Question

U use Jsoup to fetch a website. The website has multiple div classes such as:

<div class="itemcategories">
Category: <a id="cat_result_7_newamerican" class="category" rel="newamerican" href="/search?cflt=newamerican&amp;find_loc=willowbrook%2C+IL">American (New)</a>
</div>

<div class="itemcategories">
Categories: 
<a id="cat_result_6_breakfast_brunch" class="category" rel="breakfast_brunch" href="/search?cflt=breakfast_brunch&amp;find_loc=willowbrook%2C+IL">Breakfast & Brunch</a>, 
<a id="cat_result_6_tradamerican" class="category" rel="tradamerican" href="/search?cflt=tradamerican&amp;find_loc=willowbrook%2C+IL">American (Traditional)</a>
</div>

and so on.

If I use the following query selector:

categories = doc.select("div[class=itemcategories] > a[class=category]");

each child element that descend directly from div class=”itemcategories” parent is stored in the next index of the categories Elements object. So I have no way to determine which children belong to which parents. Is there a way to ‘concatenate’ all the children from each div class and save them in a separate indices of the Elements object?

Answer 1

How about doing it in two steps?

Elements parents = doc.select("div.itemcategories");
for (Element parent : parents)
{
    Elements categories = parent.select("a.category");
    // do something with categories
}

Note use of .foo instead of [class=foo] selector syntax.

N.B. I’m not terribly familiar with jsoup’s API so this code may not be exactly correct.