select pageA.page_title

from page pageA

where pageA.page_namespace = 0

# Filter out hard redirects

and pageA.page_is_redirect = 0

and not exists(

  # Filter out talk pages which have anything matching Wikiproject%articles 

  # (There are a few that omit the wikiproject but this seems to not lose any true positives)

  select 1 from categorylinks 

  join page pageB on pageB.page_namespace = 1 

  where pageB.page_id=categorylinks.cl_from 

  and pageB.page_title = pageA.page_title

  and (categorylinks.cl_to like "%articles")

)

and not exists(

  # Main article pages get disambiguation, set index tags and soft redirect indicators

  select 1 from categorylinks

  join templatelinks on tl_from = pageA.page_id

  where pageA.page_id=categorylinks.cl_from 

  # filter out disambiguation and set index and list of lists

  and (categorylinks.cl_to like "%Disambiguation" 

       or categorylinks.cl_to like "%disambiguation%"

       or categorylinks.cl_to = "All_set_index_articles"

       or tl_title = "Disambiguation"

       or tl_title = "List_of_lists"

       # ongoing AfD or prod -> filter out

       or tl_title like "Proposed_deletion%"

       or tl_title = "Article_for_deletion"

       # AfD settled the article will be deleted soon -> filter out

       or tl_title = "Afd-merge_to"

       # Redirects not of the form "R from ... which will be soft redirects with page_is_redirect=0"

       or tl_title = "Soft_redirect" 

       or tl_title = "Double_soft_redirect"

       or tl_title = "Category_redirect" 

       or tl_title = "Portal_soft_redirect"

       or tl_title = "Wiktionary_redirect" 

       or tl_title = "Wikidata_redirect" 

       or tl_title = "Wikibooks_redirect" 

       or tl_title = "Wikiquote_redirect" 

       or tl_title = "Wikisource_redirect" 

       or tl_title = "Wikispecies_redirect" 

       or tl_title = "Wikivoyage_redirect" 

       or tl_title = "Commons_redirect"

       # Filter out any other soft redirects (this is usually due to an ongoing RfD)

       or tl_title like "R\_%"))

limit 200

​