mikejohnpage / newsrivr Goto Github PK

Add arg to get_news() that allows users to supply a vector of known publishers which are matched against the url and/or website.domainName Newsriver query fields.

Store a list of all publishers (matched to the Newsriver publishers list) in a .rda object and reference the list from utils when writing conditional checks to validate provided publishers by the user.

Make get_news() check if language is in the tibble saved as language_codes.RDS, rather than checking within a vector within the function as is done now. Check out Newsanchor's stop_if_invalid_language function for an example of how to do this.

Additionally, some of the language codes (e.g., "zu") appear to return nothing. Valid language codes need identifying. Contact Newsriver for a valid list. For now, the extensive list in get_news() will suffice.

query arg returns an empty tibble if only title: or text: are specified. Omission or inclusion of both does not generate issues:

# okay
get_news("google") 

# okay
get_news("tittle:Google AND text:Google")

# empty tibble
get_news("tittle:Google")

# empty tibble
get_news("text:Google")

Find out if api_token is always a set length of characters to add another precondition check to store_creds(), store_creds_temp(), and get_news() that tests is api_token is the correct nchar().