#!/usr/bin/env python3
"""Reproduce Scholar4AI's current-window comparison, using Python's standard library.

Download /data/research-patterns-input.json and run:
  python research-patterns.py research-patterns-input.json > research-patterns.csv
Each OpenAlex work counts once; multiple names on one work do not inflate totals.
"""
import csv
import io
import json
import sys
from collections import Counter


def aggregate(snapshot):
    seen, buckets = set(), {}
    for category in snapshot['categories']:
        buckets[category['name']] = dict(category, papers=0, with_datasets=0,
            with_equations=0, with_both=0, with_abstract=0, with_fulltext=0,
            dataset_names=Counter(), equation_names=Counter())
    for paper in snapshot['papers']:
        key = paper['id']
        if key in seen:
            continue
        seen.add(key)
        row = buckets[paper['category']]
        datasets, equations = set(paper['datasets']), set(paper['equations'])
        row['papers'] += 1
        row['with_datasets'] += bool(datasets)
        row['with_equations'] += bool(equations)
        row['with_both'] += bool(datasets and equations)
        row['with_abstract'] += bool(paper['has_abstract'])
        row['with_fulltext'] += bool(paper['has_fulltext'])
        row['dataset_names'].update(datasets)
        row['equation_names'].update(equations)
    return list(buckets.values())


def as_csv(rows):
    out = io.StringIO(newline='')
    fields = ['name', 'slug', 'papers', 'with_datasets', 'with_equations',
              'with_both', 'with_abstract', 'with_fulltext']
    writer = csv.DictWriter(out, fieldnames=fields, extrasaction='ignore', lineterminator='\n')
    writer.writeheader()
    writer.writerows(rows)
    return out.getvalue()


if __name__ == '__main__':
    with open(sys.argv[1], encoding='utf-8') as source:
        sys.stdout.write(as_csv(aggregate(json.load(source))))