FindPagesByCriteria is your friend. Then use EPiServer.Filter to filter for visitor (removed the expired ones and the ones that you do not have access to). Then use LINQ, skip and take to solve paging if needed...
FindPagesByCriteria combined with multiple languages and PageTypeBuilder is not always friends so I would recommend doing it the old fashioned way to avoid some bugs I've seen :)
If you are using it to produce a listing, remember to cache the result to avoid any performance problems...

Hi,

Can you provide the code snippet for "Filter<T>(query, page, pageSize, sort, area, type, publication, year, parentId);"?

/Henrik

Daniel; i will look into it - thanks!

Henrik - certainly, here it is:

private ITypeSearch<PageTypeBase> Filter<T>(ITypeSearch<PageTypeBase> query, int page = 1, int pageSize = 20, string sort = null, int area = -1, int type = -1, int publication = -1, int year = 0, int? parentId = null) where T : PageTypeBase
{
if (typeof(T) == typeof(PublicationPageType))
{
query = query.OrderByDescending(x => ((PublicationPageType)x).Published);
}
else
{
switch (sort)
{
case "date":
query = query.OrderByDescending(x => x.StartPublish);
break;
}
}

if (area >= 0)
{
query = query.Filter(x => x.Category.Match(area));
}
if (type >= 0)
{
query = query.Filter(x => x.Category.Match(type));
}
if (publication >= 0)
{
query = query.Filter(x => x.Category.Match(publication));
}
if (year > 0)
{
query = query.Filter(x => x.StartPublish.InRange(new DateTime(year, 1, 1, 0, 0, 1), new DateTime(year, 12, 31, 11, 59, 59)));
}
if (parentId.HasValue)
{
query = query.Filter(x => x.ParentLink.ID.Match(parentId.Value));
}
return query
.Filter(x => x.MatchTypeHierarchy(typeof (T)))
.Skip(pageSize*(page - 1))
.Take(pageSize);
}

Also, both of the code blocks in my posts gets called here:

var pages = FindHelper.Instance.SearchSitePages<ArticlePageType>(string.Empty, Pager.CurrentPageNumber, CurrentPage.PageSize, "date", Facets.ArticleArea, Facets.ArticleType, Facets.Publication, Facets.ArticleYear, CurrentPage.PageLink.ID);

If i could pass in the all id´s for the children as nodes to retrieve results from, i would be home safe, i guess. Tried to circumvent this but have not found a way to merge several results into one, unforttunately.

//Marcus

If I understand you correctly what have changed is this. Before you had the following structure:

articles
- article1
- article2
-...

and you could filter all articles on ParentLink.ID == articles

Now you have:

articles
- 2013
 - article1
 - article2
 - ...
-2012
 - article3

but when filtering on ParentLink.ID == articles it fails since the "year"-folders are inbetween.

To solve this you could instead filter on Ancestors:

.Filter(x => x.Ancestors().Match(parentOfParentReference.CreateReferenceWithoutVersion().ToString()))

/Henrik

Spot on, Henrik. 

I will try this out first thing tomorrow and let the thread know of my progress. Thanks! :)

//Marcus

Hi again,

since i´m running epi ver. 6, i can´t find the Ancestors anywhere. Or is it a method reference? Care to explaing further (i´m a newbie to epi, as you might have guessed).

I´ve managed to get all the result by another method, but i still need to use a pager to the result list.

//Marcus

Hi,

The Ancestors() extension was added for the cms 7 integration. However you can easily add it yourself. Create the extension:

public static IEnumerable<string> Ancestors(this PageData page)
{
var ancestorLinks = DataFactory.Instance.GetParents(page.PageLink);
var ancestors = DataFactory.Instance.GetPages(ancestorLinks, LanguageSelector.AutoDetect(true));
return ancestors.Select(x => x.PageLink.ToStringIgnoringWorkId());
}

and then add it to the indexing conventions at startup (Iniitializable module or Global.asax):

SearchClient.Instance.Conventions.ForInstancesOf<PageData>().IncludeField(x => x.Ancestors())

Reindex...

Filter using:

.Filter(x => x.Ancestors().Match(parentOfParentLink.ToStringIgnoringWorkId()))

/Henrik