File size: 1,331 Bytes
fe1c232
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
"""
General utilities
"""

import os

def listdir_fullpath(d):
    return [os.path.join(d, f) for f in os.listdir(d)]

def get_chunks(
    items,
    num_chunks
):
    assert num_chunks > 0, "ERROR: zero chunks"
    item_chunks = []
    num_items = len(items)
    len_chunk = max(1, num_items // num_chunks)
    i = 0
    while i + len_chunk < num_items:
        item_chunks.append(items[i: i + len_chunk])
        i += len_chunk
    if i < num_items:
        item_chunks.append(items[i:])
    return item_chunks

def check_minhash_ids(
    minhash_df
):
    """
    Check if id == line number
    in minhash dataframe. This later
    helps in faster filtering
    """
    for index, row in minhash_df.iterrows():
        id_number = int(row["id"].split("/")[-1])
        assert id_number == index, \
            f"ERROR: check_minhash_ids with {id_number=} and {index=}"

def check_shard_of_filtered_docids(
    filtered_docids,
    expected_shard_number
):
    """
    Sanity check for shard number
    of filtered docids that were populated
    by quality filtering
    """
    for filtered_docid in filtered_docids:
        shard_number = filtered_docid.split("/")[1]
        assert shard_number == expected_shard_number, \
            f"ERROR: check_shard_of_filtered_docids with {shard_number=} and {expected_shard_number=}"