Select a specific range of elements?

Question

I want to select a specific range of elements within my XML files

Here is an example of the XML:

<urlset>
    <url>
        <loc>e1</loc>
        <priority>1</priority>
    </url>
    <url>
        <loc>e2</loc>
        <priority>2</priority>
    </url>
    <url>
        <loc>e3</loc>
        <priority>1</priority>
    </url>
    <url>
        <loc>e4</loc>
        <priority>3</priority>
    </url>
    <url>
        <loc>e4</loc>
        <priority>1</priority>
    </url>
    <url>
        <loc>e5</loc>
        <priority>2</priority>
    </url>
</urlset>

How to get the value of e2 to e4?

By value, do you mean the priority? – Pierre-Alain Vigeant Aug 15 '11 at 14:11 — Pierre-Alain Vigeant, Aug 15 '11 at 14:11
Lookup XPath and conditions in []-brackets ... – Hinek Aug 15 '11 at 14:13 — Hinek, Aug 15 '11 at 14:13

Konrad Morawski · Answer 1 · 2011-08-15T14:35:15.947

2

var result = urlset.Elements("url").Where(url =>
            url.Element("loc").Value.ToString().CompareTo("e2") >= 0 &&
            url.Element("loc").Value.ToString().CompareTo("e4") <= 0).Select(element => element.Element("loc").Value.ToString());

It uses standard (string) comparison - same as in alphabetical sorting, and it doesn't protect you against cases where some element would not have loc subelement at all (null reference exception would be thrown).

edited Aug 15 '11 at 14:35

answered Aug 15 '11 at 14:14

Konrad Morawski

8,307
7
53
91

I should have used `CompareTo` (>= does not work for strings) - I updated this. As for your comment: I just tested it and it works as expected, whether the xml elements are sorted or not. The example given above is inclusive of "e4", if that's not intended just use `("e4") < 0` instead of `<=`. It's admittedly less clean than `SkipWhile` / `TakeWhile`, though. – Konrad Morawski Aug 15 '11 at 14:37
Chop off `Select` if you want to retrieve entire XML elements rather than just the `loc` values by themselves. – Konrad Morawski Aug 15 '11 at 14:39
Change `e3` to `e10`. It won't select `e10`. As I understand, OP needs elements between `e2` and `e4`. – Kirill Polishchuk Aug 15 '11 at 14:44
Do we want it to select `e10`? `e10` is not between `e2` and `e4`. A problem would occur if you had `e30` (treated as string, it does fall between `e2` and `e4`). But this problem occurs whether you use `Where` or `SkipWhile` / `TakeWhile`, no difference here. A small method converting `"e30"` to `(int)30`, `"e4"` to `(int)4` etc. would be needed. – Konrad Morawski Aug 15 '11 at 14:49
Maybe I misinterpret OP question. As I understand, for example, in sequence `1, 2, 10, 3, 4, 5` OP needs elements between `2` and `4`. In this sample: `2, 10, 3`. – Kirill Polishchuk Aug 15 '11 at 14:52
Ah OK I see your point now. I understood it differently. He should probably just clarify this. – Konrad Morawski Aug 15 '11 at 14:53

score 2 · Accepted Answer · answered Aug 15 '11 at 14:14

2

var doc = XDocument.Parse(xml);

var result = doc.Element("urlset").Elements("url")
    .SkipWhile(x => x.Element("loc").Value != "e2")
    .TakeWhile(x => x.Element("loc").Value != "e4");

answered Aug 15 '11 at 14:14

Kirill Polishchuk

54,804
11
122
125

Wouldn't it just return all the `e2`s and nothing else? I did not downvote by the way, I just undid my upvote (which was for the use of `SkipWhile` and `TakeWhile` - I didn't know these). The OP wants a a range of elements e2 from to e4 (presumably including e3, for example) – Konrad Morawski Aug 15 '11 at 14:18
@Morawski, It will return `e2`, `e3` in provided sample, in other words from `e2` to `e4` – Kirill Polishchuk Aug 15 '11 at 14:21
I'd consider using OrderBy, same for Morawski's answer. Although in sample elements are ordered, it may not be the case (that should be cleared by OP). Edit: TakeWhile won't take last element (e4), just checked. Also, you should Select to get the interested value instead of XElement. – Marcin Deptuła Aug 15 '11 at 14:52

score 1 · Answer 3 · edited May 23 '17 at 12:04

an alternative way:

var urls = from url in doc.Descendants("urlset").Elements("url")
    let str = Int32.Parse(url.Element("loc").Value.Replace("e",""))
    where str >= 2 && str <= 4  
    select url;

or actually, a safer option that doesn't throw exceptions if loc is not in the form "e + integer value" (courtesy of Marc Gravell) would be:

int? TryParse(string s)
{
    int i;
    return int.TryParse(s, out i) ? (int?)i : (int?)null;
}

var urls = from url in doc.Descendants("urlset").Elements("url")
    let str = TryParse(url.Element("loc").Value.Replace("e",""))
    where str >= 2 && str <= 4  
    select url;

score 1 · Answer 4 · answered Aug 15 '11 at 15:03

1

You can use this XPath:

//url[loc = 'e2' or 
    (preceding-sibling::url/loc = 'e2' and following-sibling::url/loc = 'e4')
]

It will select url with loc = e2, e3, e4

answered Aug 15 '11 at 15:03

Kirill Polishchuk

54,804
11
122
125

Select a specific range of elements?

4 Answers4