Only American ipa is getting. Percentage of ipa scraping and aggregation is around 90, while lvl scraping is around 94 (mostly because of no lvl on the page at all). Supposed that there no incorrect aggregated ipas and lvls.
Core class is consist of fields: ipa, lvl. Field lvl is object with name, img(shown emoji), img_name,ipa is dictionary with pos (part of speech) as key and ipa html string as value. Converting ipa to aggregated html string is available with ipa_to_str method. Fields lvl and ipa (and most inner object fields) is NOT EXIST if was an error while scraping, it is on purpose and still None if other reasons.
Also possible to use word verification with word and word_filter params, second one is supposed to be func with word and header_of_pagepart params returns True\False depends on accaptability of header_of_pagepart as a word.
pip install dictionarycom-ipa-scraper or equivalent.
from dictionarycom-ipa-scraper import scraper
#some python code with page_html_as_str declaration
scraped = scraper.ParsedCode(page_html_as_str, word)
scraped.ipa_as_str() #html string actually
scraped.lvl.name #lvl of complexity