app.py

import os

from flask import Flask, request, jsonify, send_file, make_response
from flask_cors import cross_origin
import io
import json
import csv

app = Flask(__name__)


def load_pre_processed_file(filename):
    try:
        return json.load(open(filename, "r"))
    except IOError:
        print(f"Cannot load file {filename}")
        return {}

def parse_mgi_ids(req):
    if "file" in req.files:
        file = req.files["file"]
        return file.read().decode("utf-8").strip().split("\n")
    elif req.json and "mgi_ids" in req.json:
        return request.json["mgi_ids"]
    else:
        return []

def transform_data_for_tabular_formats(data):
    results = list()
    for gene_data in data:
        for allele in gene_data["alleles"]:
            allele["mouseGeneSymbol"] = gene_data["mouseGeneSymbol"]
            allele["humanGeneSymbols"] = gene_data["humanGeneSymbols"]
            allele["humanGeneIds"] = gene_data["humanGeneIds"]
            allele["significantLifeStages"] = ", ".join(allele["significantLifeStages"])
            allele["significantPhenotypes"] = ", ".join(allele["significantPhenotypes"])
            allele["significantSystems"] = ", ".join(allele["significantSystems"])
            allele["notSignificantSystems"] = ", ".join(allele["notSignificantSystems"])
            allele["notSignificantPhenotypes"] = ", ".join(allele["notSignificantPhenotypes"])
            results.append(allele)

    return sorted(results, key=lambda k: k["allele"])


PRE_PROCESSED_DATA_PATH = os.environ.get(
    "PRE_PROCESSED_DATA_PATH", "pre-computed-results.json")
preprocessed_results_data = load_pre_processed_file(PRE_PROCESSED_DATA_PATH)


@app.route("/mi/impc/batch-query", methods=["POST"])
@cross_origin()
def query_data():
    # Parse request headers and data
    response_format = request.headers.get("Accept", "application/json").lower()
    mgi_ids = parse_mgi_ids(request)
    if not mgi_ids:
        return jsonify({"error": "No MGI accession IDs provided"}), 400

    mgi_ids = list(dict.fromkeys(mgi_ids))
    # Query the dataset
    filtered_data = list(
        map(lambda gene_id: preprocessed_results_data[gene_id] if gene_id in preprocessed_results_data else None, mgi_ids)
    )
    filtered_data = list(filter(None, filtered_data))

    return jsonify(filtered_data)

@app.route("/mi/impc/download-preprocessed-data", methods=["POST"])
@cross_origin()
def download_summary_data():
    mgi_ids = list(set(parse_mgi_ids(request)))
    if not mgi_ids:
        return jsonify({"error": "No MGI accession IDs provided"}), 400
    response_format = request.headers.get("Accept", "application/json").lower()
    filtered_data = []
    for mgi_id in mgi_ids:
        if mgi_id in preprocessed_results_data:
            final_data = dict(preprocessed_results_data[mgi_id])
            filtered_data.append(final_data)
    if response_format == "application/json":
        return send_file(
            io.BytesIO(json.dumps(filtered_data).encode("utf8")),
            mimetype="application/json",
            as_attachment=True,
            download_name="batch-query-results.zip"
        )
    elif response_format == "text/tab-separated-values":
        data_formatted = transform_data_for_tabular_formats(filtered_data)
        keys = sorted(data_formatted[0].keys())
        string_io = io.StringIO()
        writer = csv.DictWriter(string_io, keys, delimiter="\t", lineterminator="\n")
        writer.writeheader()
        writer.writerows(data_formatted)
        output = make_response(string_io.getvalue())
        output.headers["Content-Disposition"] = "attachment; filename=batch-query-summary-data.tsv"
        output.headers["Content-type"] = "text/tsv"
        return output

@app.route("/mi/impc/batch-query/health-check", methods=["GET"])
def health_check():
    return jsonify({"status": "ok"}), 200


if __name__ == "__main__":
    port = int(os.environ.get('PORT', 5000))
    app.run(debug=True, host="0.0.0.0", port=port)