Parses pages and returns simple objects to lookup/match against. It's loosely an in-memory database.
This creates a matcher that will loosely match strings found in the wiki, created further below. We use the StrCloseComparer to allow for discrepencies in the search text. matcher = WikiMatcher(StrCloseComparer(), BaseIgnorer, wiki, 0.9)
This ignores all instances of any phrase provided in the list. BaseIgnorer = ListIgnorer(['Director', 'Non-Canon Name'])
This creates the wiki, searching all pages currently in it's internal list. This is appended to by the add_page function. Indexes are not created until the update function is called. wiki = Wiki() wiki.add_page(some_page) wiki.update()
This creates the page that can be searched through some_page = WikiPage('www.myfanwiki.com/wiki/Curse_of_Cannibal_Island/Actors', ActorParser(BaseIgnorer))
This is an abstract class, so each instance of this will be unique to your application. For Example: class ActorParser(WikiMatcher.wiki_parser.PageParser): def find_items(self, soup: BeautifulSoup)->ActorItems: items = [] for element in soup.findAll(ElementMatchPattern): items.append(element) return items def create_item(self, data): return ActorItem(data[0], data[2], int(data[1])) # Or Whatever formatting you need
These are what are found by matching against the WikiItem's key_names list. key_names should be unique, but this is not enforced. For example, the following will be searchable by the person's id, but not their non-unique name:
class ActorItem(WikiItem):
def __init__(self, name, person_id, age):
super().__init__(name)
self.person_id = person_id
self.age = age
def key_names(self):
return [ f'actor {self.person_id}', f'actor #{self.person_id}', self.person_id ]