dinghino / pyfinder Goto Github PK

for each segment (partial query or attribute value) find the root-word (i.e. searching -> search).

remove all punctuation and split there. This should be already in place with the internal splitter function that as of now calls a re.split(r'\W+'), but I haven't actually tested it

problem is that self.keys is a list of tuples, while keys (param) is a list of strings. normalize the param list to make it work.

for blobs like long strings stopwords should be removed from the string (before or after splitting in chunks)
this could be done removing all words shorter than x character, for testing, bug would work better having a dictionary of stopwords to filter out.