Convert SQL to linq to SQL with cross apply -

this question has answer here:

• how write cross apply query in linq-to-sql? 2 answers

any idea how convert cross apply sql query linq-to-sql?

    select *     dbo.company     inner join dbo.contact c on c.contid = dbo.company.compcontactid or dbo.company.compcontactid null     inner join dbo.company_program cp on cp.compid = dbo.company.compid or dbo.company.compid null     inner join dbo.program p on cp.progid = p.progprogramid     inner join dbo.division d on d.divid = p.progdivisionid     inner join dbo.phonetype pt on pt.phtphonetypeid = c.contphonetypeid or c.contphonetypeid null     inner join dbo.phonenumber ph on ph.phoneid = pt.phtphoneid or pt.phtphoneid null      cross apply (select top 1*         dbo.participation p         p.participationid = dbo.company.compparticipationid or dbo.company.compparticipationid null      )  part      divid = 29 

cross apply evaluate given query each left hand side row, returning first participation company, if 1 or more exist, or first participation in table, if no company participations exist @ all.

ignoring other tables now, should able achieve same result original sql cross apply projection applies same logic above:

var result = db.companies     // include / join on other tables here ...     .select(c => new       {         company = c,         participation = db.partipations              .firstordefault(p => c.compparticipationid == p.participationid                                   || c.compparticipationid == null)     })    // project out final columns company, participation etc.; 

the above linq exhibit 1:n performance issue (i.e. lookup in participations each row of company). there performance optimizations had, e.g. if many of companies don't have participations (i.e. null compparticipationid) pre-fetch first participation in table , apply in coalesce:

var defaultparticipation = db.partipations.firstordefault();  // ... above      participation = db.partipations          .firstordefault(p => c.compparticipationid == p.participationid)          ?? defaultparticipation; 

similarly if there few participations in table, prefetch of them , key them in dictionary participationid, , again coalescing default participation if lookup fails (via trygetvalue or containskey).