script to demonstrate basic python-superfastmatch pairwise comparison of all documents in a collection

sfm_compare_all.py

from superfastmatch import client
import uuid

sfm_client = client.Client(url='http://127.0.0.1:9000/')

class Document():
    
    def __init__(self,title_string,content_string):
        self.doc_id = uuid.uuid4()
        self.title = title_string
        self.content = content_string
        self.near_neighbors = []

documents = [doc_1,doc_2,doc_3]

for doc in documents:
    sfm_client.add( 1, doc.doc_id, doc.content, title=doc.title)

for doc in documents:
    sfm_doc = eval(sfm_client.get(1,doc.doc_id))
    comp_list = []
    try:
        for other_doc in sfm_doc['documents']['rows']:
            shared = 0
            for frag in other_doc['fragments']:
                shared += frag[2]
            comp_list.append((other_doc['docid'],shared))
    except KeyError:
        continue
    #this takes the top ten, but you could also do something more clever
    comp_list = sorted(comp_list,key=lambda x: x[1],reverse=True)[0:10]
    for i in range(min([len(comp_list),10])):
        other_doc_id,s = comp_list[i]
        doc.near_neighbors.append(other_doc_id)