Glider
Loading...
Searching...
No Matches
SanitizationLines.py
Go to the documentation of this file.
1import os
2
3def matching_sales(sales, catalog_collection, releases):
4 if str(sales["release_id"]) == "" or str(sales["release_id"]).lower() == "nan" or str(sales["release_id"]) == "undefined":
5 upc_found = catalog_collection.find_one({"tracks.isrc":sales["isrc_id"]},{"_id":0, "upc":1,"title":1,"label":1,"artist":1,"tracks.isrc":1, "tracks.title":1, "tracks.artist":1, "catalogId": 1})
6 if upc_found:
7 isrc_found = [element for element in upc_found["tracks"] if element["isrc"] == sales["isrc_id"]]
8 sales["release_id"] = upc_found["upc"]
9 sales["release_title"] = upc_found["title"]
10 sales["label_id"] = upc_found["label"]
11 sales["track_title"] = isrc_found[0]["title"]
12 sales["artists"] = isrc_found[0]["artist"]
13 sales["catalogue_id"] = upc_found["catalogId"]
14 sales["clean"] = 1
15 else:
16 upc_found = [element for element in releases if element['upc'] == sales["release_id"]]# and element["tracks"] == track_sales]
17 if len(upc_found) > 0:
18 if len(sales["isrc_id"]) >= 10:
19 isrc_found = [element for element in upc_found[0]["tracks"] if element["isrc"] == sales["isrc_id"]]
20 sales["release_title"] = upc_found[0]["title"]
21 sales["label_id"] = upc_found[0]["label"]
22 sales["track_title"] = isrc_found[0]["title"]
23 sales["artists"] = isrc_found[0]["artist"]
24 sales["catalogue_id"] = upc_found[0]["catalogId"]
25 sales["clean"] = 1
26 else:
27 sales["release_title"] = upc_found[0]["title"]
28 sales["label_id"] = upc_found[0]["label"]
29 sales["artists"] = upc_found[0]["artist"]
30 sales["catalogue_id"] = upc_found[0]["catalogId"]
31 sales["clean"] = 1
32
33def sanitization(lines, catalog_collection, releases):
34 for line in lines:
35 try:
36 line["catalogue_id"] = "no catalogue"
37 matching_sales(line, catalog_collection, releases)
38 except:
39 continue
40 return lines
sanitization(lines, catalog_collection, releases)
matching_sales(sales, catalog_collection, releases)