{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from pathlib import Path\n",
    "import numpy as np\n",
    "import util as u\n",
    "import json\n",
    "import pickle\n",
    "\n",
    "\n",
    "BASE_PATH = \"../\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df = pd.read_csv(Path(BASE_PATH,\"metadata.csv\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "int"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def is_manually_checked(row):\n",
    "    #check based on the audio click files\n",
    "    if Path(BASE_PATH,row[\"midi_performance\"][:-4]+ \"_click.wav\").exists():\n",
    "        return False\n",
    "    else:\n",
    "        return True\n",
    "    \n",
    "def key_number_from_number_of_sharps(nos):\n",
    "    if nos >= 0:\n",
    "        return int((nos * 7)%12)\n",
    "    else:\n",
    "        return int((np.abs(nos)*5)%12)\n",
    "    \n",
    "type(key_number_from_number_of_sharps(-3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_beats_from_txt(ann_path):\n",
    "    ann_df = pd.read_csv(Path(BASE_PATH,ann_path),header=None, names=[\"time\",\"time2\",\"type\"],sep='\\t')\n",
    "    return ann_df[\"time\"].tolist()\n",
    "\n",
    "def get_downbeats_from_txt(ann_path):\n",
    "    ann_df = pd.read_csv(Path(BASE_PATH,ann_path),header=None, names=[\"time\",\"time2\",\"type\"],sep='\\t')\n",
    "    downbeats = [a[\"time\"] for i,a in ann_df.iterrows() if a[\"type\"].split(\",\")[0] == \"db\"]\n",
    "    return downbeats\n",
    "\n",
    "def get_beats_db_dict_from_txt(ann_path):\n",
    "    ann_df = pd.read_csv(Path(BASE_PATH,ann_path),header=None, names=[\"time\",\"time2\",\"type\"],sep='\\t')\n",
    "    out_dict = {str(a[\"time\"]): a[\"type\"].split(\",\")[0] for i,a in ann_df.iterrows()}\n",
    "    return out_dict\n",
    "\n",
    "def get_key_from_txt(ann_path):\n",
    "    ann_df = pd.read_csv(Path(BASE_PATH,ann_path),header=None, names=[\"time\",\"time2\",\"type\"],sep='\\t')\n",
    "    keys = {}\n",
    "    for i, r in ann_df.iterrows():\n",
    "        if len(r[\"type\"].split(\",\"))==3:\n",
    "            number_of_sharps = int(r[\"type\"].split(\",\")[2])\n",
    "            key_number = key_number_from_number_of_sharps(number_of_sharps)\n",
    "            keys[str(r[\"time\"])] = [key_number, number_of_sharps]\n",
    "    return keys\n",
    "\n",
    "def get_ts_from_txt(ann_path):\n",
    "    ann_df = pd.read_csv(Path(BASE_PATH,ann_path),header=None, names=[\"time\",\"time2\",\"type\"],sep='\\t')\n",
    "    tss = {}\n",
    "    for i, r in ann_df.iterrows():\n",
    "        if len(r[\"type\"].split(\",\"))>1 and r[\"type\"].split(\",\")[1]!=\"\":\n",
    "            ts = r[\"type\"].split(\",\")[1]\n",
    "            beat_n = int(u.ts2n_of_beats(ts))\n",
    "            tss[str(r[\"time\"])] = [ts, beat_n]\n",
    "    return tss\n",
    "\n",
    "def midi_and_score_aligned(row):\n",
    "    return len(get_beats_from_txt(row[\"performance_annotations\"])) == len(get_beats_from_txt(row[\"midi_score_annotations\"]))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Creating json\n",
      "###########################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################################"
     ]
    }
   ],
   "source": [
    "asap_ann = {}\n",
    "print(\"Creating json\")\n",
    "\n",
    "counter = 0\n",
    "for i,row in all_df.iterrows():\n",
    "    asap_ann[row[\"midi_performance\"]] = {\n",
    "        \"performance_beats\" : get_beats_from_txt(row[\"performance_annotations\"]),\n",
    "        \"performance_downbeats\" : get_downbeats_from_txt(row[\"performance_annotations\"]),\n",
    "        \"performance_beats_type\" : get_beats_db_dict_from_txt(row[\"performance_annotations\"]),\n",
    "        \"perf_time_signatures\" : get_ts_from_txt(row[\"performance_annotations\"]),\n",
    "        \"perf_key_signatures\" : get_key_from_txt(row[\"performance_annotations\"]),\n",
    "        \"midi_score_beats\" : get_beats_from_txt(row[\"midi_score_annotations\"]),\n",
    "        \"midi_score_downbeats\" : get_downbeats_from_txt(row[\"midi_score_annotations\"]),\n",
    "        \"midi_score_beats_type\" : get_beats_db_dict_from_txt(row[\"midi_score_annotations\"]),\n",
    "        \"midi_score_time_signatures\" : get_ts_from_txt(row[\"midi_score_annotations\"]),\n",
    "        \"midi_score_key_signatures\" : get_key_from_txt(row[\"midi_score_annotations\"]),\n",
    "        \"downbeats_score_map\" : u.same_number_of_measures_with_repetitions(row[\"xml_score\"], str(Path(BASE_PATH,row[\"midi_score_annotations\"])), base_path=BASE_PATH),\n",
    "        \"score_and_performance_aligned\" : midi_and_score_aligned(row) ,\n",
    "#         \"manually_checked\" : is_manually_checked(row)\n",
    "    }\n",
    "    counter +=1\n",
    "    print(\"#\",end=\"\")\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "with open(Path('../asap_annotations.json'), 'w') as outfile:\n",
    "    json.dump(asap_ann, outfile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}