String... urls) {
        try {
            return loadXmlFromNetwork(urls[0]);
        } catch (IOException e) {
            return getResources().getString(R.string.connection_error);
        } catch (XmlPullParserException e) {
            return getResources().getString(R.string.xml_error);
        }
    }

    @Override
    protected void onPostExecute(String result) {  
        setContentView(R.layout.main);
        // Displays the HTML string in the UI via a WebView
        WebView myWebView = (WebView) findViewById(R.id.webview);
        myWebView.loadData(result, "text/html", null);
    }
}

Below is the method loadXmlFromNetwork() that is invoked from DownloadXmlTask. It does the following:

Instantiates a StackOverflowXmlParser. It also creates variables for a {@link java.util.List} of Entry objects (entries), and title, url, and summary, to hold the values extracted from the XML feed for those fields.
Calls downloadUrl(), which fetches the feed and returns it as an {@link java.io.InputStream}.
Uses StackOverflowXmlParser to parse the {@link java.io.InputStream}. StackOverflowXmlParser populates a {@link java.util.List} of entries with data from the feed.
Processes the entries {@link java.util.List}, and combines the feed data with HTML markup.
Returns an HTML string that is displayed in the main activity UI by the {@link android.os.AsyncTask} method {@link android.os.AsyncTask#onPostExecute onPostExecute()}.

// Uploads XML from stackoverflow.com, parses it, and combines it with
// HTML markup. Returns HTML string.
private String loadXmlFromNetwork(String urlString) throws XmlPullParserException, IOException {
    InputStream stream = null;
    // Instantiate the parser
    StackOverflowXmlParser stackOverflowXmlParser = new StackOverflowXmlParser();
    List<Entry> entries = null;
    String title = null;
    String url = null;
    String summary = null;
    Calendar rightNow = Calendar.getInstance(); 
    DateFormat formatter = new SimpleDateFormat("MMM dd h:mmaa");
        
    // Checks whether the user set the preference to include summary text
    SharedPreferences sharedPrefs = PreferenceManager.getDefaultSharedPreferences(this);
    boolean pref = sharedPrefs.getBoolean("summaryPref", false);
        
    StringBuilder htmlString = new StringBuilder();
    htmlString.append("<h3>" + getResources().getString(R.string.page_title) + "</h3>");
    htmlString.append("<em>" + getResources().getString(R.string.updated) + " " + 
            formatter.format(rightNow.getTime()) + "</em>");
        
    try {
        stream = downloadUrl(urlString);        
        entries = stackOverflowXmlParser.parse(stream);
    // Makes sure that the InputStream is closed after the app is
    // finished using it.
    } finally {
        if (stream != null) {
            stream.close();
        } 
     }
    
    // StackOverflowXmlParser returns a List (called "entries") of Entry objects.
    // Each Entry object represents a single post in the XML feed.
    // This section processes the entries list to combine each entry with HTML markup.
    // Each entry is displayed in the UI as a link that optionally includes
    // a text summary.
    for (Entry entry : entries) {       
        htmlString.append("<p><a href='");
        htmlString.append(entry.link);
        htmlString.append("'>" + entry.title + "</a></p>");
        // If the user set the preference to include summary text,
        // adds it to the display.
        if (pref) {
            htmlString.append(entry.summary);
        }
    }
    return htmlString.toString();
}

// Given a string representation of a URL, sets up a connection and gets
// an input stream.
private InputStream downloadUrl(String urlString) throws IOException {
    URL url = new URL(urlString);
    HttpURLConnection conn = (HttpURLConnection) url.openConnection();
    conn.setReadTimeout(10000 /* milliseconds */);
    conn.setConnectTimeout(15000 /* milliseconds */);
    conn.setRequestMethod("GET");
    conn.setDoInput(true);
    // Starts the query
    conn.connect();
    return conn.getInputStream();
}