Go back to parent node using XmlReader

Question

I'm looking at the example below, and I wonder, after I'm done reading all the image nodes, how do I get back reading the last_modified_at node from parent element? (See my comments in last few lines of code)

public class StackOverflow_6473251
{
    public static void Test()
    {
        string xml = @"
           <movies>
             <movie>
               <score>8.582207</score>
               <popularity>3</popularity>
               <translated>true</translated>
               <adult>false</adult>
               <language>en</language>
               <original_name>Transformers</original_name>
               <name>Transformers</name>
               <alternative_name>The Transformers</alternative_name>
               <type>movie</type>
               <id>1858</id>
               <imdb_id>tt0418279</imdb_id>
               <url>http://www.themoviedb.org/movie/1858</url>
               <votes>28</votes>
               <rating>7.2</rating>
               <certification>PG-13</certification>
               <overview>The Earth is caught in the middle of an intergalactic war /overview>
               <released>2007-07-04</released>
               <images>
                    <image type=""poster"" url=""http://hwcdn.themoviedb.org/posters/304/4bc91347017a3c57fe007304/transformers-original.jpg"" size=""original"" id=""4bc91347017a3c57fe007304""/>
                    <image type=""poster"" url=""http://hwcdn.themoviedb.org/posters/304/4bc91347017a3c57fe007304/transformers-mid.jpg"" size=""mid"" id=""4bc91347017a3c57fe007304""/>
                    <image type=""poster"" url=""http://hwcdn.themoviedb.org/posters/304/4bc91347017a3c57fe007304/transformers-cover.jpg"" size=""cover"" id=""4bc91347017a3c57fe007304""/>
                    <image type=""poster"" url=""http://hwcdn.themoviedb.org/posters/304/4bc91347017a3c57fe007304/transformers-thumb.jpg"" size=""thumb"" id=""4bc91347017a3c57fe007304""/>
                    <image type=""backdrop"" url=""http://hwcdn.themoviedb.org/backdrops/2ce/4bc91339017a3c57fe0072ce/transformers-original.jpg"" size=""original"" id=""4bc9133s9017a3c57fe0072ce""/>
                    <image type=""backdrop"" url=""http://hwcdn.themoviedb.org/backdrops/2ce/4bc91339017a3c57fe0072ce/transformers-poster.jpg"" size=""poster"" id=""4bc91339017a3c57fe0072ce""/>
                    <image type=""backdrop"" url=""http://hwcdn.themoviedb.org/backdrops/2ce/4bc91339017a3c57fe0072ce/transformers-thumb.jpg"" size=""thumb"" id=""4bc91339017a3c57fe0072ce""/>
               </images>
             <last_modified_at>2010-04-26 03:26:14</last_modified_at>
       </movie>
     </movies>";
        XmlReader r = XmlReader.Create(new StringReader(xml));
        r.ReadToFollowing("original_name");
        string title = r.ReadElementContentAsString("original_name", r.NamespaceURI);
        r.ReadToFollowing("images");
        int imageCount = 0;
        if (r.ReadToDescendant("image"))
        {
            do
            {
                Console.WriteLine("Image {0}", ++imageCount);
                Console.WriteLine("  Type: {0}", r.GetAttribute("type"));
                Console.WriteLine("  URL: {0}", r.GetAttribute("url"));
                Console.WriteLine("  Size: {0}", r.GetAttribute("size"));
                Console.WriteLine("  ID: {0}", r.GetAttribute("id"));
            } while (r.ReadToNextSibling("image"));
        }
        //read last_modified_at
        //reader.Name evaluates to 'images' at execution time
        //reader.NodeType evaluates to 'EndElement' at execution time
    }
}

UPDATE

In a similar scenario, I'm trying to achieve the same result successfully accomplished by the answers below unsuccessfully, and fails at the highlighted line (41) by returning false.
The problem is going from the last abbr to intervals. And here is where I can't see the difference from the example presented above.

Update Here is a complete example.

Why aren't you just deserializing the xml? – Aron Oct 10 '14 at 11:04 — Aron, Oct 10 '14 at 11:04
@Aron what do you mean? – Shimmy Weitzhandler Oct 10 '14 at 11:24 — Shimmy Weitzhandler, Oct 10 '14 at 11:24
The question drastically changed and should be closed. – Shimmy Weitzhandler Oct 10 '14 at 12:29 — Shimmy Weitzhandler, Oct 10 '14 at 12:29

score 1 · Answer 1 · answered Oct 10 '14 at 04:30

1

You can't. From the documentation:

"Represents a reader that provides fast, noncached, forward-only access to XML data."

You could read from the start again.

Or, since the data is very small, you could use an XmlDocument which allows greater flexibility.

answered Oct 10 '14 at 04:30

Steve Wellens

20,506
2
28
69

According to what you say, an XML reader can only read first element, which doesn't seem right. Please see my updated question with another example. – Shimmy Weitzhandler Oct 10 '14 at 10:26
@Shimmy I didn't say that. – Steve Wellens Oct 10 '14 at 15:29
You said 'you can't'. According to the documentation there has to be a way, I was talking about next upcoming element, the reason I wanted to go up to parent is for the sake of accessing next node. Thanks anyway. – Shimmy Weitzhandler Oct 11 '14 at 16:41

Fratyx · Accepted Answer · 2014-10-10T05:21:07.390

1

As last_modified_at follows the images you can do it as you did it some lines up:

r.ReadToFollowing("last_modified_at");
string lastModified = r.ReadElementContentAsString("last_modified_at", r.NamespaceURI);

edited Oct 10 '14 at 05:21

answered Oct 10 '14 at 04:33

Fratyx

5,717
1
12
22

[Here](http://pastebin.com/ZetsSmmL)'s code that have same example, but fails! on the highlighted line, `reader.Name` evaluates to `ChordType` and `reader.NodeType` is `EndElement`. Why is that? – Shimmy Weitzhandler Oct 10 '14 at 04:58
@Shimmy Why are you using pastebin? that code should have been your question. – weston Oct 10 '14 at 05:09
1

As I can see it the code should crash at the line `if (reader.Name != "abbreviations")` already because you do not read the `abbreviations` tag before. – Fratyx Oct 10 '14 at 05:18
@Shimmy I've tested this answer it absolutely works and they're absolutely correct about where the code crashes. Make a Short Self Contained Correct Example. – weston Oct 10 '14 at 05:30
@Fratyx, I don't understand, why should I be reading the `abbreviations` before? – Shimmy Weitzhandler Oct 10 '14 at 05:58
@weston, you're right. I initially started the question by posting the similar example. Can you notice the difference between the code in my question (which works according to Fratyx's answer), and the code in pastebin that fails? – Shimmy Weitzhandler Oct 10 '14 at 06:00
@Shimmy At line 23 of your code the reader "stands before" the `abbreviations` tag. First the reader has to read this element (`reader.Read()`) then the `Name` of the reader will be "abbreviations". – Fratyx Oct 10 '14 at 06:08
-1 The `Reader` pattern is a forward reading pattern. There is no guarantee that the underlying data structure can be read backwards (NetworkStream for instance). – Aron Oct 10 '14 at 06:12
@Aron, the structure and order of the elements in the XML file of the pastebin is guaranteed to be the same. After the `abbreviations` element, I need the reader to move to the `intervals`. @Fratyx, the problem is at line 40, the `abbreviations` part is read just fine. – Shimmy Weitzhandler Oct 10 '14 at 10:13
I'm sorry but after fixing 'my' problem in line 23 I don't get a problem in line 40. – Fratyx Oct 10 '14 at 10:22
Ok. [Here](http://pastebin.com/Un0B9i1P)'s a complete reproduction. I've tried replacing the condition on line 59 with: `!reader.Read()`, `reader.Name != "intervals"`, and `!reader.ReadToFollowing("intervals")`, they all evaluate to false. – Shimmy Weitzhandler Oct 10 '14 at 11:56
1

Replace lines 54-56 by `while(reader.Name == "abbr") Trace.WriteLine(string.Format("Abbreviation {0} - Category: {1}, Abbr: {2}", ++i, reader.GetAttribute("category"), reader.ReadElementContentAsString()));` ReadElementContentAsString() moves to the next sibling and ReadToNextSibling does it again so you skip each second entry by your solution. – Fratyx Oct 10 '14 at 12:23
Yep. A few tweaks done but this was the catch. Thanks for you effort and patience @Fratyx. – Shimmy Weitzhandler Oct 10 '14 at 12:45
You're welcome. Just as advice: Use XmlDocument or even better XDocument for reading XML. XmlReader maybe suitable for very large XML documents but the alternatives are much easier to handle. – Fratyx Oct 10 '14 at 12:54
I'm using the `XmlReader`, because it saves performance as it's *forward-only*, as [Steve mentions](http://stackoverflow.com/a/26291962/75500). I'll have to iterate on a large amount of nodes at program startup and choose to optimize wherever I can. `XDocument`, although way more flexible and convenient, iterates over from the beginning of the node for each query. – Shimmy Weitzhandler Oct 11 '14 at 16:47

score 0 · Answer 3 · answered Oct 10 '14 at 06:08

0

In your pastebin example, instead of if (reader.Name != "abbreviations") do:

if (!reader.ReadToFollowing("abbreviations"))

It fails because you are not on abbreviations tag yet.

The corresponding working code in the question is:

r.ReadToFollowing("images");

In fact there are no reader.Name != tests.

answered Oct 10 '14 at 06:08

weston

54,145
21
145
203

But its the if statement at line 39 that evaluates to false, not the `abbreviations`. It's the `intervals`. Please read update in my question. – Shimmy Weitzhandler Oct 10 '14 at 10:24
1

For gods sake man, put the code in your question. This is noob stack overflow stuff. You've got 24.4K how? – weston Oct 10 '14 at 12:20
I'd delete my question and start it anew. Please vote for its closure. – Shimmy Weitzhandler Oct 10 '14 at 12:28

Go back to parent node using XmlReader

UPDATE

3 Answers3