src/search_engine/search_engine_functions.py

from whoosh import index
from whoosh.qparser import MultifieldParser
from whoosh import qparser
acronym_regex = r"([A-Z][\w,’‘']+(?:(?:\s|&|and|or|the|of|to|in|on|at|for|an|-)+[A-Z][\w,’‘']+){1,})\s\(([A-Za-z\s]+)\)"


def search(query,index_dir):
    ix = index.open_dir(index_dir)
    mparser = MultifieldParser(["title", "content","date","acronyms"], schema=ix.schema,group=qparser.OrGroup)
    q = mparser.parse(query)

    s = ix.searcher()
    print(q)
    results = s.search(q, limit=10)
    for r in results:
        print(r['title'],r['file_name'])
    return(list(results))