#
# Pyserini: Python interface to the Anserini IR toolkit built on Lucene
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#

"""Filter a TREC run file to either retain only docids that are judged (--retain) or throw away documents that are
judged (--discard)."""

import argparse
import sys
from collections import defaultdict


def load_qrels(qrels):
    judged_docids = defaultdict(set)
    with open(qrels) as f:
        for line in f:
            cols = line.split()
            qid = cols[0]
            docid = cols[2]
            judged_docids[qid].add(docid)
    return judged_docids


def main():
    parser = argparse.ArgumentParser(description=__doc__,
                                     formatter_class=lambda prog: argparse.HelpFormatter(prog, width=100))
    parser.add_argument('--qrels', metavar='FILE', type=str, help='Qrels.', required=True)
    parser.add_argument('--input', metavar='FILE', type=str, help='Input run.', required=True)
    parser.add_argument('--output', metavar='FILE', type=str, help='Output run.', required=True)
    parser.add_argument('--runtag', metavar='STRING', type=str, default=None, help='Runtag.')
    parser.add_argument('--k', metavar='NUM', type=int, default=1000, help='Number of hits to retain per topic.')
    parser.add_argument('--retain', action='store_true',
                        help="Retain judged docids, i.e., throw away all unjudged documents.")
    parser.add_argument('--discard', action='store_true',
                        help="Discard judged docids, i.e., keep only documents that are unjudged.")
    args = parser.parse_args()

    if (args.retain and args.discard) or (not args.retain and not args.discard):
        print('Must specific either one of --retain or --discard.')
        return

    judged_docids = load_qrels(args.qrels)
    counts = defaultdict(int)

    prev_score = None
    check_score = True
    with open(args.output, 'w') as output_f:
        with open(args.input) as input_f:
            for line in input_f:
                cols = line.split()
                qid = cols[0]
                docid = cols[2]
                score = float(cols[4])
                tag = cols[5]

                if args.runtag:
                    tag = args.runtag

                if counts[qid] >= args.k:
                    if check_score:
                        if score <= prev_score:
                            print(f'Warning: scores of {qid} do not strictly decrease at {docid}')
                        check_score = False
                        continue
                    else:
                        continue

                if args.discard:
                    if qid not in judged_docids or docid not in judged_docids[qid]:
                        counts[qid] += 1
                        prev_score = float(cols[4])
                        check_score = True
                        output_f.write(f'{qid} Q0 {docid} {counts[qid]} {score} {tag}\n')
                elif args.retain:
                    if qid in judged_docids and docid in judged_docids[qid]:
                        counts[qid] += 1
                        prev_score = float(cols[4])
                        check_score = True
                        output_f.write(f'{qid} Q0 {docid} {counts[qid]} {score} {tag}\n')


if __name__ == '__main__':
    main()