satkit 0.22.0

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "58639a29",
   "metadata": {},
   "source": [
    "# ECOM Solar Radiation Pressure\n",
    "\n",
    "Solar radiation pressure (SRP) is the largest non-gravitational force on a GNSS satellite, and the hardest to model physically: it depends on the spacecraft's shape, surface properties, thermal state and attitude law. The **Empirical CODE Orbit Model (ECOM)** sidesteps all of that by writing the SRP acceleration in a Sun-oriented satellite frame as a small set of constant and once-per-revolution coefficients that are *estimated* from tracking data rather than derived from a surface model. It is what CODE and most IGS analysis centres use for precise GNSS orbits.\n",
    "\n",
    "This tutorial\n",
    "\n",
    "1. downloads a month of IGS final GPS orbits (SP3) for one satellite,\n",
    "2. fits an initial state plus ECOM coefficients to the first three days with `scipy.optimize.least_squares`,\n",
    "3. propagates forward with the fitted coefficients and measures the error against the SP3 truth for the cannonball model, the reduced 5-parameter ECOM, and the 9-parameter ECOM2.\n",
    "\n",
    "> **Experimental.** The ECOM API is new and may be reshaped in a minor release; the physics and conventions are stable.\n",
    "\n",
    "The model, the DYB frame, the sign and eclipse conventions, and the parameter sets are described in the [Empirical SRP: ECOM guide](../../guide/ecom/); the API is [`satkit.ecomparams`](../../api/satprop/#satkit.ecomparams). In short: with $\\hat e_D$ pointing from the satellite **to the Sun**, $\\hat e_Y = \\hat e_D \\times \\hat r$ along the solar-panel axis and $\\hat e_B = \\hat e_D \\times \\hat e_Y$,\n",
    "\n",
    "$$\\vec a = \\nu\\left[D(\\varphi)\\,\\hat e_D + Y(\\varphi)\\,\\hat e_Y + B(\\varphi)\\,\\hat e_B\\right]$$\n",
    "\n",
    "Each coefficient is an acceleration in m/s² along one DYB axis: the constants $D_0, Y_0, B_0$ (all models), once-per-revolution terms $D_c, D_s, Y_c, Y_s, B_c, B_s$ (ECOM1; the reduced model keeps only $B_c, B_s$), and the ECOM2 even harmonics $D_{2c}, D_{2s}, D_{4c}, D_{4s}$ in $\\Delta u$. The full table — axis, harmonic, which model uses it, typical GPS magnitudes, field names — is in the [Empirical SRP: ECOM guide](../../guide/ecom/#coefficients).\n",
    "\n",
    "where $\\nu$ is the Earth-shadow factor, $\\varphi$ is the argument of latitude $u$ (ECOM1) or $\\Delta u$ from orbit noon (ECOM2), and the physical $D_0$ is *negative*, about $-10^{-7}$ m/s² for a GPS satellite.\n",
    "\n",
    "**Runtime and data.** The fits below take a few seconds to a minute each (finite-difference Jacobians over a few hundred propagations), and the data are ~32 daily files (~1 MB each) fetched from BKG's public IGS mirror. The notebook is committed with its outputs and is not re-executed by the documentation build.\n",
    ""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c6aa63ca",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-29T14:37:30.538234Z",
     "iopub.status.busy": "2026-08-29T14:37:30.538066Z",
     "iopub.status.idle": "2026-08-29T14:37:31.120670Z",
     "shell.execute_reply": "2026-08-29T14:37:31.120431Z"
    }
   },
   "outputs": [],
   "source": [
    "import datetime as dt\n",
    "import gzip\n",
    "import os\n",
    "from pathlib import Path\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import requests\n",
    "import scienceplots  # noqa: F401  (registers the \"science\" style)\n",
    "from scipy.optimize import least_squares\n",
    "from scipy.optimize._numdiff import approx_derivative\n",
    "\n",
    "import satkit as sk\n",
    "\n",
    "plt.style.use([\"science\", \"no-latex\", \"../satkit.mplstyle\"])\n",
    "\n",
    "PRN = 20                      # GPS satellite (PRN G20)\n",
    "START = dt.date(2024, 1, 1)   # first day of the window\n",
    "FIT_DAYS = 3                  # length of the fit arc\n",
    "PROP_DAYS = 30                # length of the prediction\n",
    "CACHE = Path(os.environ.get(\"SP3_CACHE\", \"sp3-cache\"))  # where SP3 files are kept"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "90b5c435",
   "metadata": {},
   "source": [
    "## Truth: IGS final orbits\n",
    "\n",
    "The IGS final combined product gives satellite centre-of-mass positions in the ITRF every 15 minutes, accurate to ~2.5 cm. Daily files are named `IGS0OPSFIN_<year><doy>0000_01D_15M_ORB.SP3` under the GPS-week directory on the BKG mirror (no login required). We read one PRN, drop any flagged epochs, and rotate the positions into the GCRF with the full IAU 2006/2000A transform — the frame in which `satkit.propagate` works.\n",
    "\n",
    "SP3 epochs are in **GPS time** (header line `%c ... GPS`), 18 s ahead of UTC in 2024, so they are read with `scale=sk.timescale.GPS`. Reading them as UTC is a classic mistake: the ITRF→GCRF rotation of the truth is then off by 18 s of Earth rotation (1.3 mrad) relative to the Sun/Moon geometry — the orbit stays self-consistent, so nothing looks obviously wrong, but the fit residual quadruples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "c7b44369",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-29T14:37:31.122174Z",
     "iopub.status.busy": "2026-08-29T14:37:31.121923Z",
     "iopub.status.idle": "2026-08-29T14:37:31.237123Z",
     "shell.execute_reply": "2026-08-29T14:37:31.236901Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G20: 2976 epochs, 2023-12-31T23:59:42.000000Z → 2024-01-31T23:44:42.000000Z; fitting on the first 289\n",
      "epochs in Earth shadow: 0 (none — no eclipses in this window)\n"
     ]
    }
   ],
   "source": [
    "GPS_EPOCH = dt.date(1980, 1, 6)\n",
    "BKG = \"https://igs.bkg.bund.de/root_ftp/IGS/products\"\n",
    "\n",
    "\n",
    "def fetch_sp3(day: dt.date) -> Path:\n",
    "    \"\"\"Download (once) the IGS final SP3 for `day` into CACHE.\"\"\"\n",
    "    doy = day.timetuple().tm_yday\n",
    "    name = f\"IGS0OPSFIN_{day.year}{doy:03d}0000_01D_15M_ORB.SP3\"\n",
    "    out = CACHE / name\n",
    "    if not out.exists():\n",
    "        week = (day - GPS_EPOCH).days // 7\n",
    "        r = requests.get(f\"{BKG}/{week}/{name}.gz\", timeout=120)\n",
    "        r.raise_for_status()\n",
    "        CACHE.mkdir(parents=True, exist_ok=True)\n",
    "        out.write_bytes(gzip.decompress(r.content))\n",
    "    return out\n",
    "\n",
    "\n",
    "def read_sp3(path: Path, prn: int):\n",
    "    \"\"\"(times, ITRF positions in m) for one GPS PRN; bad epochs dropped.\"\"\"\n",
    "    times, pos, current = [], [], None\n",
    "    for line in open(path):\n",
    "        if line.startswith(\"*\"):\n",
    "            y, mo, d, h, mi = int(line[3:7]), int(line[8:10]), int(line[11:13]), int(line[14:16]), int(line[17:19])\n",
    "            current = sk.time(y, mo, d, h, mi, float(line[20:31]), scale=sk.timescale.GPS)  # SP3 epochs are GPS time\n",
    "        elif line.startswith(f\"PG{prn:02d}\") and current is not None:\n",
    "            x, y_, z = float(line[4:18]), float(line[18:32]), float(line[32:46])\n",
    "            if any(abs(v) > 900000.0 for v in (x, y_, z)) or (x == 0.0 and y_ == 0.0):\n",
    "                continue\n",
    "            times.append(current)\n",
    "            pos.append([x, y_, z])\n",
    "    return times, np.array(pos) * 1e3\n",
    "\n",
    "\n",
    "def load_truth(start: dt.date, ndays: int, prn: int):\n",
    "    \"\"\"Concatenate daily files into GCRF truth (times, positions in m).\"\"\"\n",
    "    times, pos = [], []\n",
    "    for i in range(ndays):\n",
    "        t, p = read_sp3(fetch_sp3(start + dt.timedelta(days=i)), prn)\n",
    "        for tt, pp in zip(t, p):\n",
    "            if times and tt == times[-1]:  # duplicate day-boundary epoch\n",
    "                continue\n",
    "            times.append(tt)\n",
    "            pos.append(pp)\n",
    "    pos = np.array(pos)\n",
    "    gcrf = np.array([sk.frametransform.qitrf2gcrf(t) * p for t, p in zip(times, pos)])\n",
    "    return times, gcrf\n",
    "\n",
    "\n",
    "times, truth = load_truth(START, PROP_DAYS + 1, PRN)\n",
    "n_fit = sum(1 for t in times if t <= times[0] + sk.duration.from_days(FIT_DAYS))\n",
    "print(f\"G{PRN:02d}: {len(times)} epochs, {times[0]} → {times[-1]}; fitting on the first {n_fit}\")\n",
    "\n",
    "# Is this an eclipse season for this satellite? (cylindrical shadow test)\n",
    "in_shadow = 0\n",
    "for t, p in zip(times, truth):\n",
    "    s_hat = sk.sun.pos_gcrf(t)\n",
    "    s_hat = s_hat / np.linalg.norm(s_hat)\n",
    "    along = p @ s_hat\n",
    "    if along < 0 and np.linalg.norm(p - along * s_hat) < sk.consts.earth_radius:\n",
    "        in_shadow += 1\n",
    "print(f\"epochs in Earth shadow: {in_shadow} ({'eclipse season' if in_shadow else 'none — no eclipses in this window'})\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fd8385e2",
   "metadata": {},
   "source": [
    "## Force model and SRP variants\n",
    "\n",
    "Everything except SRP is the same for all three variants: EGM96 to degree and order 12 (more than enough at 26,600 km), Sun and Moon, solid Earth tides, relativity. The variants differ only in the `satproperties` passed to `propagate`:\n",
    "\n",
    "| variant | free coefficients | `satproperties` |\n",
    "|---|---|---|\n",
    "| cannonball | $C_R A/m$ (m²/kg) | `satproperties(craoverm=c)` |\n",
    "| ECOM reduced | $D_0, Y_0, B_0, B_c, B_s$ (m/s²) | `satproperties(craoverm=0, ecom=ecomparams.reduced(...))` |\n",
    "| ECOM2 | $D_0, Y_0, B_0, B_{1c}, B_{1s}, D_{2c}, D_{2s}, D_{4c}, D_{4s}$ | `satproperties(craoverm=0, ecom=ecomparams.ecom2(...))` |\n",
    "\n",
    "`craoverm=0` makes ECOM the *whole* SRP model rather than a correction on top of the cannonball. Coefficients are handled in nm/s² inside the fit so that all parameters are O(1–100).\n",
    "\n",
    "Two setup notes. This window has no eclipses (checked above); for an arc that crosses Earth's shadow use `settings.integrator = sk.integrator.gauss_jackson8`, because the adaptive Runge–Kutta steppers can abort at a shadow boundary with *too many consecutive step rejections* (the fixed-step multistep integrator is immune, and fits an eclipsing satellite just as well — see the benchmark below). And keep the default `egm96` gravity model with `solid_step1` tides: EGM96 is a *tide-free* model, whereas `jgm3` and `itugrace16` are *zero-tide*; adding Step-1 solid tides (which include the permanent tide) to a zero-tide model double-counts it, ~5–7 cm cross-track over 3 days at GPS altitude."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "bf5c971d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-29T14:37:31.238310Z",
     "iopub.status.busy": "2026-08-29T14:37:31.238241Z",
     "iopub.status.idle": "2026-08-29T14:37:31.240663Z",
     "shell.execute_reply": "2026-08-29T14:37:31.240461Z"
    }
   },
   "outputs": [],
   "source": [
    "settings = sk.propsettings()\n",
    "settings.gravity_degree = 12\n",
    "settings.gravity_order = 12\n",
    "settings.use_sun_gravity = True\n",
    "settings.use_moon_gravity = True\n",
    "settings.tide_model = sk.tidemodel.solid_step1\n",
    "settings.use_relativistic_correction = True\n",
    "settings.use_spaceweather = False      # no drag at GPS altitude\n",
    "settings.abs_error = settings.rel_error = 1e-11\n",
    "settings.enable_interp = True          # so one propagation can be sampled at every SP3 epoch\n",
    "\n",
    "# name: (initial coefficients, absolute finite-difference step per coefficient, builder)\n",
    "VARIANTS = {\n",
    "    \"cannonball\": ([0.02], [1e-3], lambda c: sk.satproperties(craoverm=c[0])),\n",
    "    \"ecom-reduced\": (\n",
    "        [-100.0, 0, 0, 0, 0], [0.1] * 5,\n",
    "        lambda c: sk.satproperties(craoverm=0.0, ecom=sk.ecomparams.reduced(*(np.asarray(c) * 1e-9))),\n",
    "    ),\n",
    "    \"ecom2\": (\n",
    "        [-100.0] + [0.0] * 8, [0.1] * 9,\n",
    "        lambda c: sk.satproperties(craoverm=0.0, ecom=sk.ecomparams.ecom2(*(np.asarray(c) * 1e-9))),\n",
    "    ),\n",
    "}\n",
    "\n",
    "\n",
    "def propagate_states(x, build, t0, t1, at_times):\n",
    "    \"\"\"Propagate [x,y,z (km), vx,vy,vz (m/s), coeffs...] and sample at `at_times`.\"\"\"\n",
    "    state = np.concatenate((x[:3] * 1e3, x[3:6]))\n",
    "    res = sk.propagate(state, t0, t1, propsettings=settings, satproperties=build(x[6:]))\n",
    "    return np.array(res.interp(at_times))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "28efadeb",
   "metadata": {},
   "source": [
    "## Fit: initial state + SRP coefficients\n",
    "\n",
    "The unknowns are the GCRF state at the first epoch (position in km, velocity in m/s, so the parameters are similarly scaled) and the SRP coefficients. The residual is the propagated-minus-truth position at every SP3 epoch in the fit arc; `least_squares` builds a finite-difference Jacobian, so each iteration costs one propagation per parameter.\n",
    "\n",
    "Two practical details: the initial velocity is seeded by differencing the first two truth points, and the finite-difference steps are set explicitly — scipy's default relative step on a coefficient whose current value is *zero* is far below the integrator's noise floor.\n",
    "\n",
    "The Jacobian is finite-differenced with **absolute** steps (1 m, 0.1 mm/s, 0.1 nm/s²): scipy's `diff_step` is relative and silently falls back to √ε for parameters that are currently zero, which is far below integrator noise for the harmonic ECOM terms. The initial velocity comes from a five-point one-sided stencil (a two-point chord over a 15-min step is ~250 m/s off at GPS altitude)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "dfeb0af0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-29T14:37:31.241647Z",
     "iopub.status.busy": "2026-08-29T14:37:31.241579Z",
     "iopub.status.idle": "2026-08-29T14:37:34.127709Z",
     "shell.execute_reply": "2026-08-29T14:37:34.127438Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cannonball    fit RMS  3.812 m (3D)   coefficients (m²/kg):    0.023\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ecom-reduced  fit RMS  0.050 m (3D)   coefficients (nm/s²): -105.768,    0.969,    0.561,   -0.179,   -0.465\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ecom2         fit RMS  0.048 m (3D)   coefficients (nm/s²): -106.042,    0.971,    0.556,    0.771,    0.172,   -1.570,    0.010,    1.197,   -0.232\n"
     ]
    }
   ],
   "source": [
    "def initial_velocity(t, p):\n",
    "    \"\"\"O(dt^4) one-sided stencil from the first five positions.\"\"\"\n",
    "    h = (t[1] - t[0]).seconds\n",
    "    return (-25 * p[0] + 48 * p[1] - 36 * p[2] + 16 * p[3] - 3 * p[4]) / (12 * h)\n",
    "\n",
    "\n",
    "def fit(variant, tf, pf):\n",
    "    c0, cstep, build = VARIANTS[variant]\n",
    "    x0 = np.concatenate((pf[0] / 1e3, initial_velocity(tf, pf), c0))\n",
    "    steps = np.array([1e-3] * 3 + [1e-4] * 3 + cstep)  # km, m/s, coefficient units\n",
    "\n",
    "    def resid(x):\n",
    "        return (propagate_states(x, build, tf[0], tf[-1], tf)[:, :3] - pf).ravel()\n",
    "\n",
    "    sol = least_squares(resid, x0, jac=lambda x: approx_derivative(resid, x, abs_step=steps),\n",
    "                        x_scale=\"jac\", ftol=1e-12, xtol=1e-12, gtol=1e-12, max_nfev=400)\n",
    "    return sol.x, np.sqrt(np.mean(np.sum(sol.fun.reshape(-1, 3) ** 2, axis=1)))  # 3D RMS\n",
    "\n",
    "\n",
    "fits = {}\n",
    "for name in VARIANTS:\n",
    "    x, rms = fit(name, times[:n_fit], truth[:n_fit])\n",
    "    fits[name] = x\n",
    "    unit = \"m²/kg\" if name == \"cannonball\" else \"nm/s²\"\n",
    "    print(f\"{name:13s} fit RMS {rms:6.3f} m (3D)   coefficients ({unit}): \"\n",
    "          + \", \".join(f\"{c:8.3f}\" for c in x[6:]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6a31a4f",
   "metadata": {},
   "source": [
    "The ECOM fits sit at the accuracy of the IGS final orbits themselves (~2.5 cm per axis, i.e. ~4 cm 3D), some 75× tighter than the cannonball, and the coefficients are physical: $D_0 \\approx -106$ nm/s² is the expected $-P_\\odot\\,C_R A/m$ for $C_R A/m \\approx 0.023$ m²/kg (which is what the cannonball fit found), and the Y and B terms are below 1 nm/s². The reduced 5-parameter set and the 9-parameter ECOM2 give the same answer here — for a cube-like GPS bus the extra harmonics add nothing."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0b4d788b",
   "metadata": {},
   "source": [
    "## Prediction: propagate a month with the fitted coefficients\n",
    "\n",
    "From each fitted state and coefficient set, propagate over the whole window in one call and compare with the truth at every epoch. The RTN decomposition tells us *where* the error goes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "57619169",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-29T14:37:34.128896Z",
     "iopub.status.busy": "2026-08-29T14:37:34.128816Z",
     "iopub.status.idle": "2026-08-29T14:37:34.495404Z",
     "shell.execute_reply": "2026-08-29T14:37:34.495152Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3D RMS error (m) on day N                1       3       7      14      21      30     R/T/N rms on day 30 (m)\n",
      "cannonball                          3.81    3.72  143.22  869.20 2267.23 5126.79       21.1 /  5126.7 /   5.9\n",
      "ecom-reduced                        0.06    0.05    0.76   13.10   36.58   97.82        3.3 /    97.8 /   0.8\n",
      "ecom2                               0.06    0.04    0.74   12.88   35.90   95.86        3.5 /    95.8 /   0.8\n",
      "cannonball    first epoch above 10 m: day 3.1\n",
      "ecom-reduced  first epoch above 10 m: day 12.3\n",
      "ecom2         first epoch above 10 m: day 12.4\n"
     ]
    }
   ],
   "source": [
    "def rtn(pos, vel, d):\n",
    "    r_hat = pos / np.linalg.norm(pos)\n",
    "    n_hat = np.cross(pos, vel); n_hat /= np.linalg.norm(n_hat)\n",
    "    return np.array([d @ r_hat, d @ np.cross(n_hat, r_hat), d @ n_hat])\n",
    "\n",
    "\n",
    "tdays = np.array([(t - times[0]).days for t in times])  # duration.days is fractional\n",
    "results = {}\n",
    "for name, x in fits.items():\n",
    "    st = propagate_states(x, VARIANTS[name][2], times[0], times[-1], times)\n",
    "    d = st[:, :3] - truth\n",
    "    results[name] = (np.linalg.norm(d, axis=1), np.array([rtn(s[:3], s[3:], dd) for s, dd in zip(st, d)]))\n",
    "\n",
    "print(f\"3D RMS error (m) on day N          {'1':>7} {'3':>7} {'7':>7} {'14':>7} {'21':>7} {'30':>7}     R/T/N rms on day 30 (m)\")\n",
    "for name, (err, comp) in results.items():\n",
    "    cells_ = []\n",
    "    for day in (1, 3, 7, 14, 21, 30):\n",
    "        m = (tdays >= day - 1) & (tdays < day)\n",
    "        cells_.append(f\"{np.sqrt(np.mean(err[m] ** 2)):7.2f}\")\n",
    "    m = (tdays >= PROP_DAYS - 1) & (tdays < PROP_DAYS)\n",
    "    rr = np.sqrt(np.mean(comp[m] ** 2, axis=0))\n",
    "    print(f\"{name:13s}                    \" + \" \".join(cells_) + f\"     {rr[0]:6.1f} / {rr[1]:7.1f} / {rr[2]:5.1f}\")\n",
    "\n",
    "for name, (err, _) in results.items():\n",
    "    above = np.nonzero(err[n_fit:] > 10.0)[0]\n",
    "    print(f\"{name:13s} first epoch above 10 m: day {tdays[n_fit + above[0]]:.1f}\" if above.size else f\"{name:13s} never above 10 m\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "331f2636",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-29T14:37:34.496604Z",
     "iopub.status.busy": "2026-08-29T14:37:34.496507Z",
     "iopub.status.idle": "2026-08-29T14:37:34.803342Z",
     "shell.execute_reply": "2026-08-29T14:37:34.803109Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA8sAAAHWCAYAAABAJhvDAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjcsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvTLEjVAAAAAlwSFlzAAAPYQAAD2EBqD+naQABAABJREFUeJzsXQV4FFcXPXF3Dwnu7lKc4k6pG3V3/+sCdXcvbWkLLbS0pS0tbaFUgFLcXRJC3N32/84Ls2yWTbJJdl7YyTvftyyZfTvnzZ0zb+fOu+9eF5PJZIKCgoKCgoKCgoKCgoKCgoIZrif/q6CgoKCgoKCgoKCgoKCgoJxlBQUFBQUFBQUFBQUFBQUbUDPLCgoKCgoKCgoKCgoKCgpWUM6ygoKCgoKCgoKCgoKCgoIVlLOsoKCgUA+kpKRgzpw56NWrF/r164fHH38clZWVyoYKCgoKCgoKCgaDcpYVFJoR/vjjD1x66aXCyRsyZAjatWuHbt264dZbb8X999+PSy65xNw2NTUVs2fPhp+fH1xcXBASEiIcxDZt2qBVq1bisyVLllTbf1lZGZ555hnRrm/fvmjZsiVcXV3F9/v37293Pzdt2oQbbrhB7OOMM85Ax44dxevqq6/G008/jdGjR5/ynZ9++kn0adiwYYiJicGkSZOwfv36U9oVFxfj0UcfFfseNGiQ2P+KFSvs6hedYu63Q4cO+Ouvv+Dt7Y1HHnlE9NcZ4GgbZWVl4Z577sHkyZOFTdq3b4/HHnsMpaWltfYjLS0N0dHROPvss+vV/59//hlDhw7F4MGDhcbIVVJSckq7I0eO4K233sLUqVOFfg8fPoz6Yu/evZg1axYGDBiA3r1747LLLsPx48dPaZednY1FixaJByhRUVGYP38+GgpeTzw+Xi98addnXFwcunbtijvvvBO7d+9u8P6NjG3btiEiIgITJ05ssj7cdttt6NSpk/n8hYeHm1/BwcFwc3Mzf9YQTTojOGYHBQXht99+q7PtV199hYsvvlj8xlx++eUN4vv6668xbtw48++Ov79/tfPAv7VzwGua4BjSs2dP8bvG372mQEJCgri++TvJcbQu2Ptbe/fdd5s1Sf316NFDjCktWrQQY828efPEGGb9O3f++ecjNDTUbEPycDwiD383PvroI1RUVOhiCwWF0w4sHaWgoGBsFBcXm66++mqTr6+v6b333jOVlJSYP9u+fbtpyJAhLCFnuuiii0757ty5c8Vnd955p3nbjh07TL179xbbb7nlFvP2iy++2BQSEmLas2ePedt///1natGihalfv3519rOystL04IMPmjw8PExPPfWUqaCgwPzZkSNHTFOnThWcQ4cOrfa9Tz/91PTcc8+ZysvLxd+pqanimLy8vEz//vtvtf3PmjXL1K5dO1NKSorYtmTJEpObm5tp6dKldfZv2bJlgv+nn34Sf2dnZ5u++eYbM+/pDEfbKD8/33TBBReY9u/fX42D9uH3a8OMGTNEu9mzZ9vdf/bB1dVVvBPsG/t49tlniz5reOONN0wzZ840XXPNNSZPT0/Bc+jQIVN9sG/fPlNYWJjp1ltvFfsuLS01nXXWWaaOHTuaMjIyzO3+/vtv08SJE03XXnutKT4+XnB99NFHpsaiQ4cOYl/btm0Tf7MPH3/8sTgenq9//vnH1FxB/Y4aNeqU7WvWrDG5u7ubunXrZmpK5ObminPE85eXl3fKOPzVV1+J81hfTTorpkyZImzx5Zdf1nkek5OTTfPnzxft58yZ0yhe/pZxP6+99topnyUmJprGjh1r5uBYFhkZKX4fDx8+XK3t3r17TVdccYVNDv7eOeo8FhYWit9VaqNVq1Z1tq/Pby2Pl7bgmKaB9wAPP/yw2M7xhrq1xoIFC04Zp48ePWo688wzxfbp06c34ogVFJwHyllWUGgG4M08f9zo2NkCb+r4A23LWebNv7WzrDnZ3E4H5vjx48KZrckBIq89zjIdZO7jpZdesvk5b7AGDhxYzVkuKysz3XXXXae0Xb9+vdgXHWwNvGHjNt4EWGL06NGm6OhoccNSG+bNmye+v3LlSpMzQQ8b8aELNVDTzTFv3Gzh/fffN/Xt27dezjI5IyIixA2uJehA1qbrM844o0HO8uTJk02BgYHVHtbwJtrFxUU40LZw//33O8xZHjlyZDVnWQPPIbezf80V1F1NzgQfABUVFZmaGtqDE2tnWQMf8DQXZ5ljT1JSkt3nkXZxhLOsOYK2nGVi06ZN1Th4rjIzM09px99EW33ZtWuXeDjj6PNIh7YuZ7khv7XWzrKGYcOGic9efvnlUz7j75wtnrS0NJO3t7f4zPJBq4KCUaHCsBUUDI5//vkH77zzDsaMGYOZM2fabMMwq5tvvrle+2UoF0MLGbLFkEItRPXQoUOntJ0+fTpiY2Nr3d/Ro0fx4IMPihC0W265xWYbhpEx7NcS7u7ueO65505p2717d/G+ZcsW87Y333xTvE+ZMqVaW4brJicnnxJWbo26wotPV+hho6uuukpowJ59ajh48KAIj547d269+v/ll1+K0G32wRJaH19//XWb3/Pw8EB9QS3/+OOP4nrx9fU1b2eIJo/tvffesxn63RCu+oLh88SBAwfQHLF//37ce++9NX7OMGwujWhqMBy2NlD/DNlvDuDYwyUf9TmPjgDDh2sDw4pvv/32ar+BXGpkCY51n3/++SnfLSoqEiHc5eXlaAo09rfWElxmQ+zZs8fu7zCcvXPnzs16LFJoXlDOsoKCwfHhhx+K93PPPbfWdmPHjm0wR2RkpFiz6unpiY0bN+Kuu+6qtp6JN4/fffddrfv45JNPxHe4jrW2m80zzzyzzptRIj8/X7wHBgaKdzo4a9asEWu16ORbQnP6fv/99xpvTrhOlmtkCa6Z5t/XXnut2OcVV1whnHw+cGBb9pG8y5cvN+/jl19+ETbmTRrXjXFdmXZuNDz77LNiPTmdtHXr1uGNN97AhRdeiPj4eGFf9i8zMxMPPPAApk2bhrCwMLGemOs1GwJH2qimfWrgueU6RDq2Pj4+9ernypUrq/VBA4+fa5///vtvmw5sQ7Bq1SqbXNq2wsJCcW7qCx47r5OtW7c2uo/cD/Haa6+JBwY8V//99x+WLl0q1hRyfTPXnWsPeLgukWsytfWQXLduqZnt27eLh1DcL3XFteFcg02d8sb4pptuMu/P8iEc2w4fPlysqeY7r31rsF9cR0zdt27dWjxwoL779OkjriF798fJMa4Jpr6065Ev2pO5GC666CK0bdvW5lrXf//9FzNmzBDHQ16+8+GR5RjFa5EPI6hbOkn8m/vktcc1n1wL21js27dP5ALg/rRroLZzSKeMYw7HE669T09PFw+b6Ghb5m3Izc0V54/f4Rp7nmM6c1y7r2Hnzp0iDwTtT4cqLy8P55xzDgICAvD++++f0tcrr7zSvMaaDi8fjhHUjrad5/Hll18W2/kwi2MUt3t5eYl+3nfffRgxYoRop63Rru08WoPjJ88nxzieF/apsQ8sudZ31KhRop/UJB/SXn/99SIvgeVaYTqBzI/B/v7www+ij9rDKl5PmpPIvAb8zDJXAc8ztcM1wdwnz8k333xzSl943Z133nnma5O/Dxxf6kJjf2stofFxLG3MWKSgYGg09dS2goKCvujcubMIl1q7dm2Dvl9XGHb//v3N2x555BGxjS+GgjHM115w7Se/t3DhQpMj8MMPP4j93XvvveLv3bt3m/tljc2bN4vPGLZbG7TjswzDZgj6J598IrZffvnlYm3rhRdeKP5myDHx5ptvijW/lsfGMGeG9d52223VOGhnftff31+swyQYDty1a1exno3rZhlCqHFz3Vr37t1PGxsNGDDA5OPjc0pII9e+33fffbWG99WEwYMH2wxLJrS181xbWFM4c31CJdnHmsI3ea742bvvvlujNmoKw9bCFp999tlGh2G//vrr5lBQ6k3bL9eQUydce5mTkyPWWg8fPlyMAVxrqIW0c80421iufeYaRk13b731lnkd+Oeffy62c42khg0bNggtsp0Wlsm1km3atBFhtxq2bt0qtHDZZZeJv/kZ15Nbh33auz+CIarWYarp6ek1rnX97rvvRA4ErtfX8Pvvv4t1xbSd5Xp3nnNt/Przzz/NfR4zZoxoz+vNHrB/tsKwV69eLXRiibrO4ZYtW8w5Jbg85corrzTFxcWJF5GVlSXOL8NptWuO6+p5zYSHh5tzCnDbr7/+KvYzadIkcT65rp/LaJgnwhaoc7YfP358te0vvPCC2G69JKGiokLYWsvpwDW42rVrfQ3aOo+WYdh8UQ/cJ/HBBx+IbS+++KJd50C7Hq2v42PHjonrSwPHVi4Z8fPzO6U/tYWEc5ut4+I1yyUj2m8E9XX33XeLtl9//XU1DQYEBJh/IyyXldizZrm+v7W2wrA5FnDJAM8Zz5U1ahqnqSVeD7GxsdXynygoGBXKWVZQMDi49pI/eExa5ChnmTcXvXr1Ej+0vNmzBG/4NMeAN2JMjsKb2brQs2dP8Z0VK1aYHAGuw+XNItcxEnQ8uX/eVFqDjhY/401nfZ1ly5uq0NBQ04EDB8Q2LTlWQkKC+cbcGlyDa72+V+P4/vvvq7Xljamt7ZrzwZvmprYRj4NtrB0Cbuc6Ze3Gqr7OcqdOnUR7y2Ri1uuSbT0MaoizrK3vt7yJtV6X/PTTT9fbWaYTzrX2to6hLmeZN7V0FuhQnn/++WYHwpKXOtDW9WvXG69FfkaH0XrNI6/NHj16VNvOtiNGjDilP1pCH20M0farPcwhrr/+erHNMuHQVVdddcq50dbJDxo0yLzN3v3Z42RZOjZcv0wHgdqzhtY3LWGc5VjHJHGW0JwYOuQNdZbZl/POO++Ua6Ouc2jpmGkP1pgojMkFiRtuuEF8RqfaEn/99ZfYPm3atGrbuY3js2ZrPpiw1JMluJ1J9NjeMrEdE2JxP9b6IScfXtnjVNZ1Hq2TR3Hs4ANHPrhoqLPM43nooYeqOcu19achzjKvH8sHS5pDzvXN2oNlXs90NC+55JJT9stEY/Y4y/X9rbV2ljdu3CjyUAQHB5sWL15s8zu2xmk+UOU54ENa7YGSgoLRocKwFRQMDi08tab1VSwhxVBDhj7yxbCxXbt22Vw3yrYMLXv44YdFCaodO3aY1y5ZlqpgeBnLCXE9M8MZGWbG8N7G9JMhg5b9ZMhlTeVIGIbGMkkM7eY6RqLqfsH2WjYtrFtr01BwTS37ZRmetnjxYnFsDKW0xllnnSXebYXocQ2dJbSwaOvtLMtC5OTk1KuvjrYRzxvDPBl2ybXnGhhKyjD1Dz74QIQONgQyzp3eXE899ZQoN8byK/aCodAMQZ0wYYIID2eI7hdffGFzGYK2zp8hslpI5WeffSberbXHMF+GfTIUm+tHLWHruHndW4bDs18Mp2d4qQZNQ1zXruHYsWPi3TKkn2WBrPVq7/7qC4bxZmRk1Hrt2QqvtlyrTmhrQLXjsRdc984Q3YEDB4pxgSXGaoOtc2jrc4Y587qnDrmmlrbiGGsJjtMcg2gD69Be9kcLgWeYfU3LWrj9mmuuESH4HCcsS8YR1M+GDRvM2xcsWCBKRTkC1uuHOXZwW2JiYr3288ILL5hDvfnbxhBqvcBQ9D///FPYQePkizrgum3Nbhx7k5KSbJY5q88Slfr+1vKa47mndlhuin2jPRlaXxtWr15t/u2nBhnGzjXOWli6goLR4d7UHVBQUNAXvBniTR5/yK0dW4I/rB9//LFY48abWjrAmgNmCa55fv755+3ipEPAdV50xriulzUk6Uiydm1Na6PYT/4As5811eDk2kRtvRmTRXENpDW4No7O2SuvvCLqQWpgzUjC1powOnSNWbdVm6OhOSNcW2sNOv9anxuL+jhwetiICXu4dpLJsbjGUQPXUnIdOtcE1gauX+T6P0twLedDDz2ky7kjl2XdZN7M8qGFDJ3YC16XWsK0hmqPa1KtnT9Ne1zLSy3UVdeV1ybBNbMEnbAbb7xRaI41xumQ8aEQwZt2DV26dBFjAPvBdboE19xbrwm3d3/1haOuPU3P9V0vy/X9lg+3eFxr165tcFIq68/5IIA1crkG2xZ4jKwbTMfM8hzXxWMJrhnmNcjEdlxrTPD/TI710ksvibW6zLPAtcAc87nOVy9w3XZ9cxOwdjHX3FuuJ3aUQ28NrhGnhrle+7rrrquxnbYW3xFJ3urzW8vfdV7z69evF44u16pr+TZqAx+AatejgkJzhJpZVlAwOHgjQ/AmtCZoT4jpfNpylBsKOmK8OaQjwptkWzOoDekn+2jLUaazxiRBt956q7j5tgRntJix2NZMlbaNN/eOhpZ8xdZNv3YTXt+EV42BHjbijAZn6JjEzPoGjUnKmKSIx6q9tJk+fkf7mxqhVixfvEknOnbsWK0P1v3iTJs2Y2kvqEVLLk2bdXHVZIPTEdReTc5mfbRHR4jQHiQQn376qXDSOCPMWWom0LI188UokCeffFI4dewLE0LxfDH5kyXs2Z+zX3ucSWSCr4ZEgtT3+Bx1jJy15iw8E4Rx1pQRJLxWeE6ZLI0z23Rg+ZCMSccc+fuhB5gcS0vipxfqyhCtzTBr15Xs31pGcDzxxBPiwTQrZBQUFDisHwoKRoRylhUUDA4tCzZnZmsCw/4s3xsC3gQvW7bslO0MYdQyqXImpK5+0oGyzO5pbz95w8YbbO5Hc7IIhpkxLI5hfAw74ywLb9wtwdl0wjLDrKOgPbW3zEyrQZupq+vJvqOgh414zp955hkx66/N1hGMQuCNNbPcbt68udpLy77L7K+Wf9eEkSNHVuuDBvaRURPMnGw5m90Y1MSlbePslmW48OkM6oo3wrauO2qPM4zasoHawIzBBEM3CTp8XIbx4osvivD6cePG2Qyx54wur33qgBEhdIJTUlJE9nL+X4O9+3Pma8/6gdUFF1zQ6P3QkWWWaM4m2oos4TFyZruxM5jaLClnlHm98/ywRBfPGR2z77//XoT86zVj6yzgA1xeUwy3t+WA0pnltaBFajQ0oqixv7UEs2hz3OSYxmzgCgoKNUM5ywoKBgdLU3DWlusl61ozZwua41qTA2s9i2gL2o1vbSGlXA/Fkia8sbU33Nuyj7z55E2cpRNIcAZEu5FkSKFWCsQSvNnjjKh1GLA1tBmc+oSG8sk91/6xHI+tEHg6/gxTrgvaMVjfFGtrvOvqkx42otNz//3349dffxXrYDWw/idnCunA8pxbv7RZYN7oW/5dE2gfhhNb3yBqf9d0k16TzWoDnUeGHTKE1jIUmw4ja0fTQeDMaH25OJPD/WqOp6OuudqgrUW01h7DibnWlA8GrGvgWs928XgYMcC1kNpDAjq1dNTGjx9fKz/DUVkaiTrhQxGGnzKcU4si0WDv/rSxxN76ttQ6dbNixYpTzou2rvP888+Ho1GX3rj+lzPujQXHFV6PdL6t16nyoQSXqvDaaeyDJOqES3i41vadd94R4b4Ey9px33zAwZD3+qxhrc951FCf67gx4fvWv1u2+mnrMzrBfJjIhxe89rT11ew3HyZwvTTtRSeVaEw5svr81tqKQKB2GELPh38cqy1LXzl6HFJQcHo0dYYxBQUF/cGyMSwvxGy6LNnCbJyWpTSYpZXDAbMjW+P2228Xn40aNapamRVrsKwJ2zHbqGU5icTERFPbtm1FhuWasq5qYEZn8rCkEsuZsHSK5Wc333yz4LAulcTMtsw2yqzJlq+YmBjRnmWPtCyzzITKDK9atuqvvvpKZBJdtGhRnXZkJmLuj6VTLMESPNzOMj228Nhjj4nP33vvvWoZolmmhyWVLMGMpmzL7MeWuOiii8zlVGxlg7bO0G0NR9uImZqZETU6Orra/jp06CCyf2uZfW2hvtmwCdqO39Eyt7Jv7CM1a0uXzELO7OSWpZbsBTMLs3TPLbfcIvbNEkyzZs0ytWzZ0ma2WWY6ZhZcclEjtKE1WELJntJRzHRMm9aWWdsSLP9TU/ZuZmNmxmJm19WycPMaZFbvoKAgUf7NElopmieffNK87YknnhBtmT1XA88x9aCVm2FG5W7duonvsmwQs6uTRysbxLI8PFcsicQyN1OmTBHjBe1an/0R48aNE6VucnNzRUbob775Rmz/+++/RXt+bgnakOMJxyVNJ8xYz2zEvCYswSzn3Mejjz5abfsXX3whtl966aV1ng9qQTvX1qWjOC4uXbpUZCB+5pln7DqHBEs38XOWfrIGx2+W2WLGb+16pV2ZaZnbk5OTzW2TkpLEfmyV5KoL2vjOzOiWYLZtW2OiBp4Pfs5M2dbb63Me+VtAu/JlbVdb0Mpx3XjjjXW25e8hq0ZQ58xcrYE24jhAHWul0FjmjJg3b57Y//Lly4WumDGd2Llzp9gPP2P2bl57HCdZJkqrHEEt8zpgG54nlhRjKambbrpJHB/7UltJpvr+1mplHnmNWeqBePvtt83XqPU5euWVV8Rn7du3N1+rCgrNEcpZVlBoJuBNOB0t3uy0bt1a3FwNHDhQODj8wbZ2tug8denSxXwDzRdLRfEHurYfcL7oBLK0FG+OWWrogQceEOVG7AF/5D/77DNxQ80bjT59+oibFdYXPvvss03ffvtttRsBrURKTS86iJY3FOwHnSAeN0t58GVdWsca5CC/tk/efNM2vJGnPbSbI77YZ1u1QFmvlsdBm/CdNzTWDjq3c9/cDx9s8CEGS7ZYngdupyPKG0zexGjbebNjqyyNXjbSyl7V9KpJJw11lgnWqWZpGvaHNmF9aOubSjo21BxvVC37w4cCdGTtBR9m8BjpbNJx47ngzah1OS1+TttbclEPLIVmeVPPm3eWbqmtHiodU5aO0fbDY+D+//3331Pa0iEnh6YXnkP+bV1vmjphfWbqlzWweT3xwYutUnLcD51IlpShTtn+nHPOEeVirPXEftEJoFPDMk8//vijOD7y8DjoaLAMXG0aueOOO+q1P4Llajh+Uft0dnl98AEgv6vtl7rlwxINP//8syh3w+28xjjuWdaSJji28NrS7K6VT+PDK8351R7UWT5stAQf5vFzrS3PZVRUlHjxwQ2dFe0zlqyq6xzyGKhl7TssE8TzZ+0s0mGmw02b8Nh4jVx33XXVHCMeL+v/avviAxnqwl6whjPtwAdnluDDK9qNDzgsQRtpDzz4ogNoOSbYOo98gECnUvsOr3E+2OADBu0BEl8sd1fTw00eJx1GS52Rg46fLVB7/F3T2lL/5NPA0lPUFq8FPvTUfntoW2qJNuW1rT1sJHhtsV42P+ODQz5I27RpUzVefp+a41hB27BUFh1x/i7TLrxGbdU+ru9vLWuqa+UjtbGJdbttPYyhPvlQiw/XeT1ajqHUsPXDJQWF5gIX/tPUs9sKCgoKCgoKTQuut2SI9H///eeQ/TFM/ttvvxVrXS1DwJks6o477hDrKhnarqCgoKCgcLpCrVlWUFBQUFBQcChYsujiiy8+ZR0y11SyjBjLljHzuoKCgoKCwukM5SwrKCgoKCg0c2hBZo5IjEQwOzrLI9WUxInldRxRHkpBQUFBQUFPKGdZQUFBQUGhmWPv3r3mTObMrtxYsAYvZ46ZAdi67uzKlStFXd5777230TwKCgoKCgp6olk5y3PmzKl3SRoFBQUFBQUj480338TAgQPNtatbtWplrtfaULB+8dKlS0V5MNZnZikblmm66aabBAdLEDmilrKCgoKCgoKeaDYJvvijzdqidJhZjF1BQUFBQUFBQUFBQUFBwalnlg8ePIjXXnsNU6ZMEQXUi4uLq31Of//JJ5/EoEGDMHjwYJx99tlITk42f378+HERBsYsnwoKCgoKCgoKCgoKCgoKdcEdToA9e/agV69e+O2331BYWHjK53PnzsXixYuxZs0a+Pr64sYbb8SECROwYcMGuLu74+WXX8YTTzyBa665pk6ubt26oV27dvXuY2JiIuLi4iADsriMxsOHLNu2bUOPHj3g7e2tK5fSg7JdU+hB6U7ZTunOua4ldc0q2yndOde1pK5ZY9uOk6s7duyovtHkRHjkkUdEcfSioiLztpycHJOfn5/pww8/NG9LSEgQ7T755BPTu+++ay4WP2fOHNNzzz1XK8e0adMa1LepU6c26HunM5fReDZs2CB0wXe9ofSgbNcUelC6U7ZTunOua0lds8p2SnfOdS2pa9bYtrPlBzrFzHJt+PPPP1FQUCBCsDXwSUKLFi2wbNkyHDlyBC+88II5HJsZOFn/8dlnn3VoP2SWwKipFIejIeuYjFg+ROnBOWxnNI3LGhuMaDsj6s5ovxUyuYzGY0Q9KN01HOq3QumuKfQww0l/K5zeWd63b594j42Nrbadf+/fv1+EYmu47LLLREbO2hJ8JSQkYNKkSeL/DOGeNWuWePH/DPHWSmr4+PiI96KiIvHONvysoqICbm5u8Pf3FzUmCYb8uri4mNsyOyjDySlOtg0ICBDZQbW2rq6u5nBzfsbw4bKyMrGd342MjERWVha8vLxEv/iwgCBnaWmpeJEvODhYtCPYlmU88vPzT2lLhISEiD5w/TczlPI1btw48X2uE2dfS0pKzG15bKzHyX2yz3l5eeIztqUNtHXl7APtwrbWNuT/uV3jCQoKEv3j9+uyt7UNre1taUOtrbYvHqNmF7bl9zUbWtubfbK0ob32njx5stiHpb35t6UNre2ttbW2t6UNre1NG7G9xsu2/KwmG5JPOzeW9ra2oS1787ogT132tqVZS33XZG/NhuPHj7db3/ba0NLeljak9ri/mjRry97cbsuGtWmWPPxeTfZ21BgRFhYmjlFrq+cYwR8i7bt6jhFTp0418+g5Rmj6ro9mGzpGUA9sY6lZPcYI/lawjaVm9RgjeDxsY6+9GzNGzJ49u172bugYMXPmzGr61muMqEnfeowR1AP3Z2lvPcYIW/rWY4wgj7W+9RojqAf2j231HCMs9W2tWUeOEeyDxmOtWUePEdo9Xn3s3ZAxgjxsY61ZR48R5GEbezXbmDFi2rRpop2mWb3GCPJyv7bs7eXAMYL6ro+9GzNGaLqry97fffcdFixYID6jnTixegpMTh6GPW/ePLGtsLCwWtvhw4ebOnbsWG2bnmHYJSUlJllgeLkMyDomWTwZGRmmjz76SLzrDaUH57Cd0TQua2wwou2MqDuj/VbI5DIajxH1oHTXcKjfCqW7ptBDiRP8VhgyDDs8PFy880mE9hSL4FMMPkW1xPz583XrB59QyAKzfcuArGOSxRMaGorzzjuvmk70gtKDc9jOaBqXNTYY0XZG1J3RfitkchmNx4h6ULprONRvhdJdU+ihwkl/K5zeWW7durV4T0pKEtPwGvj3qFGj6r0/LQyboQJXXHGF3eE8nOJn+IKMMGyuuea73mHYGo/eYdgaj95h2CwntmjRIpEVnf3WMwyb/dTCWvQOw+Zx6xliqdkwMzNT7E/vMOy0tDSxXUYYNm0TExOjexg2t8XHx+sehs39M2eDjDBs8mt20HOMsNS3nmOEpm8ZYdjsMx/06h2GzW3R0dG6h2Hz+9SdjDBsy/7pOUaQw1Lfeo0RNelbjzEiIyND9F3vMGz22VrfeowRtvSt1xjB/bGd3mHYlvrWMwzbUt96h2Fr93h6h2FT3+yP3mHY7DOXPMkIw9b6qXcYNvfHl95h2ATtKSMMW9OdvfZesmQJli5dajMM24XTy3ASPProo3jssceEELTSPzQqB0sm8aJzSxw7dkz8cPOg67vAe/r06SJ+vb6gUHhiZUAWl9F4Nm7cKGptcx173759deVSelC2awo9KN0p2yndOde1pK5ZZTulO+e6ltQ1a2zb2fIDXeFEOHjwoLkGlgY+xbjuuuvw0ksvmZ96zJs3D507dxZJlmTBclbbKFxG45EJpQdlu6bQg9Kdsp3SnXNdS+qaVbZTunOua0lds83Pdk4Rhv3444/j448/NjvLnB1kVuuFCxeiffv2eOKJJ8T2oUOHiql6rlX+5ZdfzKG2MsKwCYY5yAjDZqFt7l/vMGyGLbOd3mHY6enpgsdI2bBpK4bpyAjD5n74fb3DsFNSUsR2vcOw66PvxoZhs31UVJTuYdhsw3J2eodhcxsrAcgIwyYn+yEjDFsby/UcIxiypW3XOwybx8AcCnqHYbM9fw/1DsNmG+pORhg2Xzwee+3d0DGC2y31rdcYwf/b0rceY4SWkVjvMGzawlrfeowRtvSt1xih6ULvMGxLfesZhq3x2dKso8cILq3idr3DsGlj/l/vMGzahv+XEYZNfh6T3mHY2vHrHYbt6uoquGSEYXM5LvmbXRi2DKgwbOOGtKgw7MbBaHqQyWU0HplcRuORyWU0HplcRuORyWU0HplcRuORyWU0HplcRuNp1mHYpzP4dMJoXEbj4VO1MWPGiHe9ofSgbNcUelC6U7ZTunOua0lds8p2SnfOdS2pa7b52c4pwrCdAZzGNxqX0Xg6dOggwvMZjqI3lB6U7ZpCD0p3ynZKd851LalrVtlO6c65riV1zTY/2yln2YGloxhTL2vNsozSUYz3N1LpKPafXB07djT3V8/SUREREVLWLJOTfdN7zTLXsMsoHcVs9qp0lHOUjtKe3Oo5RljqW88xQtO3Kh11+peO0qB36Sjtwaqea5bJaUvfqnTU6Vc6iudJRukoDXqXjtL0rfeaZeY7UaWjGl46ip/LKB3FbTJKR7mc0Lnea5Y1X0mtWdYBas2ycdcwqDXLjYPR9CCTy2g8MrmMxiOTy2g8MrmMxiOTy2g8MrmMxiOTy2g8MrmMxkOoNcunIZw1tKA58ciE0oOyXVPoQelO2U7pzrmuJXXNKtsp3TnXtaSu2eZnO5Xgy0HQUv0bictoPDKh9KBs1xR6ULpTtlO6c65rSV2zynZKd851LalrtvnZTq1ZduCaZcb6y1izrK3x1XvNssZjlDXLMusss59sK2PNMo9bzxqqmg0zMzOlrFlmTUa1Ztk51ixrWtK7zrLe6xEt9a3WLJ/+a5a1/um9ZtlS33qtWa5J32rN8um3ZpntZKxZ1o5F7zXLGo/ea5a1ezy96yxnZGSI/uhdZ5l9DgsLk1JnWeunjDXLfDV0zXJmZibySyuRW+6K7FITUrLzkVtSgTIXd2QWlCCrsBR5JRXw8XDFQyPjpKxZ1nSn1izrALVm2bhrGNSa5cbBaHqQyWU0HplcRuORyWU0HplcRuORyWU0HplcRuORyWU0HplcljwVlSZkFJYhJb8UafmlyC4uR1ZRGbKLqt6zisrN2zJPtOOrqMz2TK6/pxsCvd0R5O2O+AB3/Hz1AN2Px/qYGusHupgsU+8pNNhZ5tMMPo2RAVlcRuPhEzM+/QoNDRVPqfSE0oOyXVPoQelO2U7pzrmuJXXNKtsp3TnXteRM12xZRSXSCsqQnFeC9IIys7Nr+U6Hl+8ZheVVDnJBKSpteIYBXm4I8fFAiI+7eA/2dkeorwci/T0RdeLF/0f4eSDYxwOBXm4I8HKHm6uLw46nPmgoly0/UIVhOwgMI+B0v5G4jMZDB5lhQXo7yoTSg7JdU+hB6U7ZTunOua4ldc0q2yndOde11FTXLB3f1PxSJOeViplf4ehaOL6a08v/0+Gl40sH2Rr0XS2dXvHu5YoukcFVTm/ASec3ws9TfM5ZYXc312Y73iln2UFg7LzRuIzGc+DAAdx00014/fXX0a5dO125lB6U7ZpCD0p3ynZKd851LalrVtlO6a758TDUOf2EQ1vl+J50djUHmH+n5hYho7gSyTU4vpy1pTMbesLp5UxvbKAXekT7I9rC6Y0O8EK4rwdCfD0Q4OkGV4vZXpnh3hVO6ispZ9lBCb441c/PZCT40j7TO8GXxqN3gi+NR+8EXzy3y5cvF8epd4Ivfl9LuKR3gi+21Xj1TPDF/pBH7wRf9dF3Q5P3aDYkl57JezQbkkfv5D20Ib+v2UXvMYLQvqvnGMG2Go+eY4SmbxkJvsirZ/IeS33rnbxH07feyXu0MYJ91jN5jzZGWOtbrzGiJn3rMUbw74Yk76nvGGFL33qMEbb0rdcYQVuwf3on+LLUt55jBLk0Hr3HCOt7Mb3GCH7XlmYdPUaQx1Kz/G6ZmxeO5xYjObtQrOctdvFCUlaecHhzSk3IKalEal4x0gvLkS5mfstshjoHebkJhzfQ0xXB3m7i1SXKH2FeLoj080DbqCAEuFYgxMsFYX5eiA4NNJ+b2scI2sm7KkFZkW1987gbmuAry84xgvupj70bM0Zox6YSfJ1Ga5a1gUIGZHEZjUdmgi+lB2W7ptCD0p2yndKdc11L6ppVtlO6O72updLySqQXcg1vGTIKqmZ+Odsr/i4sQ3p+ifg75URINGd9S8pPTW7l5+mGMF8PhJ6Y8eX/w/08ERNwcraX76KNr4cIdbZc3+vIY6oLRuNpDJdas6wj+ERDVnY8WVxG45EJpQdlu6bQg9Kdsp3SnXNdS+qaVbZTutP3WiouqzA7utYvhjbznQmwknJLcPzEemBruJxY50vHNsjTBZEB3uga6Ycz24ci2t8TMYFeiPTzRJhfVRu29XJv/Bpfo41DuU7qK6kwbAUFBQUFBQUFBQWF0x7lFZXmmV6u+9Uc38SMXORVpAmHl45vlfNbIur7WoMTuNpsL190eEe1DRFOb0yAl8jqrH0WZjXrK7N0lMLpAeUsOwja2gwjcRmNp0WLFnj66afFu95QelC2awo9KN0p2yndOde1pK5ZZbvmrjuuX80vrRCzvJrzyzDoFM3pzSvBsZwqx5drfXOKy23uh+t7I/y9EBvgKZJc9Y0NODHjWxX+bO38Wie5ctTxOBJG04OPk9pOOcsKzQZRUVG49dZbRZIABQUFBQUFBQUFx0Nb90vn91hWAbJLIZxdLdT5eC4d35OOcVmFyeaa3xaBXsLxjQ/2xsD4QPOMLzM7h1k4wFwXXF5Wqu7vFHSBcpYdlA2bGdiY7U1GNuyUlBTxrnc2bI1H72zYqampgkfvbNjp6ekiG/Z5550n7KNnNmz2MyIiQko2bO6HfdM7G3ZGRobYn97ZsKkHbpeRDZu2iYmJ0T0bNrfFx8frng2b+4+Li5OSDZv8Wn/1HCMs9a3nGKHpW0Y2bPY5PDxc92zY3BYdHa17Nmx+n7qTkQ2b/dPOsZ5jBDkt9a3XGFGTvvUYI6hx9l3vbNjss7W+9RgjbOlbrzGC+6MG9c6GbalvPbNhW+rb3mzYpRWVKKh0F87ukYx8JOWVIrPMFQlZBVVrf1n+qKhSOMG5NkKf6fxG+3sg2s9dlDXqFhUCP9cKhPm6IyrQR8z+8m86vq0iguFaWVVVpHZ7lyLA2wvFRYVC37S13tmwqZGwsDC7NduYMUK7trRzpdcYwf2xnd7ZsAket4xs2JoPo7Jhn0bZsGWuYZDFZTQemdmwlR6U7ZpCD0p3ynZKd851LalrVtmuKXVXUFqBYznFOJZbgsScEpHhmc5uGuv6FpaJsGftb2sH2N3VRcz6MrtzhJ8nwv08xHvEiXftb8+KIrSPDUeAl77zc+qaVbZzhB5UNmwdwacXRuMyGo9MKD0o2zWFHpTulO2U7pzrWlLXrLKdI/XAGTvW9rV0fFPF+wmn98S7ts16/W+gl5uY4dUc3i4RvhjeOtj8dwRLHvl7okVQVfZne9b9VlT4SykXpK5ZZTu99KDCsB0EhjIwxMFIXEbjkQmlB2W7ptCD0p2yndKdc11L6ppVtqtLD3SAs4vKkXrC2aWjW+XsVr2nnthWNTtcjKKy6jV/fT1czY4uk111CPNF/xhftAjxEzPDcYFeaBHkLdYH+3o63qk12rWkrtnmZzvlLDsIjLM3GpfReGRC6UHZrin0oHSnbKd051zXkrpmm6/tKipNIsGVpfN7OC0HOeVuOJpdjIQcvqocYOsEWJzQpQPMhFec9eUa4H4tAhAf5I24IC/EBXmLbWxjywGWGf5vtGvJ2XXXHHgczaWcZQdBRoiJbC6j8TC5R8+ePaWkrld6ULZrCj0o3SnbKd0517Wkrlnj2a6wtEKEQPOVzLJHJ9YDJ2RXrQ2umg2uygRtskoC7eXmgqgAL7QM9kZ8kBeGtAyq5vhGnpgdDvHxaHDpo/oeT2NhtGtJ2a752c7FxPgOhUYn+GIGNmaQkwFZXEbjkcmljknZTunOua4ldc0q2yndOde1JPuaZXQzZ3yPZBXhSHaxKIOUnFdazTHme55VIiwPNxcR4sxZX671jfKv7vhGWvzt5+EixaFQ452yXVPoodIJ7vdt+YHKWbZCnz59RCmChpSOYmpzGaWjjh49KqV0VFJSkqFKR7GtdWkdI5SOIqeM0lHJyclSSkdR36p0lHOUjuK+bWnWkWOEpb71HCM0favSUad/6SgNepeO0pwWVTrKuKWjqJ3k3GIkF1Ygr8IdR9JykFZYhsxiE1ILy3A8pxhpheVVibEKT4Z1svpkhK87Inw9EBNY5QCHeLmIba3CAxDiCYR6Vc0St4sOQ06OfWOypb71Lh2l6dve0lENvY/QSvjUZ0xuyBhh1NJR/FxG6Shuk1E6yuWEzvUuHZWYmNjg0lGsmmMJ5SxbQZWOMm4a/k2bNmHw4MFYu3ateCiiJ1Q5EGW7ptCD0p2yndKdc11L6prVz3bFZRXmhFhaMizO/B7OKsbhrCLze6FVQqxgb3dEBXDW11O80xEOcKtAx+gQtAr2RusQrgv2hqe742fIlO6U7YyshywnKLOrSkfpCD7xMBqX0Xj4JItPt2SsPFB6ULZrCj0o3SnbKd0517WkrtmGJ8dKL3HB9kNZwuk9lFlkdoCP5hQLx9g6HJoI8HITDm+bUB+MbR8qHF/+n84vnWKGQnvZcII5Wycj34nSnbKdkfXg7aS+kkrw5SDIisGXyWU0HplQelC2awo9KN0p2yndOde1pK7Z6iivqBQZoOn40gHmGuEUG3WCWUu40uK5N53cNiGc9fVB/7hAs+MbYbEmmP/39mjYemCj6UHpTtlO6c5+KGfZQWDcv7auzihcRuORCaUHZbum0IPSnbKd0p1zXUvN5ZqlE5yUV70+sGXNYCbMonOcmFsiZo01xJwIg2aNYM7+9okNMDu+wW7l6NEyQswU61EfuK5jUjzKdkoPzWO8U86ygoKCgoKCgoJCg6GtD2aG6G0JmUgtzRbO78HMIhxiaHR2cTUnmAj0chNOMNcGs1TSoPhAERLN2WHOEnNbbTPBVWsS/dRZU1BQ0BXKWXYQmHHNaFxG4+nSpQu2bNmCDh066M6l9KBs1xR6ULpTtlO6c65ryRmu2fyScuH08nUgo+o9gVmiLZJnWa8PDvX1EA4vnV+GRbcJ8UGrkJPrgsN9Gx4S3djjOZ25jMYjk8toPDK5jMbjaC7lLDtwup9p2Y3EZTQeJudo06aNlCQdSg/Kdk2hB6U7ZTulO+e6lpryms0tLjev/80sKhfvx/NKcCynBMdyS0RoNB1jOsQa/Dzd0DbUR8z6donwxcg2weawaDrB0QFeCPcoR1yE/hlv1XinbNcUelC6a362U86yg8DaYEbjMhrPkSNH8OCDD2Lu3Llo1aqVrlxKD8p2TaEHpTtlO6U757qW9OJh1Qc6v1wnzNBovvYlZ+NYoQl70wuxL70QGYVlp3zP18NVrA1uEeiFjuG+mNQxDO3CfISDzBcdYtZJrQ1azVW9ocY7Zbum0IPSXfOznXKWHQSVWfD0tx0L1i9YsAC333677s6y0oOyXVPoQelO2U7prnnwlFVUmsOiD2QUiv8zOVaSxau0ovoa4XBfd7QP90OHMB9M7hSG9mG+IiyaIdN8hfi4w9/TrU5nWK9jOl15ZHIZjUcml9F4ZHIZjcfRXMpZbuZx+M2JRyaUHpTtmkIPSnfKdkp3zs/DzNEMi04vKBWzvyydlJzHVwn2ZRRhR0q+mB0uO+EMe7q5iLXBDI3uEMbQ6BDEBnpVe0X7e8LDzaXRjnBDj8mZeWRyGY1HJpfReGRyGY2HUGuWdURCQgImTZqE2bNn44orrkB+fj4qKirg7u4OX19f5ObminbaulcWqicqKyvh4eEh2rq5ucHf3x85OTnmwtj8gdLa8gQylp4hAmzLmPrs7GxzWz4N4ecEPysuLkZZWZnYzu8ePXpUvDMlOvtVUFAg2pKztLRUvMgXHBxsDodiW/aPx2PdlggJCRF9YOiWp6eneCUlJQkePz8/0deSkhJzWx6bdszsc15enviMbWkD9plgH2gztrW2If/P7ampqYInKCjIbntb29Da3pY21Npq++IxanZhW35fs6G1vdknSxvaa2/2MyIiopq9uU9LG1rbW2trbW9LG1rbmzYip5Yen235WU02JJ92biztbW1DW/ZOTk4W+6vL3rY0a6nvmuyt2ZDXILfbY297bWhpb0sb0jYxMTE1ataWvbndlg1r0yy3xcfH12hvR40R3H9cXJy5rZ5jBPm575rs7agxwlLfeo4Rmr7ro9mGjhHsc3h4eDXN6jFGcFt0dHQ1zeoxRvD71J299m7MGMH+adBzjCAn+1GXZhs6RpAno8wN/x1KxYGcCuzLLMbhnHLsTitAasGp4dEeri6I9PdA6yBPDIr1xbUDY9Em0A2tAj1E2HRwUHUbnrR3JQL83VFcXIjjGRmi75b21mOM4Hm31rceY4Qtfes1RnB/PH621XOMsNS3tWYdOUZY6ttas44eI1JSUoS96mPvhowRjB5kf6w16+gxgn0OCwuzW7ONuY9gP/m5ptna7N2Y+wjuj9ts2dvLgWMEoencHns3ZoxITEwUx2GvvZcsWYKlS5fi+PHj5r6a+2yyvDIVMH36dHz33Xf1tkRVCQP9E1rI5DIaz8aNG9GvXz9s2LABffv21ZVL6UHZrin0oHSnbKd01/TXUkl5pVgTTMf3eG4J0gvLkM7M0SfqC/PFBFpaBmlvd1d0ivBFpwg/dI7wFTPEzBYd5leVNZqh0sE+7o2eFTbaOKTGO2U7pTvnupaynMBXsuUHqjBsB4FPPIzGZTSeqKgo3HXXXeJdbyg9KNs1hR6U7pTtlO7kXEuFpRUiPJqvI1kMjS7A9pR8bE8pEOuHtZrCDJFmtuhwPw+E+3mKWeB+LQIRHeCJLhF+aBfsjk7RwXBz1T882mjjkBrvlO2U7pzrWvJ2Ul9JOcsOghbOYiQuo/G0aNEC8+bNE2EfekPpQdmuKfSgdKdsp3TnGBSWm7D1aA52phaI1560ArFumLPEafmlyC+tXlc4JsAT3aL8ReKsrpF+6BLpJ2aKWVKpthlhhifKcJSNOA6p8U7ZTunOua4lNyf1lZSz7CAwll+GEyaTy2g8XOuwatUqjBo1Svc6b0oPynZNoQelO2U7pTv7wqQZCs0aw3zXZohT8kpxILMIW5PzcSizat0f/dzWIT4iPLpXjL95lpgllKJO1BVmmSVmk1bXrNxxSI13ynZNoQelu+ZnO+UsKzQb7Nu3T6xFkLFmWUFBQUGhacHySnvSCoXzuy05X2SRZpi05ghbItjbHVEBnmgV7I2zukWgXaArBrWNROcIP/h6ypsNUVBQUFA4vaCcZQdB75nKpuAyGo9MKD0o2zWFHpTulO2ai+64Jnj/ifrCnBFOFjPDJeYyS3zn51qtYc7+9oj2x+zukegS4YuYQC9E+nkiwt8TkX4e8Pao7hAzK6uW7V3G8ciA0fSgbKdsp3TnXNdSgJOOd8pZdhCYkpwp0o3EZTQemVB6ULZrCj0o3SnbGU13LNiRXVSOHakF2JyUh/VHM7EzvQTbk/NRXF5pbhfi4y7CosUrwAvdo/1x7aAW6BUTgJ7R/gipZ5i0EWzXVFxG45HJZTQemVxG45HJZTQeR3MpZ9lBYG0yo3EZjUcmlB6U7ZpCD0p3ynbOprvS8kqRQIuO8NHsYjFDnJxXImaHtf8XlVU5xR5uLugc5o1+8cG4sFeUcITbh/sKB9nL3bXZ2e505TIaj0wuo/HI5DIaj0wuo/E4mks5yw4Ci3gbjctoPCxqHhsbK971htKDsl1T6EHpTtnudNUdk2oxNPpwZhG2HM/D5uP52JSUJ9YRM1SaibREwqwTM8OsOzyybYj4W5RZimQNYj8UFeQhKCjotDgmZ+KRyWU0HplcRuORyWU0HplcRuNxNJeLiTFOCrUWo1ZQUFBQUFCoGZWVJuzLKMTGY3SEq2aJj+eW4PiJ9cM5xeXmtqw9zDDpXtH+6NsiEH1iA9Azxh8BXur5vYKCgoLC6eUHql8mByErKwshISGG4jIaj0wudUzKdkp3znUtqWvWPuSXlJtDpI9kFWHDsTysPZyJbalF5trDzCjdJtQHsYFe6NciUGSZ1maMmWiLs8Yebg176m80PSjdKdsp3TnXtaSu2eZnO+UsKzQbbNu2DRMmTMDPP/+MHj16NHV3FBQUFE5L5JWUY0NiLv5NzBWJtBJySpCYUyxmiQtOOMQa2ob6oEeENx4cE41+LQLETHFDaw4rKCgoKCicbmgWzvKrr76KTz/9FCkpKXj++edx7rnnOpzDy8vL4ftsai6j8XCx//Hjx6UkGFB6ULZrCj0o3Snb2asHhk1nFJbheF5VqDTLLK1PqHKQd6cVgAu0/DzdRCbplsHe6B8XiFjOEAd4iTXEfOfMcZC3OwoLC+Hr69vkx6R4lO2UHk6v60gml9F4ZHIZjcfRXIZ3lrOzs8Us4vr16/HJJ5/gpZde0sVZllGPUTaX0XhkQulB2a4p9KB0p2xnrYfMwjL8m5CDdQm5Iut0QnYxEnKKRSh1eeXJlCXMNM3s0qPahuDuEa0wMD5QJNRyc3VRulPX7Gk5DqnxTtlO6c65riV3J/WVnMJLOXjwIH744QcsX74cq1atQkZGBry9vc2fM0fZU089hW+//RYuLi6Ii4vD66+/jujoaAQHB2P06NGi3aFDh3DXXXfp0seCggJ4enrqsu+m4jIaj0woPSjbNYUelO6ar+24ljjpREItZpimc/zP4UzszywRn4f5eohZ4o7hvjizfaiYFY45MUPM95gAL3g2sPySs9uuOfDI5DIaj0wuo/HI5DIaj0wuo/E4mssuZ3nv3r3CUeWaz9TUVHOh55YtW6JPnz6YPHkywsPDoRf27NmDXr164bfffhPhXtaYO3cuFi9ejDVr1ohQsBtvvFGsTd2wYYN4skCD3XrrrcLRfvfdd3Xrp4KCgoKCgl4oLK3Af4m5WJuQgz1phTjCWeLsYuEka8m1CHdXF/SODcCo1oF4ZFwkBscHoV2Yj3iYrKCgoKCgoOAgZ3n16tW4//77hROqVZgKDAwUjjKzjBUVFVXtxN0d5513nnBaW7VqBUdj0qRJ4v33338Xs8eWyM3NxTPPPIPXXnvNvGbqf//7H+Lj4/HFF1/gkksugZ+fH95//3389NNPmDlzJo4ePSpmnB0J2kQWZHEZjadDhw5YsWKFeNcbSg/Kdk2hB6U7Y9iOv7d5JRU4llsiZon/PpyNv47kYHNSngid9vd0E3WHW4V4o2d0uJgl1maK+c41xj4ebiI/g4y68qeT7RSPsp3Sw+lzHcnkMhqPTC6j8Tiaq0Zn+Y477sA777yDGTNm4OqrrxahzLGxsdViwOksM0Sa2YU589yzZ0+89dZbuPDCCyELf/75p5g5HjRokHkbw7BbtGiBZcuWCWfZ0unu3Lkz9u/fj/79+zu0H6WlpVJuSmRyGY0nICAAQ4YMEQ9P9IbSg7JdU+hB6c55bMff0iNZxfj7SDb2phfiUGYREnNKhIPMl2XWaZZhGtYqCFf2j8WQlkGiRrE9a4mV7hp3jtQ1q2xnVD2o3wplO6W7RjrLF198MfLz87Fv3z7hINcEHx8fdOvWTbzoXP/333+44oorkJeXh2uvvRYywD4S1v3k33SK+WQ9ISEBbdu2FeHjrq6u6N69e437Y1ttJps3M7NmzRIv/p8z15zJ1o6d0GbXKysrUVFRIV5ubm7iiUZOTo74jOurGf6mteXsPMPJy8vLRVs6cUxEprVlH7Vwc37GfvM4uJ3fTU9PFwMdM71pYeYEObmdL/Jx9pwRAATbcgDmebVuS7AWGfvAGQ3G+POl8dC5ZF9LSkrMbXlsPGbuk33mOSfYljZgnwn2gTZjW2sb8v/crvEEBQWJ/vH7ddnb2obW9ra0odaWEQVvvvmmiDzQHGa25fc1G1rbm32ytKG99mY/aUNLe3Oflja0trfW1trelja0tjdtxP9r55Ft+XdNNiSfdm4s7W1tQ1v2Zn/JU5e9bWnWUt812VuzIXMS2Ktve21oaW9LG9I2/LsmzdqyN7fbsmFtmuU2/l2TvR01RnD/7IPWVs8xgvzaZ3qOEZb61nOM0PRdH83WZ4xgbNaxEg/8sjMJqw9mYkNqKY7lVh1XlJ872oT6okWABzqH+iMmIBQdooMR6FqGNsFeaBnqZ6HvclSUl6HEjjGC29gfS83qMUbw++S0196NGSPYP00Peo4R5LTUt15jRE361mOM4NjK9pb21mOM0Ppcl70bO0bY0rdeYwT3Rx621WuMsNa3tWYdeR9hqW9rzTp6jNDu8epj74aMEdQ321tr1tFjBPvM9vZqtjFjhHbPpWlWrzGC++Mx2bK3lwPHCIIc9tq7MWOEpru67P3dd99hwYIF4jPaiVVzrOFi0uKrT+CFF15AYmKiyBrdEPBAOBv9xBNPYOjQoXAkHn30UTz22GNCCFqCryeffBIPPPCAOOHahUmMGDFClIqiEYYNG4YzzzwTbdq0wZw5c8Tsck2YPn26+E59QVE5OrS7qbmMxrNx40b069dPrGXv27evrlxKD8p2TaEHpbumsx1LMXGGeF96IQ5ylji3RKwl5kvMFucUo7CsUqwn7hXli9EdwjG8dTCGtgpCmJ8+CU+U7pTtjKwHNd4p2yndOde1lO0EvpItP/CUmWU6og11lAk+uViyZAnmz5/vcGfZFrTEYnwSYeks82lCZGQkOnXqhLS0NN37Ievky+QyGo9MKD0o2zWFHpTu5NmuotKELcfzsOpgFlYfysafh7NFmSbCy90VcUFeaHFiLXG/FgHi/31iAzAwPgi+nm6QAaU7ZTsj60GNd8p2SnfOdS0FO6mvdIqz/OCDDzZoRwxfZlItber89ttvhwy0bt1avCclJVUzDP8eNWpUvfenhWHPnj1bhJTbG86jhQnICMNmODHf9Q7Dpg3Jo3cYNjOsk0fvMGxtXzxGzS56hmFHRERICcMmp1Z8Xc8w7OTkZLE/vcOwqW9ulxWGHRMTIyUMm+OjjDBs5myQFYat5bDQc4yw1LeeY4Smb0vNlpRXIqUYSMovx8G0XBzLLUNasQlHsgqx/lg+sosr4O3uigGxfriqTziGtwlBt+gAET7t6uJSo73TjueKB716L9XgNpZQlBGGTd3JCsPWoHcYNvtRl2YbO0aQ05a+9QrDZt9lhGFb61uvMGxrfesZhs3jlxGGrUHvMGxN33qHYTPKk/aSEYbN/sgIww4LC5MWhs3PZYRhc5uMMGyXEzrXOwybUdI8DnvtzYnepUuX2heG3RCQsGPHjjhw4AD0hK0wbBqVgyXDx+ncEseOHRM/3DxohoTXBw0Nw6ZQeGJlQBaX0XhkhmErPSjbNYUelO4aZ7tskzd+25+J3w5kYs2RHBzNKYblL2Skv6eYMY47MUs8pl0oBrUMEjPJp+N5MhqPTC6j8cjkMhqPTC6j8cjkMhqPTC6j8TSGy64wbFugN/7KK69g7dq1wvu29K/prXMm6MiRI9AbzLxN0ClnUjHtKcZ1110nQsdZvopPG+bNmyfWJbP+syzIKrItk8toPHwSyDXrfNcbSg/Kdk2hB6U7+9cXpxaUiuzTu9MKsPJAFn7fn4HD2SVgkul+LQJxTs8odI7wRbtQH1GKieHU3h6OCZ82mh6U7pTtlO6c61pS16yyndKd/bBrZvnSSy8VmcLooXNK23pWmRnHOMPL/+uBxx9/HB9//LHZWaaDzIzWCxcuRPv27cW0/UMPPSTqKHOqnmuV3377bXNYeH3Qp08fMVNd3zBsfqbZQEY2bPZH7zBs/s396x2GzT7xM73DsNmWfKGhobqET1nam99l/2WEYVtee3qGYWv70TsMmzkG7NV3Y8Ow2YZ20DsMmzx8SKN3GLa1vvUcI/h9zb56jhHk577rY2/Nhtl5+fjnSA7WJhbgaH4F9qfliwzUx/PLRL1iDd2j/HBGnC9GtArEsFaBaBmp7xjB42Hf9Q7D1jIf6x2GTR7qTkYYNo9ds4OeYwT7oNlBzzBsHpctfesxRmhVGvQOw+YxWetbjzHClr71uo+wPFd6hmFb6lvPMGxLfesdhs19cT96h2HzvPLY9A7DZp94bDLCsPldbtM7DJt/c596h2H7+/uL/cgIw87MzDT/v75h2IxArbezzLUnixcvrnENMI1E55Xhz84OFYZt3LAMDiCbNm0SD0Qsk8E111CT5s4jk8toPDK56sPDn7NdqQX4dX8mftmXKZJvsV5xqK+HmCVuE1I1SyxCqYO8ER/khVYhPuLz5m47Z+CRyWU0HplcRuORyWU0HplcRuORyWU0niYJw6ZzUVuyLHaGs7oKCqczdu3aJTK0y1izrKCgoA9KyyuxP6MQx/NKcTyvRIRSJ+YU43BWMdYezUFGYRk83VwwrHUwHhrTBuM7hKJXTABcGV+toKCgoKCgoFAP2OUs07HYu3evSOJVE9544w288847cHY0NBu29pmMMGx+T8sQq2cYtsajdxi2xmOkbNj8rpadWO8wbPZf49UzDJsgj95h2PXRtyPCsLk/GWHY2pIDvcOweYwywrDZXvuunmMEj//f/cex8nAu/kwsxKoDWSgoqxSfi+97uyE2wBPxQd64dkA0+kZ5Y1ALP8SEh5htmJ+fV+cYoelbzxBLyzBstpERhs02MsKw2UZGGDb3q2eIpeUYYalvvcYIbtN49A7DZlvuT+8wbFv61isM21rfeoZha9eUnmOEpb71DMO21LfeYdjab7reYdjcr1ZbV88xouo3JV9aGLYWYq7nGMG/uV8ZYdjZ9bB3Y8YITXfSsmEzHfvo0aPxzDPPiA5agl9nwq2bbrrJbPTmGIbNY6fBZUAWl9F4ZGbDVnpQtmsKPRhBd/xNOZRZhDVHc0QY9c9705GQU2qeLR7XIQxntAxCiyAvxAR4OaxmsRFsZ3QemVxG45HJZTQemVxG45HJZTQemVxG42kMV4PDsJkFmzPLU6dOrfHmhk8XmjP4JESWAGRxGY1HJpQelO2aQg/OpLvDmUX4ZV8G9mcU4Wh2sQilPpZbgqTcEpRWVD3D7RrphykdgjG9RyxGtAmBn4McY2e3XXPlkcllNB6ZXEbjkcllNB6ZXEbjkcllNB5Hc9nlLF9//fU488wzMXHixFOyYdNRZpbqp59+2iEdUlDQC3ygo2XxVVBQkIuyikqRmfqH3enitTO1AO6uLmgVwkRb3mgT6iNmjlsEeqFtmC8GxQci3M9TakIQBQUFBQUFBYV6h2GzpvGOHTtqbcOaxj/++COcHQ0tHaX3WiO91xHUZ+2Lo9cs673WyLKtjPWIMtYa1bT2Rc81yzLWI+q91sh6zXJdmnXUmuXmNkZ4+gViZ2KaSL61P6sUaxIL8PuBLOSUVCDK3wPj24fgzNb+GNU6EC2jwtQYocYINUY0szFC3Ueo+wh1H6HuIzxPozGi0aWjLr/8cnz00Ue1ttEW1TfXNcsyj18Wl9F4ZHKpY1K2ay66S80vNc8W700vFOHUWUVVdWMJrjUe3DIIY9qFYkrncPSNrV9maiPbTvEo2yndnf7XkUwuo/HI5DIaj0wuo/E0hqvBa5YvvfRSvPXWWyIcuyYMHjwYu3fvRnOFHc8cnI7LaDwsHXX++edj4cKF6NKli65cSg/Kdkbm2Z1agEUbj2PFoX3452jVzNfg+CCMaBMsQqqZfCv+RB3jFkHe8HJ3bTCX0Wwnk8toPDK5jMYjk8toPDK5jMYjk8toPDK5jMbjaC67nOXvv/8er7zyClauXCnCJqxx9OhR7Nu3D80ZDCUwGpfReBi+tnXrVnMYm55QelC2MwJPfkm5SLrFBFxHsoqxLiEHKw9m4UBGEXzcXTG+YxjeP6sLpnaJQKS/PsforLY7HbiMxiOTy2g8MrmMxiOTy2g8MrmMxiOTy2g8juayy1nmTBw99MWLF9fYprknTXJWATQnHplQelC2czaekvJK/LQnHT/uycC/CTk4mFmEvJIK8+cc4rtF+mFChzBMnBKGka0DEOjrDb3hDLY7XbmMxiOTy2g8MrmMxiOTy2g8MrmMxiOTy2g8TeIsT5gwAW+++aY5YY01jh07hgEDBsAISEhIwKRJk+qd4EtbgC4jMUdSUpJ41zsxh8aj96L71NRUwaN3gi9tXzxGzS56JfhiPyMiIqQk+CIn+6Z3gi+eJ+5P7wRfTK7A7TISfNE2MTExuif44rb4+Hjdk/dw/3FxcXYn7wkIDML3W45i8c5MLNubLRJwdQzzxsAWfrigZyQifd0Q4e2CuEBPdI2PRGF+rtnehQV5yCqp6q+eY4SlvvUcIzR9y0gCyD6Hh4frngSQ25iwUu8kgPw+dScjCaBlaJ2eYwQ52Y+6NNvYMYKctvStxxiRkZEh+q53gi/22VrfeowRtvSt1xjB/fH49U4UaqlvPROFWupb70ShKSkpwl56JwqlvtkfvROFss9hYWFSEoWyn/xc70Sh3B+36Z0EkNB0rncSQM2HaUiCL2vYleCLIdYdOnSotc3PP/8snOrmmuBLZnkTWVxG49m4cSP69esnstz17dtXVy6lB2W7042HM8esYcwEXKxt/M+RbCzbnY7kvFK0C/PBhb2icUHvaHSJ9Gs0lyNhNB6ZXEbjkcllNB6ZXEbjkcllNB6ZXEbjkcllNJ7GcNmV4Gv79u3o3r17tW11OcqEtaO8ZcsW9OrVC80FfMphNC6j8bRp0waff/65eNcbSg/Kdk3JU3oipJrlmtYczcGhrCKkF5RVC6nuHuWP83tG4cLe0egfF1jvpTRGGx/UNatsp3TnXNeSumaV7ZTunOta8nNSX+kUZ/nxxx/HbbfdhjPOOKPBO33hhRfEVH5zcpYZOiArFl8Wl9F4+IRpxowZIuRCbyg9KNs1hR42HcvBwu2ZWLA5WTjHbUN9MLRVEGZ0jUBsoBdaBHohPrgqS7W/l12rcJrN+KCuWWU7pTvnupbUNatsp3TnXNdSuZP6SqfcLb322msYM2YM5s2bh5kzZ9ZrZ4zofvTRR7Fu3Tr8+OOPaE5gjL0MJ0wml9F4uG7mvffew9VXX42oqChduZQelO1k6CGnuBx70wpEhuovt6Zgw7E8kZV6Tt8YXNYvFt2jT61e4CgYbXxQ16yyndKdc11L6ppVtlO6c65rqcRJfaVTnGU6EVzgzCRX8+fPx0033YQRI0bU6p0zLnz58uV4/vnnERkZKRZJc4G4gsLpBCaie+ihhzB58mTdnWUFBUeCa41/3ZeBfxNzcTirGIezikQ5Jy1bta+HK8a2D8VtAyNxXv9W8HBT46+CgoKCgoKCQmNhMw6Pa5SZBInh2BMnThQZyAYPHozY2FiR4ZDZzTIzM0XmuQMHDmDTpk3Ce3/ggQdw1113ObWj3NBs2My4xs9kZMPWsjnrnQ1b49E7G7bGY6Rs2LSLlp1Y72zY7JfGq2c2bLYhj97ZsOuj78Zmw+bf3J/e2bDJw+/VJ4tlbn4Blu/NwPzN6VhxMAdMxchM1e3DfDCiVSCiugSLLNWdY0LQJdQdLpUVYh/uri4N0mx9xwjaWPuunmMEv6vx6DlGcH/kkZENmzZkG72zYfO7bKN3Nmx+h21kZMPWO9OtNkZwX5b61muMqEnfetxH8Ni4P72zYdvStx5jhC196zVG0Ibsn97ZsC3trWc2bEt9650NW/tN1zsbNnnYRsYYwTYysmHze2yndzZs7of71TsbdnA97d2YMULTnZRs2Lt27cJHH32EX375RfyfhtRAY/bp0wfTpk3DZZddhtDQUDTXbNg8oTwBMiCLy2g8MrNhKz0o2zVUDwnZxfh4QxLe+fcYEnNKMCAuENcMbCHWHUf4177+RulO6a4p9KB0p2yndOdc15K6ZpXtlO4akQ3bGl26dMGzzz4rXvTC6aXT6+fTAlnOoTOATy2MxmU0HplQelC2q00P5RWV2JVWiCNZRUjIKRHvDK/enpKPHSkF8PFwFaWcrh8ch35xgUp36po9rcchNd4p2yndOde1pK5ZZTulO/tRr3SonMZnGLbCqWCIgNG4jMbDhzsMsZfxkEfpQdnOWg+VlSb8si8Db687ht/2ZyK/tGq9McOm44K80CrYG0NaBuGhMW0woWMYgn3qf10o3SndNYUelO6U7ZTunOtaUtessp3Snf1oXO0QBTMYS280LqPxtGvXToRWcI2C3lB6ULbTkF9Sjg82p+ONtVuxL70QvWP8cf/o1hjWOliUdooO8IKba/1qHCvdOR7qmlW2awo9KN0p2yndOde1pK7Z5mc7583EdZpBW3RuJC6j8XC9/aFDh6qtu9cLSg/N23YFpRX442AWbv9+D+Kf+gt3/rAffWID8Nd1/bHxlkH43+g2GN4mBC2CvB3mKBvFdkbnkcllNB6ZXEbjkcllNB6ZXEbjkcllNB6ZXEbjcTSXmllWaDbYtm2btARfCs0DrHP8894MbDiWiz0n1iAfzSlBZmHVA5noAE9cPbAFLukWiB6tVLkyBQUFBQUFBQVngsOc5ZSUlGZdu1ZWkW2ZXEbjkQmlB2PbbmdKPp5adRgLt6SgvNKE1iHe6BThh8Etg3BusDdaBnujS6QfescEwNXVxVwKQW84g+2aO49MLqPxyOQyGo9MLqPxyOQyGo9MLqPxyOQyGo+juRzmLI8dO1bM3Dk7GlpnmbXCeEMso86yVoNW7zrLGo/edZa5L/IYqc4yj4vfl1FnmcemHauedZa5X/LIqKFqr74bW2eZn7Pknb31EXdnV+DZ1Qn4dlcGYgI88PTEdpjYxg+x/u427O2J0tISYUPyREZG6lpDlTbkKyIiQvcaqrQ17aP1Qc8xwlLfeo4Rmr5l1FClnbgfvess83PuR+86y+Sh7mTUWbasK6/nGMF3S33rVWeZ7WzpW48xgvsmp951lsllrW89xghb+tZrjGAfeZx611m21LeedZYt9a13nWX2kceld51lrfax3nWWycX9yKizzD6yb3rXWSavdg+hZ51lb29vaXWWtXtJ3eosX3nlleJA6jOrvHz5ckHcXOss86TwxMqALC6j8ciss6z0YAzbscTTkexi/Lo/E59vTsbqQ9loF+aD/41qjUv6xMDT3bVZXksyuYzGI5PLaDwyuYzGI5PLaDwyuYzGI5PLaDwyuYzG0xguu+ss//vvv9ixY0e9ds6nCwoKCgrOCmaq/nF3OtYm5OBodjESc0pwLLcEFZUmkYRreOtgLL6oB2Z0jYC7m8qNqKCgoKCgoKBgdNicWb7jjjswfvx4TJw40a6dJCUliRk7W1PXzWVmmWaU9cBAFpfReBj5wNALhoUwJEVPKD04j+12pRbgid8PYdHWFHi6uaJ/iwC0C/NFi0AvtArxFvWPB8UHIcS34fUvjXYtyeQyGo9MLqPxyOQyGo9MLqPxyOQyGo9MLqPxyOQyGk9juGz5gTanR6655hr079/f7h3HxsZi/vz5aM5w1nTozYmHDjIvHL0dZULp4fS33fbkfMz+ZBO6v7wW/xzJwZszOiPrkZH46/oB+PjcbnhyYntcOygOEzuFN8pRNuK1JJPLaDwyuYzGI5PLaDwyuYzGI5PLaDwyuYzGI5PLaDxSSkd17ty5XjtJS0tDRkYGmjNkrteWxWU0nn379uHaa6/FO++8gw4dOujKpfRwetqutLwSqw9l4bV/EvDdrnTEB3ri7ZmdcVm/WLvXHzcERruWZHIZjUcml9F4ZHIZjUcml9F4ZHIZjUcml9F4ZHIZjcfRXPXKhs1sZqmpqSKrmCWYYeyVV17BhRdeiOYKZlQzGpfRePiUaeXKlVKebCk9NK3tSsorsfZoDv49sf6Yr8NZxdifUYjCskpR1unDs7tiens/hAUHQW8Y7VqSyWU0HplcRuORyWU0HplcRuORyWU0HplcRuORyWU0Hkdz2bWn3bt3izJK69atq7EN03Y3Zzhr7bDmxCMTSg9NY7vMwjI8t/oI3vv3GDIKy+Dv6SZqIMcHe2NY62Bc2jcGY9qFoHdsgAjJl/WU04jXktGOSdlO2U7pzrmuJXXNKtsp3TnXteTrpPcodjnL119/vUjixfewsDAsXrwYs2bNMtdq++abbzB37lw0Z7BGl6x06LK4jMYjE0oPcm3HEk9vrzuGR349KEKtrx7YApf0iUbPmACRybqpz5PReGRyGY1HJpfReGRyGY1HJpfReGRyGY1HJpfReGRyGY3H0Vx2OcsHDhzAtm3bRCFnrRh0nz59MG7cOPH3bbfdhnvvvRfTpk1zSKcUFBQU7AEf1v1+IAu3fb8HO1ILcHm/WMyb0A7RAV7KgAoKCgoKCgoKCvo7y+3atTM7ysTMmTPx0EMPmZ1lhmDv378fzRk+Pj6G4zIaT3x8PF566SXxrjeUHvSxXXZRGfamF4qXqIu8JwP/JebijFZBWH/jQPSLCzwtz5PReGRyGY1HJpfReGRyGY1HJpfReGRyGY1HJpfReGRyGY3H0Vw26yxb44ILLkD79u3Ro0cPTJ06VcSBjxo1CmPHjsXZZ5+Nv//+G3feeSeys7Ph7OCMeXR0NGbPni3WabMuL9c1cqE4j5vT+pYnoaioSLx7enqKdnyxNBFr+ebk5IjPvL29xfpIrW1gYCAKCwtRXl4u2gYEBJhtx7aurq7ic4KfMYFaWVmZ2M7vpqSkwMvLS7zYr4KCAtGWnEzCxhf5+BAjKytLfMa2DJvn8Vi3JRiqwD5QDjwWvpjhnN9jJAH7WlJSYm7LY2OiN+6TfdaSZrEtbcA+E+wDbca21jbk/7md+yIPH8jYa29rG1rb29KGlm1px4iIiGr25vc1G1rbm32ytKG99uZxkdfS3tynpQ2t7a21tba3pQ2t7U0bacepteVnNdmQfNq5sbS3tQ1t2ZvHxnZ12duWZu2xt2bD5ORks76LKoDPNhzDD/uysTWlCMn5VXologM80TPSB5f3DsfULhGifU02tLS3pQ35eWhoaI2atWVvbrdlw9o0S57IyMga7e2oMcJa33qOEbQD+1eTvR01RljqW88xQtN3fTTb0DGCduJ+LK97PcYIfs79WGpWjzGCPNSdvfZuzBhBG2v9s8fe9trQeozgu6W+9Roj2E67DavN3o4YI7hvclraW48xglzW+tZjjLClb73GCEtN6DlGWOrbWrOOHCMs9W2tWUePEewjj6s+9m7IGMF+8DNrzTp6jCAX92OvZhszRrCP/K6mWb3GCPLyu7bs7eXAMYK8tKe99m7MGMFKTeyPvfZesmQJli5diuPHj2PDhg2ot7N88OBBDB8+XNzEfvjhh5gzZw7+++8/4TBrJ2XKlCmnFHF2RtgqRm0PKBRZcfiyuIzGk5mZia+++grnnHOOcJD0hNJD42wXGBSMD9Yfw0MrDiKtoBSj2oSI2ePu0f7oGO6L9mG+CPRufKZDo2lc6U7ZTunOua4ldc0q2yndOde1pK5ZY9vOlh9o191m27ZtsXfvXuzatQt9+/YV2/r37y/K8Hz66aeIiorCzTffXO8OKSjIxOHDh3HddddhwIABujvLCg3H3oxi3P7Ff1hzNEck6XpifDu0CpEXuqOgoKCgoKCgoKBA2D01wylvOsiWoNPBl5YEjNP2zRWWa7qNwmU0HplQeqg/Cksr8PLfR/H4b4fQKtgbf17XX5R70hNG07jSnbKd0p1zXUvqmlW2U7pzrmtJXbPNz3aujtgJY8EnTpyI5gwt5t5IXEbjkQmlh7pRUFqBHSn5WLojFff8uA9tn/0bD684iGv7R2LzrYN0d5Rlniej8cjkMhqPTC6j8cjkMhqPTC6j8cjkMhqPTC6j8cjkMhqPo7nsmlnmsuZFixZh+fLlYuEzF4BbgjWYua65OYMLxY3GZTQemVB6sO0cf7whCUu2p2LL8XxkFJ4cRyL9PTG7eyTuGt4SIa4l8PFwM9R5MhqPTC6j8cjkMhqPTC6j8cjkMhqPTC6j8cjkMhqPTC6j8Tiayy5n+fbbb8err74qMpGFhYWJ7GHWiZOaO5hpzmhcRuPhUgIuG+C73lB6qI6vt6fi2m92IauoHOM7hOL2YS3RKsRbhFt3CPdFlL+nyKpI5OZWZeiUAaNpXOlO2U7pzrmuJXXNKtsp3TnXtaSu2eZnO7uyYTOB18svvyyyCFs7ygRTj3fu3BlHjx5Fc82GzVB0pluXAVlcRuORyaWOqQr5JeW47fu9+OC/JMzqFoEXp3RE69Dak3Up2yndNYUelO6U7ZTunOtaUtessp3SnXNdS5VOcL9vyw+0ay+dOnUStZZtOcoEa1QtWLAAzRlanTMjcRmNRyaXOiZg7dEc9H3tX3yxJRnvz+6CJRf3rNNRVrZTumsqPahrVtlO6c65riV1zSrbKd0517WU46T3+3Y5ywMHDjQXo64JR44ccVSfFBR0wcaNG0XJKL4r6Iej2cW46dvdGPrWegR7u2PTLYNw5YAW5jBrBQUFBQUFBQUFBWeAXWuW77jjDtxyyy344IMPbH6emJiIu+66C5dccgmaK3x8fAzHZTQemWgOeiivqMS6hFxsOJaLXakF2JNWiD3phUjKLUGIjzuentherE12d6tfGExzsJ2z88jkMhqPTC6j8cjkMhqPTC6j8cjkMhqPTC6j8cjkMhqPo7nscpb79OmD9PR0zJ8/32HECgoKzgk6ya/9k4AX/zqKxJwSeLm7olO4LzpF+GJo61h0ifTD9C7h8Peyu4y7goKCgoKCgoKCwmkHu+5mIyMjMXLkyBqzCHNm+ffff0dzRlFRkcgWbiQuo/HIhFH1kF4CnPXpVjGbfFm/WFw9sAX6tQiARz1nj5uj7YzEI5PLaDwyuYzGI5PLaDwyuYzGI5PLaDwyuYzGI5NLL57K/EyYSovgGhIrluI5q+3scpbPP/98PPDAA3XOPp+uePbZZ7FkyRKkpqbi7bffxoQJE5q6SwoKTod1ifm47Ntt8HRzwZobBmBgfFBTd0lBQUFBQUFBQUFnmCormGIaLu4edbctLULq0mfx+JGj+N0nCi3dXTF39Ey0j2znlOfJrtJRzox169aBhzh48GCsWLFClL/KyMiosf5WQ0tHsfi1rPphsriMxlNcXCwS0bVq1Ur3J1tG0gOvn3f/PYabv9uDQfFBIqt1pL+nbnxGsp1ReWRyGY1HJpfReGRyGY1HJpfReGRyGY1HJpfReGRy1cZjKilAyvcv4+WD+7DWOxy9g0Nx39QrER0aZbt9WQmSPrgZ08rDsNc7BBdEhOOf40exz8ULn/Xph3MGjsfpbLsGl46yB+eddx70wsGDB/Haa69hypQpIhScTo/1zfyTTz6JQYMGCaf47LPPRnJysvhM20YMGTJE1ISmAR0N7lcWZHEZjYcOcmxsrJQQEKPogcm6Llq4Hdd9sxuX9Y7Eb1f11dVRNpLtjMwjk8toPDK5jMYjk8toPDK5jMYjk8toPDK5jMajB5epOB8pXz+DR1+8Ghe/cTfe++kjlFdU1MhDR/nQuzdiXHIhXg3pAv+wFphfWIkBX76LnccOndreZELB4idwfbEPDviG48+Zc/DejCuw8dI7MMVUgLs2/IPiEuc7TzbDsOfOnYvjx4/j9ddfFzHmn3zySa07SUpKwrfffgu9sGfPHvTq1Qu//fabzYNnfxcvXow1a9aIms833nijCLXesGFDtdrQ/HvatGnw9HT8zX55ebnD99nUXEbjOXToEO69914888wzaNOmja5czqiHxJxibE7KE1mtD2QWYW96If44mIVgH3csOK8bJrf2hqe7/sXkndF2zY1HJpfReGRyGY1HJpfReGRyGY1HJpfReGRyGY2nLi5TSSHy/vwMXx3chTTvEEwaPBm9OvarsX1lUR52v3M9Znq0QUJAe3R2qcBnCalY9PEz+HjKFQgICKi+/4pyZH92Py5DBFL8w7F++sXoFhqBhMM7MP7HhZj0/adYe/ZViAmPNX+neOWHePXQAXwf3htfj56GPuFVs89evoH48tJ7sHvPRnh7+cLZzpN7TWt8WVf5oYceQnR0NO655x6kpaWJJwY1Qc8aqpMmTRLvTCJm7ZTn5uYK54czz3SUif/973+Ij4/HF198YS5nxdlkOv+vvvqqLn2UFZIhk8toPFlZWfjqq69w33336e4sO5Meft2XgUd/PYi/j1QVcPfzdEPbUB/xenFKB1zSNwbBPh7iWpMBZ7Jdc+WRyWU0HplcRuORyWU0HplcRuORyWU0HplczsxjKi9Dyebl2Hh0DwqCYzH8jBnw9varkYuJs/Z9cBtmu8Vhh2cUvIsq8MDKFXj8wA48MOnSU/dfUY70BffhfI9WKAmMxOZpF6JTcBh+WLUI5+7ej0u+fx8/zbkXXh4n1yMX/fgKnkjPw9/BnfDb+NnCUSbiW3fDj5POw5DlX2PqV29j1YU3IyAoAiWbfsLK1V/hkdjRuLPnAMxq07FaHzx8A9GqXV/IgiPPk01necGCBSKLGB1lYujQoSKcma+aZpavvfZaNAX+/PNP4dhb9i0uLg4tWrTAsmXLzM4yHX86SdxeGxISEszOOWelZ82aJV78P51xzWHQ6nfRTgSfyPAzLUbe398fOTlVzgfDfrUscERgYKCYIedTD7bld7Ozs81tXV1dzTPo/Ixh52VlZWI7v8v/0/Hz8vIS/eLxE+QsLS0VL/IFBweLdgTbenh4ID8//5S2REhIiOgDH4hw5p0vjYeh7+xrSUmJuS2PrbKyUuyTfc7LyxOfsS1toIXKsw+0C9ta25D/53aNJygoSPSP36/L3tY2tLa3pQ21ttq+eIyaXdiW39dsaG1v9snShvbam3bhPiztzb8tbWhtb62ttb0tbWhtb9qI/dd42Zaf1WRD8mnnxj8gELd+uxNvrU/B4Dh/fHJuV/QJc0NsgIf5wRPtzbCdCs9As93qsrctzVrquyZ7azbksdurb3ttaGlvSxvSptxfTZq1ZW9u19raq1l+hy+trV5jhKZvra2eYwTba9/Vc4zgNo1HzzGCIE99NNvQMYLHyjaWmtVjjODfbGOp2fqMEZb2trahtb7Zxl57N2aMYNv62LsxY4SlvvUaI2rSt15jBPdnaW89xgj231rfeowRtvSt1xjBz9g/ttVzjLDUt7VmHT1GaDzWmnX0GKHd49XH3g0dI9jGWrOOHiN4rGxTm2b9C1OQcWQ3ENMR/hEtax0jAjyAgx/ejitMYfjbJxJIT0Drvc/j3cFjMaxTX9FO0yztnZeVgbzP7sRMj1bI94/AymGT0NXPH8/+9B4eTDiOvK/fwb2jzz1pb94bLH8Zt+eZsC8gCL8Mm4SWXr7iuM7oNR4LS8tw9qGjuGTBs3hr2jVw9/CE1/af8M2mP/BS9FA83ucMnBEebbYpbRgZ2QYLB47E1PV/Ydanz+FDz0IkJOzEZS3HYWhUHO7p0Eu0tx4j2P/62LsxY4Smu7rGZK5Ppt9LUFeMrK4zwRdnaG+77TbhaGpO488//yycUXamJjz99NPCGdUTjz76KB577LFq6cBffvll3H777cIglv0bOHCgMARDr1966SWxbplrlvfu3YvNmzfj3HPPdWiCL/LzxMqALC6j8WzcuBH9+vUTmujbV9+nW6e7HorKKnD+59uwbHc6XpnWCTcOiaszOsRoepDJZTQemVxG45HJZTQemVxG45HJZTQemVxG45HJpRcPZ31NpYVw9Q2qk4czuMmLHsbdSZlYFNAK7iYTLg70wouzb0SAl3eNybDGVkYhyzcM742eiojCTNyy8jvsdPPFVwNHYlKfoSfb86H7wgdxeUoelge2wvqzLkOXkHDzZ49+9jQeL3LDa61b4KbxF4vtRT+/ibf/W4k7Ivpj/qjJmNOxxyn9+HDZh7jqWCpurMzAPN9KrDu4DWfHnYkzW3XEknEza7w//GXPRpy9+hc6kyiEq5h9XjHlfET4+Dqt7mz5gafMLHN9smVSLOKnn36qs9zSgAED0BTQnkLx6ZEl6EynpKRg4cKFuOuuu8RTCYJPN7Zu3dokfVVQOB1QUFqB6R9vxpqjOfh+Tm9M7lw10CooKCgoKCgoKFQ5n7k/v4lXt/6LpT7RCPf0wl1DxqN/y+42zcO5x9zFc3F+Wgk2BbfDvJ79UXR0O55Lz8X2BS/i5wtvQ6CFE8n95335KK4p9kVSYDDWzrjY7Pj+HtkCkxa+gfPWrcLvgQHo366n2F60/DW8ceQQloT1xqJRU8ztCRdXVzxy4b1IXfA0bj2UiOAFj2Mm8vHLge24L3Ykruvc26ajTMwaOgs5W1bhjr178V1xKZJanIlBkS2Ec+1Sy0TK+E59sTmmLT7fvxNBnl64olMP+HnomwS2KXCKs8zZ2NWrV5vDJ4h///23zh3dfPPN2LlzJ2QjPLxKKJy2t+wzneLIyEhRI5ove6GFYc+ePRtXXHGF3eE8DCWQFYbNY5URhq3x6B2GrfHoHYbNffDBCZcX6B2GrYWXywjDJuwNw04vKMU5X+zE9tRC/HhZL/SN8LAZWm3L3tynjDDs+ui7sWHY/FxGGLYWKqh3iCVfssKwLcNH9RwjLPWt5xih6VtGGDbtJCMMm5/LCMMmh6wwbMvwUT3HCGt96zVGsJ+ywrB5rDLCsG3pW48xwpa+9Roj2EcZYdiW+tYzDNtS33qHYWu/6XqHYZOn1jDsff9g09bf8VuFO1pEtcFZQ89CdESkbRuaTChf9S4uOJqCVSHdMDXQDwnZaRi/7h/MTTiEW0eddYpmi396FfceTcA/QR2xYvK56OMXjMoOvTFi8wrM2LMXUxY8jy+nX4XQ4DBhr5IfX8aLCYn4MbQ7Fg4Zh2i4if2JEGJXX3w24QLM+mURJq74Fj/sWoOOWYew8uBWPBQzEtd36IGJES3EebMeI56Yeg1Sli/AJQUu6FtRhq0xIzEqOh4vDRlT7d7X0t600dX9x6Fzi874JuEgukZE4tyY1qgoKERheUWtmm0bHIwb23QRn7mUlaMMLjWOEeSVFYat6c7eMXnJkiVYunSpfWHYt956qxBn7969zaHOl19+OebPn28zwRc7wJnaV155RZeSTHWFYf/yyy9i1nvHjh3o2rWruW3Lli0xatSoOjN5OyoMmyfFenZbL8jiMhpPcz+m1QezcPnincgpLseyy3pjcMsgXXiMaDvFo2yn9ND015FMLqPxyOQyGo9MLqPx6MVVmZOC/B2rcbjSBR36jIWPX3CNPPRdCn98Bfft2IrXgzshBJXINrmgg0s5fp59FVqHVeVnskThindx1dZN+DqwDZZNOhdj41qjsrQIt3/6FF6t8MMLbVrijnEXmNsX//EJ3v/zO9wcOQBvDB2HG7pVX+q3cs0yTN6yFWNdS/DlmGkwbf8Vy7b8hfNjhuOhvmfgsf7DbR7nsaTDGPfDZ0grL8fIsmx87xODM+Pa4LsJs+HuWnN1kkqTCR/t2Ypfjx3BwIgY3NS9Lzxca054pXRXzzBsZpZm6SXOFGsJY/jkYM6cOagNembDrg0jR44UTxTWrl1rdpaPHTsmZog5OywLfDoja+CRxWU0Hj454sOV8ePHmxP6GFUPBzIKsTOlAEdzirEvvVBkuv4vMRdDWgZhxZV90Das/qn7jaYHmVxG45HJZTQemVxG45HJZTQemVxG45HJZTQee7m4dhdu7nCpxaHTULJhGT79+RPcHdID2W6e8N++B/M6dsLFvcfY5Cn58zM8tn0jXg/phhcGj8ZtPQZgy5aVOOufVRi6+H2smnEJOkS3Mrcv/mcR3lz/G74I74tPR00RjjLh6umDl+Y8BMx/AnceOgrfn+bj2vEXo/iPj/H76q9wZ+xoXNW5J67v2ueUPoweMhUL4YLZW7dh3A9foF9ZDt6OHY4p8e3wcN+Ta5Kt4esThFUX34b7163E1qwM3BPXBg/1HVqroyz66uKCKzv3Ei970Fx1Zy9OmVnWkJmZKeob82nDmDFjsHLlSps74Mzyvn37xIy0FtqiFy699FJ8+umn2L59O7p162bezlJRTEhGh5lT8zfccIPoL2e8tbXK9qJPnz4iTLe+YdhamICMMOyjR4+Kd73DsJnlnDx6h2GnpqYKHr3DsJnYa/To0fjvv//Qtm1bXcOw2c+IiAgpYdhayDKxMaMSdy3bg83JJ0LH3FzQKsgTXSN8cEGvaEzrHIrSE/u1N8RSs2FycrI5+7aeYdjUN7fLCMOmXWNiYnQPw+Y2lrPTOwyb+2c1ABlh2OTX6tjrOUZY6lvPMULTt4wwbPaZS4j0DsPmNv6e6R2Gze9TdzLCsC1vWfQcI8iplR7RMwybnLb0rccYkZGRIfqudxg2+2ytbz3GCFv61muM4P54/HqHYVvqW88wbEt96x2GzfxBtJdNex/chiMr3sLzjJL2DsaA4DDcMeZ8tIyMsT1G7PoDn//+KW6KHIjZLVrh4th4LPtvOT4w+ePa4EA8Ne58YRfN3u77/8aPP74tZnAf6N4fjwwYYbZ38tHtmP73ChS7uuPHUdPQNr4jXLb8iL9+fh9TW4zG1Z374Ln+w0+xd3ZmBub+8QU+LHXHw1nbEVNehLujBqJvVBwWnjEBXm5uNY4Rq48fwUNb1iCppATntuqIR3oORERIaI2aZT9pY8ts2HqMEdwft9kak70cOEYQms71DsNOTEwUx9GQMGz6C3Y5y5a47rrr8Pbbb9fahg416yDrgccffxwff/wxDh48aD4Z3bt3F8m72rdvL4zL0lBMREaDcq0y+8ub0/qioWHY7IN246g3ZHEZjUdmNmzZeuBA+NivB/HYb4dwRqsg3Dm8lQizjvb3hKurY6I+jKYHmVxG45HJZTQemVxG45HJZTQemVxG45HJ5cw8pvJSMeu7+uAOpPiFY8wZM9AiPLZGrvJju7H64/twdmh/+Hp5Y6i7CcsLShDs6oJfZ85Bh8i4au3LDm7Aqk8fwtSYEbikUy+8N2JileNVUY7nFszDvSXeuDc6HE9Pv9LcfsvH92Jci7EY3rI9lk6YLWZcLXHk6G6M/GERXE0V+KoiEanpx3BJ3Bh0iojFymkXwtNGrV7xwMTNDbf/8iVeP3IYlQCmtmyHL86cDn8HJ7dyZj046zVryw+0y1luTmios8ynFHyCIgOyuIzGI9NZlq2HZ/5JwdzfD+GJcW3xwJg2uiyLMJoeZHIZjUcml9F4ZHIZjUcml9F4ZHIZjUcm1+nEYyorRmV2MlxDWsDFvfYozcr8LCTNvx0XVIThL9YK5syyqQIvdOuJS3qPOoWrIjMJu9++DmeGDUSbiDj8OOV8hHh548DOvzFx1XIUuXnijxmXot2J8OiK5P3Y9+6NGBM9Aq0j44Uj6+V20hEyVVZg7qdP4uESTzzsXYb7IkNw5J/FmBF3JhAYiXWzLkWwjdJNxMHjh3HmT1/icHmVOzQkMhY/TDpH9Kcu2yUX5qOkogKtAuqXA8YZ9dBcrlm71iwrNAza+m4jcRmNRyZkHtOba49h7u8JeHZSe9w9smptjR4woh6MdkzKdsp2SnfOdS2pa1bZ7nTTnQiz/vMzvPfPT3jbvxUKXD0wOyoaj0+7Av6eNhJoFecj+6ObcTlisC0gGj+Nm4VeXh546Ju3cMPOHUjPzcJDky+p1j59/u24JLgXPP2CsXTiSce0Xdeh+M3bH6OXf40zl87HqpHj0cKlAmmL5+LiyEFw8QnC4nGzqjnKBNc63zTtRpT9uRiPp2Th530JONhyIjy8/bBy4tk1OspE25jW2HXpHfj+6H74uLljYnzbWtcEW9ou2tff8HpwJJx1vFPOsoNKRxGySkdp8f16r1nWc60Roa0j0Hj0XrPMPrVq1Uocm96lo2grGaWjlh3Ix32/JuCGAZG4skeg2J+j1yNqNtT+r/ea5frou7Frltlez/WImg3ZRu/1iLQht2l20XuMoB1klI7icckoHcX+yCodxWPQcz2ipb71Xo+o6VvvsjDaGMH2MkpHWetbrzGiJn3rMUZw33quR9RsaEvfeowRtvSt1xih/bbqvWbZUt8OGyNy01C47ScsS02De0QrTB4yE16ubjWWjvL66yPct2sbXgvtiZmhoYguzMA7aZn49ZMX8ONZ1yDMx8/cNtDfD7kf34n/lfljZUAElo2diUH+IeKzVy+4B8GLnsfDiUnw+f4j3D75UhRkpaPiy/txj0s4dnkG4qczJsCzpBRFLq5meweEtMTySedg9M9fY9jK5bg2Zx++jhiCve5++P6M8QhxdRf2srZ3fkEBbh9xDrpnJuOTXVvRw8cX9/YahJa+/uZjrW2MmBwVL2yYd0IDNY0RtHF9NNuYMYL8bKf3mmUeP/er95plV1dXab6Gtl9dSkc1dzQ0DJtmlJURXBaX0Xhkcsng+X5nGs5asBUX9orCR+d0c9ja5OZgO9lcRuORyWU0HplcRuORyWU0HplcRuORyWUPD2dmTSWFcAmMqLNt2d61+PWrpzEntC/S3KpmhSNM5Vh45jSMad/zlPbFfy/E/JVf4vrIQXhx8Bjc3nOA2L7hr8WYsG0n4txd8fsFNyHUP7iqJNOSJ/D2vl24K6w33hw2/pRM0JXlZbh5wdN4s9QTzyAdl6XvxLOesXghsD0+HDkJl3c6tQ8aDmdn4obfl+KXjHR0D4nAByMnoV9EdKNs5wg0V901F9vZ8gNrzz2uYDe0JzBG4jIaj0wuPXk4ALy9NhGzP9uK6V3C8fzYWN0dZaPYrqm4jMYjk8toPDK5jMYjk8toPDK5jMYjk6s2HlNJAdK+fAx3vHozOix4Ff3eeRTz1yyvlkHbEuVHtmDL5w/joohB6BzTGokX3YD9o89E17I8TPrtB3y7Y3219qU7VuGfXz7EbZGDcHnHHritR3/zZ/2GnY1fBg7G0XITxi14Ganrl6FgwT34dud63BvWG7d072ezZJKruweemHY9booMw70IR3T4cOEoPzNwVK2OMtE6OBQ/nnUFyq66G5vPvrxWR7ku2zkSzU13joSz2q5BYdicov7yyy9FeSnWNp41a5aYSldQOJ3BUmJa1vaePWsfpE8npOWXYl1CDrYl52N7SoGol7w3vRA3DI7DS1M7oiCvKgRHQUFBQUFBwTnApFVJm1fg3+IytO02HL3b9651NjnjvRswC9HYENIFc6IikJi0H5dv24I/kxPwzowrq62zrUg7goRP7sEFMcMRGRSGbyedW7UuuEN//BQcgXMXvY5z/voVXxTlYXb/MSjb/Rf2fPk4Lokbh97hMXhr+PhTZuX69huP5Z5+GP/PKvRdvx59K0z4MXoYZrXphBcGj6mx71xP/NrMq3BxahLWpx7H0Og49AmPsttOsmYiFRRqgs0wbK0GLWPLBw0ahM8++8z82c8//4wLL7zQHHNO9OjRQzggYWFhcHY0tM4yHxawnYw1y6xZx3UBeq9ZZk1Gfk/vNcvcF3mMVGeZx2W5vrwha43WH83CIysT8ceRKtsGermha4Qvukf7Y0aHAAxtGSBspB2nZm+91izz2NhO7zXLrHdrr74bu2aZn4eGhuq+Zpk8LGmn95plvljfW9aaZfavJns7aoyw1LeeY4Smbxlrlmkn7kfvNcv8nPvRe80yeag7GWuWaWPz2kodxwi+W+pbrzGC7bR7KRlrlsmp95plclnrW48xwpa+9RojLDXhkDHij0+x4N/luDe8D4pcqsoTTfJ2w/vTr4TPibtys2ZzsmFa/CBuLfLAIv+WWDpyKoZExCA4wA9vfjoXt5b6YHKALz4/6xqUFRUDuako+/xOnB3UC9t9wrFi7Cz0jWtVTbMlOWm48tt38INbEF4pPowuGQdwXdwoFPkGY+2sOfCvMNVow0O52Xhk/SocKirEee274YrWnehM1DhGaLXE9c5rwPPOz6w16+gxglzcj4w1y+wjv6v3mmXy8rt6r1n29vYW9pSxZjktLU30R7c6yzTUkCFD8MEHH6Bz587m7Xv37kX//v0FybBhw3DLLbeIA3jzzTeF8/HWW2+hua5ZphBkza7L4jIaj8zSUY09Jq5HPufzbWgf5oO7R7TCyDYhaBVSNRA6msteGI1HJpfReGRyGY1HJpfReGRyGY1HJpfReOriqsxNQ+Kyl/FqUgo2eYehY3gsHpx8CVoEhtpsz0zTX61chDnRQ3F5x+54sNcg/L3qC9yUnI0YT3f8fu4NiDlRishUWSnWBb94YB8eDu2B+aMmY07HHuZ98fMlX7+Ii9NLcYaHCV92aA3TX5/jhsAu+M47CiumnI8RMfE2j8cFJsz59gN8kVXlgLQLDMYPE89Gp2DHTnwZTQ+ni+4Ujz62s7t0FHf+zTffiJkQS1x99dXCUR45ciRWrFhhLvY8fvx4pwpr1QN84iLr4pHFZTQemWjMMe1KLcCFC7djUscwLLywB7zcXZvVeVLXkrKd0p1zXUvqmlW2M5LuKnNSUHJ0B7J8ghDdto+YQKqJqzIvA9vfuxnTfTsjM6gjRniYsDgrG1988TZ+mHAWhrY+OeFElG5dgTUrPsK18eNwdptOeG/kZLi6uKDNrJvR86/FmLh1O0Z88TpWnn0lWgSFo2jZi/h29wY8EjUU9/UeXM1RJlxcXTF79p0IWrUQs/ccxJBtu+AeMQQHXLzw2ZhpNh1lgsfDmbvPzr4ed6anILu0GMOi404pyeQIOLsemopHJpfReBzNZfOqaNeu3SmO8vvvv48///xTTFd/9NFHZkdZC7Xg9L6CgkLjkFtcjlmfbhGzyJ+e161OR1lBQUFBQUGh8eAsbfHv7+P9f3/FIyHdkeXmhda//4RXh03EsJh2p7SvLMxF4oe34lyfjnALjMDW6RejdUAQ0g5sxIyfvsSEn7/GtyMn4MzO/UR7rgveu3geLoyfgB7hMfh41BThKGvoOexsLHP1wozNGzBi4Vv4IH839paW4/boYTirdQfMGzDCZr8ZcTZu9AX4o0cyntu8jhvwUfd+GBzVos5j5nfrSpyloNDcYdNZ5lM0S+zZswd33XWXuKgefvhhtG7d+pTvaLHszRUyHxbI4jIaT8eOHbF69WrxfjoeE1dEXP7VDhzPLcH6mwbC38u9WZ4ndS0p2yndOde1pK5ZZbvTUXcMj85Y/RmWpCQj3T8SEwdORP82XW22FWWQlj6F+bs24ZaIAbi0TQdM8ATmb1uHGX/+iue7peCOYVNOti8pRM7823CFawxSfYKxZsr5wlEmItr1xc/nhGL64ncwfdXPWHx4K0Z5VCDxjwU4N34cPH2D8O2Es+Dj7nFKP3oOnITVrbth3PLFGONetVzsvDadMN/KsbaFPuHR+HzsDIfYzoh6cDYemVxG43E0l801yzNnzkSXLl1w6aWX4tChQ7j++uuRkJCAwYMH4++//z5l3eTu3bvRvXt3czKM5pjgi4vnCRkJvpKSkkRf9E7wlZqaKnj0TvDFrOr8v94JvrTi7jy/eibvIcjFv+uTvGfer/vx2B/HsPjCbpjUPqjO5D2aDfl/TX96JvhKT08X/dQ7wRf1TX4ZCb7YZyYm0jvBF3liYmJ0T/BlrW89xwjyaT8feo4R7I/2AFfPMULTt4wEX+Tl//VO8EWe8PBw3RN8kYe6k5Hgi33m/+21d0PHCH5mqW+9xgj23Za+9RgjuC/uU+8EX9yXtb4dNUbwWPk5/7alb7MNUw9hz5cP49Lgntjv7ge/ygrku7rjiogwzBs+E15ubtVs6PLXJ1ix8RdcEDMCF7XpjLeGjhNc+SkJePj7t/CmRyQebRGLW86YCpfSQrh+8wgezXfBK4Ht8d3YmRgX1+YUzeZmJuPiX77AHxUeGF+UjE1+0aj09MWvk85FvIeXTXtr+i6pqMB/2emIDghEW4+qc+/IMYL25f9tadbRYwSTxXIfeif44r543Hon+OK+eNwyEnyxHfetd4Iv/p9t9E7w5WExtuqd4IvJYvmuW4KvXbt2iQRemqHYhKHZq1atQosW1cM6mAWba5kPHz4siJtrgi8KhSdWBmRxGY0nMTERTz75JO6//37ExcWdVsf0674MTPhwE/43qjXmTmivK1dDYTQemVxG45HJZTQemVxG45HJZTQemVyO5mF4dMmaL/HdmmX4xC0Mhb7BOLvbQJzTZTAibFRhqchIxK73bsL44H4IDYnC4gmz0dbdDW9+9QLuK/HBMH9vfHPujQjwrHJYi//6QtQXnh4/DmPi2+Hr8bOqlWGqLM7HfZ8+hedMgbirMgM3Zu7AW16xeDagPZ4dNAp39xpUY9/LKyvx6pa1+CHxMNoFheLhvmcgzj/QcOfodOAyGo9MLqPxNIbL7gRfnFXevHkzPvzwQ+GZ8+/LLrtMPGmwBOsr00nm9uae4EtmHThZXEbj4Uw5M7ZfddVVujvL9TmmLUl5mL1gK8Z3CMNj49rpytUYGI1HJpfReGRyGY1HJpfReGRyGY1HJlddPKayYpQf+A8phXnwatkTEeE1r60V2aC/nosn9u/HM6E90N/HEyH5GbhpxzZ8dWAXvj3/JgSyfvAJVGYnI+m9G3FeUG94+IdixbQLEePrLz67dc5j6Pz9mzjneDbGfPI8lo0Yh4Ddq7F1/Q84r9VEdA+PxsKx06s5yoSrtz/+d/adCFzzPR5ONOH58BFwhQue6D8Md/UcWOuxcl939DlDvBxhO0ehOerO2XhkchmNx9FcNmeWmzMaOrOscPpDZukoe1BRacKXW1Nww9LdaBfmg5XX9EOAneuUFRQUFBQUnA3lR7Zg16LHcb1Xa/ztEynq847388Kb0y5F26Dqs8Ra2aR39+3EHeF98Xj/YXiQTmdlBZZ//ybOS85Fe3cX/Hzu9YgIDBUzyrkf3IgL/TrjH98o/DPjEnQLjTilD2v+WoJp23ciqKIEY4rT8WVQW7QMCsOqaRcizLsqdLgm7MnOwPq0ZPQJi7S5bwUFBeP5gfVOtXvttdc6sk+GgRavbyQuo/HIhK1j2pdeiId+OYAZH29Gr5fXIuCRlaJE1Mi2Ifj1qr4NdpSNdp7UtaRsp3TnXNeSumabr+3K9q7BrsVP4d2Fz+C7/35DaUVFjTxlBzdg48f34sygvjgW1hofDx6JN0I9sCsnC4MXvYONCfurJ9z69hl8v2sD7grvixu69hGOMmeLXNzcMWnmLfihU0ckllVgxGevYOvip5HyxuW4wbcDfvYIxRdnzqjRmR0ybDb+nn0VurTogD9b9MEFnXrhrxkX1+ooa8fEGsQXd+imm6OsrlllOyPrIctJxzubd+e//PLLKdtYS5n48ccfccMNN6Br167mpBRMkHPOOec4rFMKCkYCf/SfXHkYj/x6EEHe7hgUH4gzWgXhsn4xGNEmBP3ial6/pKCgoKCgcLrBVF6Kgm+fwwt7d2JeaHeUMeJx43/osHk9Xhs2GROs1gqWH96MIx/fjfNbjEFEcCRWTbsIET6+QM/BmLnzb0xZ+QNG/7AISwYOxpiOfVH0w0tYvXMdrow/E9NbtserZ4w9Jayye+9xWBXfAdN+/xG9MwDfmDNR5uKGBaOmYHLL2pc0dYqIwbKZc3SxjYKCgrFgMwybU9A//PCDGJhGjRqF0aNH44EHHhCfxcfHi2y1ll9jBrV9+/bZLCnVXLJhM1Mbs63JyIbNdeRsp3c2bGaHJY/e2bDJy33onQ2b6+tffPFFPProo+b193plw+ZxkYPH88Qfx/DS2mTcMywOtw2KgLe7q0My3Wo2ZH+0mwg9s2Ezazn3qXc2bGYitFffjc2GTe6wsDDds2GzfVRUlO7ZsGmbyMhIKdmweRxaUkc9xwhLfes5Rmj6lpENm98nh97ZsMkTGhqqezZstqfuZGTDph2049ZzjNB+z+vSbGPHCE2vddnbEWME+6hlJT5ljDiyEzm/v4vXc4ux3jcKLQNDcf3Q6TijbadTx4icbOCbR/FcVgGeCu6CWzr2xJ2deyFp10rcvH0rtnoG45Xeg3Bl3+FV9k7ei8Iv/4ezokfgoFcIVoydid5xraqNESV56Th76fv4y9UP1+fshb+LC14M7oz+4dH4Zer5KCssOmWM0PSdV1KChft3IrW4EBd36YVoNw+HjxHcF8G2eo4Rlvq21qwjxwh+R6tgo3c2bB47baR3Nmz2mZx6Z8OmfsgpIxu29Tih1xhBXn5X72zYXl5e4v8ysmEzTxG/o1s27Hnz5uGdd97Bt99+K5xHS9BZnjp1KgYMGCCMz6+/+eabGDlyJJ577jk01zXLPPk86TIgi8toPDK5NJ5VBzIx+r2NeHZSe9w9Up+HSUY7T0bWg1F4ZHIZjUcml9F4ZHI5O09lYS4q04/CNSQGrgFhtXKV7liFLUuexrmRQ5Dm5oOxnpXYVFCIRDdfvNizL24dUhVZSJgqylGw6CF8dPggbgnviyf6D8eDfU8mrypM3IXrvn4Xn3pH49ZgH8wN90P2H5/ispjhWOsZjN+nXoCBkbG2bVFWgsd+/xqvJiSi3MUFl3fqiZeGjIGXm+0lSkp3DYeynbJdU+ih1Al+K+zOhv39999j/vz5pzjKmvfOjMKWoON80UUXoTmDT1xkCUAWl9F4+ATt33//xcCBA8WTJD0hnsC5uuOGb/eIkOs7h7fSlctI50ldS8p2SnfOdS2pa/b0sZ2Y9fv1XXy3bjnmBXXCXo9AdPNwwUPDJmFkZKtTuEp3/YndCx/FtPgJCAkKw7ZJ56JNYDCKkvbinq/fwm3bNuFw+jE8P2UOXMuKUfDFA/gu4RBuixqCa7v0xgN9hlTbn29cF7x81s3oseoL3JdViJ9SU5AdNxH5bh74bsLsGh1lwtPDC/MmXIAnTCZxHG5WWamtoXTXcCjbKds1hR4KnNRXsuksc+p6zJgxNr+wevXqU7Z169ZNTJErKJzO2L17t1hSICsb9gt/HsHe9EJsvHkgXF3lpctXUFBQUDAGKvMzUfLf91iXlozjQbEYMmAiWofYTi7FWd/Cb57C0/v3YG7kYIwLj8DFbuX46eh+zPrjZ1wTHYM3pl5sLo3E5FzHPn8Q57UcD1//ELGOOMrXT3zmE9sRr149Dy2/eAb3JqVg29sP4o7s3VjvHoC50WdgVusOeGPoOJvlWVz8QnD3xQ9gQMJBfLB3OyJ8/XBjt75oF2hfzVNX7lNiiRkFBQWFejvLXFtXE2oq8CyzdtbpCMblG43LaDwykVbmhsd/O4Tbh7VEz5gAXbmMdp7UtaRsp3TnXNeSumb1sR2d2QNfzcVlgd2wzisMyEqC66EPcFVcHF4Yfy78PU7OmpjKSlCw8EG8kZiIuWG98UjfoXik31Bxb3ZXbjpe+fwp3H28EvsXvIBFY8+C//41SFjxHs5tOQ6pnv74Z9I5ZkdZg4uXL+6a8yg6rfket+zciSnhg0R94Vu69cOzg0fVOPOrHdOo+LbipReU7pTtmkIPSnfNz3Y2nWUtUYy909cMb83IyEBzBu0la3ZdFpfReGSB4WO3fr8XEX4eeOTMNrrzGe08qWtJ2U7pzrmuJXXN2ve7ULblZ+xZ/yO+LHNHQXAspg6ejEFRrWyeo9Itv2Db4qcwI34sKn0CsXzUFPRECT7+8UPMSyjHP5+9iqUzL0e74DBUFuWhYME9+DwtA/8L74c7ew4wO8qEa2A4brvqabT/7g3MSclD76WfYEJBEn5pPRmlHr5YOeU8URLJFriP6WdMx8RBU7AtMw1RPr6I86+9goPSXcOhbKdsZ2Q9lDqpr2TzsWC/fv3w6quv2r2TZ599Fl26dEFzhpbpzUhcRuORhaU70rB8XzZemdYJ/g2sndycz5O6lpTtlO6c61pqrtdsZUG2eNUFk1jrez/e/uF99DfF4jX3SHyZnY8zf/0eZ383H2lFVRlpNZRs+lHMKJ/dajz8giKw/qzLMCG+LWLiu+Deq5/GqpZhKCzIRv9Fb+Prr55D5ovn4OXsYlwX3g+Xd+qBZweNPiXaz8XdA8NGz8H6mZdgeGxrrInrhzPbdceamZegZ1hkncfg6eaGfhHRdTrKp9s5cjYuo/HI5DIaj0wuo/E4msvmnTzrKA8aNEik2b7nnntqnGFmeu+5c+fiySefxKJFi2AEJCQkYNKkSfUuHcWnxvxMRukoLXW63qWjLFO061k6Svuu3qWjuG+2o200uzi6LMz+Y2m4YekujG3tjykdT9pfj7Iwmg3Jr/HoWTpK+7/epaPqo+/Glo7iS8+yMJoN+X29y8LQhrSVZhe9xwgeh/ZdPccIS33rOUZo35NROoq20LMsjKW+9S4Lo+lb77Iw2hhhOX7rXTrKUt+Wmg0oSkXK0ufwWm4pvvdrAXh6Y0p8e9zYfyyig4Kr2TvQxwu582/D6znFeDhiAK7p3AsPdekL35J8LFvxPu7MLkbPBa/g7cHjMLVLb+T99gGSVi/AzDaTUeYdiO+GTYJvaTmKXIvMY0TbIefj73b7cPHq5ZidVQm3iFGogAtu7tQTj/QYhJLi4hpLR4UHheDzGZedHCPcPYW9HDlG2NK3HmOELX3rNUZo9yp6l46y1LeeY4SlvvUeI7TzKqN0lC3NOnqMIE99NNuY+wgeP9vpXTqK/eF+9S4d5eLiUi97N2aMqK+9LUtHWcNm6Sjitddew6233ipqkE6ePBk9evQQ9ey4c9bf3bZtG5YvXy4OZObMmYLECGho6SgFhYpKE2Z9ugV/H8nB1lsHoUWQtzKKgoKCwmkOzg5XJGxHuU8gPFp0NdfUtYWyfWux+/OHcVHkEOx198P5YcGoSDmIxS6BCHN3w/yxszC2dUfR1lScj/xP78b7GTm4I7QX/td7MOYNGGGe9eXt15G13+Cy9f9gtVc4ritMQOuidLwQ2Q+ePv5YOfUCtA+qOSlWpcmE344dxr6cLAyNjkMvO2aIFRQUFBTq5wfW6CwTbHzTTTchMTHxlJAefo1PjW677TY8+uijhllz2lBnmU9K+GRDBmRxGY1HLy46yQcyCnHPT/uxbHc6ll7SE8NiPJz6mJoDj0wuo/HI5DIaj0wuo/HowVX8zyKsWzEfDwZ3xV8+kQg0leP8uFa4r/9YtImKrta2ZNNP2P/105gaPx4mvxB8P+kc9AiNgKmyEvv+WICrtm3Fn94RuDk6Ek+0iodpxZt42xSA/wV1xU3d+uLVM8aeci/F4/E3FeGpP77FM2k5KIYLprfqgDeHjUO0r2OT4RhND86su+bCI5PLaDwyuYzG0xguu+ssW35h2rRpWLVqFdatWyempjmVzhnmXr16Yfz48dIO+nRHLc8cnJbLaDw7d+7EWWedha+//hpdu3Zt1L7+OZKNBZuSseZINnalFaKkvBJB3u745uKemNolwhySIgNGO0/qWlK2U7pzrmvpdLlmWTqpbPtvSEs6gO2ewWjZdSg6R7e0Wa3DVFmBomUv4bMtf+OmmFHoHBSC12OicGTXP3jv6CF8nzQfn46fjTNbdRCcJas/xeFf3sFZbaaiwicIq6dfhNYBQWJfLq6u6Dj6UvzW7RCe//p1PJZUge8TDsI/oDe2u3iJhFvP2VhHrB2Pe0gMHpp5He6vrERpZQV83PWZfDCaHk4X3SkeZTulB2Nfs3VmH+Lgztq0fCnUDFlFtmVyGY2H6xX27NljXrfQUMz97SAe/vUg2ob6YETrYFzWLxZdIv0wpFUQAk4k9FJ6aDiU7ZTtmkIPSnfObbuKlAPI+fRuPFkRgDeDO6HIJQ84sBD9fb3w5vhzMSAy1tzWVFKIgoUP4IPEY7g9ajAu6dAd746YAC83d5iGTcM1qz7Fldu2YdzyJbjB3w3XFhzG5uQE3Nd2BuDlh1VTzzc7ypbwiGyD+659HlMS9uHZHRtR4emDlzr1wti41nYdD0sx+dRQjskRMNq1dDroTvEo2yk9GP+ateks33HHHXjxxRfrtaM777wTL7zwApornFUAzYnHEfhqawoeWnEQD41pg0fGtoWbq+364koPDYeynbJdU+hB6e70sl1F+lEUr1mMjbnZ2Ocfg049R+GM1p1tcpXtX4/Uz+7DdRED8aN7MP7XZwgujInFtj++wLzMLAz55hM81KU7Hhg2GS5ZSSj47F68VuyOByP64YauffDa0HFwPTHry1ni9mPmYHmXA3jh1y/wdH4l3nBpCUS3xISY1vhw5CTE+gXUOsHQs2VHLGhZtW65LijdNRzKdsp2TaEHpbvmZzubjzA//fTTes2+MfPaggUL0JyhZXMzEpfReBqLxJxiXLlkJ87vFYXHxtXsKBNKDw2Hsp2yXVPoQenu9LCdyLT7x8c48vKFmHIkBSMK/HBlSi6GrfgOfT95Eb8c2HVqmaWP78K0mFFY6R2ObyfMxtwBI9A1rj3Ou+ghrB0/A3eUHMPju7Zj8DuP4Pl378YclxixRvneXoPwuoWjbIli71A8eMmDODrndnw3YTbWz7oUyyefW6uj3BAo3SnbNYUelO6U7ZTuGjmznJGRgfj4eLRs2VKkG68NzI595MgRqWs0FRSaAo//dghe7q54Z1YXm2vPFBQUFBRORdnetcje8AO+yS9GenAcevcejXHtusHD1e2UNceF3z6DnZtW4MK205Hp5oWlIydjdEgI/vx1Ph47loKpq3/C5Xs24Jl+Q+C38XtsWv8TLmw1CWXe/lg98Wz0Da+elMu30xA807YvJv61BPcdTsCDYb3RKiAIH/cbhks7dq/zdIV6+2Baq/bqtCooKCg0U9jMhs0aW/XekYuLcJybazZsJj6TFV4gi8toPMyM99tvv+HMM8+sd2I6ziq3eeZvPD2xPe4c0arO9koPDYeynbJdU+hB6c7xtqPzW/TTq/hxw2+4PnoIMlzc4VdZgXxXd3T0dMdLo6ZjcusOVW1Li1Hwxf34/cgeXNZiNKICgrBs4tloF3iydFJp0l68+e3beNgllDcv6FBRiC2ewegSGoEfJp6NeP9A3Y/J0VC6U7ZTunOua0lds8a2nd3ZsFn4ee3atejevXudjjMd5O3bt2P48OFozmAhbVkCkMVlNB46yJMmTRIFyOuL1/9JgK+HK64Z1MKu9koPDYeynbJdU+hB6a5umMqKUbJ2CVL2b8APLn6oiGiDQX3GoFtQxCnnqDIvHQWf349XswrwUMwITIxvizeGjkNLlwr88+PbeCg5BVN++RoXhgTg0dgIBP67GO+WeeHxmJEYFR2Hr8bORLBX9Vr1nrEdcfWlT+CslP14Z99OJLr74LKoeFzeqQf8PByrEaU7ZTsj60GNd8p2Snf2w6YnPHjwYPTs2dOuGWaGabOM1MCBA9GcUVJSYjguo/EkJyfj6aefFu/1QX5JOd5ZdwxXD2xhznZdF5QeGg5lO2W7ptCD0l3tqEjej5yXz8fLf/2AzmWRuLnUD3ckpmDQskUY8/V7WJ101Ny2/MgWZL52Ke4r8sCDoT1xT69BYpa4TWAw3ALCMPy8B7Bi9tV4yyUDP6RnoOOOg4j264tHg7vhxm598eOkc05xlC1nC1q27YV5Ey7Ax2fOxE3d+zncUSaU7pTtjKwHNd4p2ynd2Q+bd/6///476otff/0VRkBCQoKYfZw9ezauuOIKkQSBs+fu7u5iRjI3N1e08/HxMSc3IyorK8VnbMsHCP7+/sjJyTHP1DNMXWsbGBiIwsJC8WSPbQMCAkSIsNaWDyn4OcHPmGytrKxMbOd3tT54eXmJfhUUFIi/yckbCb7Ix5lUbS0523p4eJiTOli2JUJCQkQfGJXPp5p8aTx+fn6ir9rgyrY8Nh4z98k+5+XlmdvSBlqCOPaB+2Fbaxvy/5rdiKCgILvtbW1Da3tb2lBru3v3bjzxxBOYMWOGsIfWlt/XbGhtb/bp7b8OILekHFf1CRf2ssfe7Cc/t7Q392lpQ2t7a22t7W1pQ2t700Zsp/GyLT+ryYYiec6Jc2Npb2sb1mRvoi5729Kspb5rsrdmQ+3Y7NG3vTa0tLelDWmb2jRry97cbsuGtWmW22qzt6PGCO5fs4veYwT7pn1XzzHCUt96jxFEfTSrjRGWNrTH3uyztWb1GCO4zVqz5jGivByuh/9DxbHt+Ku4AhUtuqNvl4GI9PY9Vd97/kLesmdxY9QQLHMPxjXtu+HBvmcgIC8dP6+Yj6ezKzBy2ReY7euOOZV5yDuyGU9FDsYuV28803sIbu7eH2WlpdU0WxIcj/PO/h8mFuTgv7xMHCkowsjIFugWEQVTeQWyCnJtjhFEfezd0DGCGrPUt15jRE361mOMIK+1vvUYI2zpW48xwpa+9RojuD/2j231HCMs9a3nGGGp7xrHCAfdR2jfrY+9G3IfQR5bmnX0GEGe+mi2MWME+8l2mmb1GiO4P+7Xlr29HDhGEPWxtyN9jbrsvWTJEixduhTHjx+HXWuWmzMaumZZ4fTHxo0b0a9fP2zYsAF9+/a16zuVlSZ0fWkNekT746uLeureRwUFBQW9UZF6GAWLHsSX2QV4IKIvUl2rZmbdYMI5MS3wyPDJ6BwcBlNlJUrWfoWty9/BpfFnIsndB5+MnoqZrauXRSo5uBEfrfwKT5R6I8mtakZ4UEQ0Xh82Hv0jYtQJVVBQUFBwWj+w/pm8FGxCewJjJC6j8TQEP+/LwJ60Qtw6tGW9vqf00HAo2ynbNYUenFl3dGpZa7hk/VIk7fwbqXnZNfKUbv8daW9cjjtco3B11GCMat8TG2Zegh29u+LJkkT8dXQfun/5Hq767Hksfe92vLbqa4yLnwCXwEj8O2vOKY4yURTaFtdd+RQOX/sgtp59BXafexX+mXmpLo6y0fTgzLprLjwyuYzGI5PLaDwyuYzG42gu+xZgKtQJmRP0sriMxtMQvPJ3Avq1CMDQVkH1+p7SQ8OhbKds1xR6cFbdcS1x/hf34+fcQjwa1hO7PDlW/YXOXh44t10P3D5guFj/y8zUxSs/wsY/vsANrcZiNzzxztBxuLpzr6rwz8hYdOk/Gddu+AGvrl+J1/JC8YFrLFzCY3Fum854d8REBHp61Xo8LAXVIzTCYcdWG5feMBqPTC6j8cjkMhqPTC6j8cjkMhqPo7mUs+wgMJ7eaFxG4+GahpkzZ9pdNmpHSj5+3puBT87tVu+6ykoPDYeynbJdU+jBGXVXsuknZH79FB6PHoi3YyIwNrYlHoiKQMXuv/Fj2hE8XVyCl3ZtxJVBPhiUvA2bikrwcvwEdAgKx9rRU9EnPKra/lxc3RA0YDoeGjAd91VWICE/D+HePjU6yY4+nuaoB2U7ZTulO+e6ltQ12/xs1yhnOT09HYcOHUKXLl3EIu7mDC48NxqX0Xjatm2Lr776Sizotwf3/rQfbUJ9cG7P6jeU9kDpoeFQtlO2a66645Pwsu2/o2DdN1icV4RVPlGoDInFkI59cXbHHoj1C6hqV1qMoh9fwbYNy3Flq0nYDw+8Ong0burWt+rB3oCxuCgjEYdXLcAbxxLxfkUUXvZuD3cfF9zWYwCe6D8c3nWMg5wlbhsY7DS2UzzKdkoPp991JJPLaDwyuYzG42guu9YsX3fddbjwwgtx9dVXm7d9+OGHiIuLE2WmWrVqJZImNWdoGdqMxGU0HmbjY0Zsy6x8NeHH3en4YXc6np7YHl7u9V/ar/TQcCjbKds1R91VFuWhYME92LHocUysCMc1fh2xw80XB1OO4q5//0Crz97ABd98gJV/LcXPb1yDh/ftxZmtJsEUGIH1Z83Bzd37VYuAcQuLQ+iYa/HiLa8h+ap7kXTxjSi88k48N3h0nY6yI45HLxhND8p2ynZKd851LalrtvnZzu5fzJiYGDz++OPi/3SM6UCHhYXhjjvuwOHDh3HvvfcapnyUgjGxffv2OrNhF5RWYOWBTMz5aicmdQrDOT0ipfdTQUHBGDCVFqH84AZUFucjO6oDgiJa2XRUyxN3Iv/z/2GhyR/3tJmKMN8A/DV6KoZGx6EyNw0pa5bg4x3/4e2yAixMSwf8usHXzQ3Xd+uLR/sNg38ddYZ93T3ES0FBQUFBQUEHZ3nNmjXYsmWL+e+bb75Z1KdavHgxhg4dKrb16tULzRms+WU0LqPx1IS/D2fjnp/2YXNSHgrLKsW2kW2CseC87vVeq6xB6aHhULZTtnN2HlNlhUimVfTHJ1jkGYG5od2R5F71G9rFxxsz2/fA+R27o3tgEEp+/wDJf36Oh+OGYYFrCC5t2xWvDR1rXifsGhiBmAnX4d6x5bgzJxXrszLgHxSJziHh8HRzk3ZMpwOPTC6j8cjkMhqPTC6j8cjkMhqPTC6j8Tiayy5nWSssTnz22WdYu3YtzjnnHLOjTLDgc3MGHx4YjctoPLbw0X9JuGLxTgyIC8QT49shzNcDPaP90SsmAK6uDXOUCaWHhkPZTtnOmXkqC7JRsOgh7D60HXe3n4Y/Sipxbqt2mOVpQua+dfgrtQBvFuThqW3r0baiCB1LsrGm9XRUuLrjo6HjcFmnHjb36+LmDo/QWJwRGiv9mE4XHplcRuORyWU0HplcRuORyWU0HplcRuNxNJddznJISAhefPFFEXZ92223CW/92WefNX/OpElRUfVPgmQkFBcXV3uoYAQuo/FYY2dKPm5YuhuX94vBe7O7wq0RzrE1lB6U7ZpCD0p3+tnOVF6K8oMbUViQhSPeYagIj0fX0Ah4uZ38GS07uAG5Xz6Clzxj8FzLSYjzDMTPZ47H+Lg2VQ1Gn4NrkvYgY+Mv+CczFd+ZfJES0wfXx7bGbd37I8rXsU/dle6U7YysBzXeKdsp3TnXtVTspL6SXc4yHeVp06bh4MGDCA0NxYIFC0RSr+eeew4//fQTVq1ahZEjRzqkQwoKsvDAzwcQF+SF12d0dqijrKCgYByYKitRsm4xCpa/gQ+9ojEvpDty3KrWCHvAhH7+/pjWoiV6H9+OAwc24qPwAdji6oN7eg3Cw33PgI/VWmH32E7w9InErJAQzGqiY1JQUFBQUFCwDy4mO6o2f/HFF7jggguQlpYmZpm10jtHjhyp1o4OtLNj+vTp+O677+r9PZqxoetbT1cuo/FwqQAzYXt6euJAZjE6vfAP3p3VBVcNbOFwLqUHZbum0IPSnWNtV5mTgoJFD2Pv0d24re0k/FXugqs6dMWFAd4wJWzHhsT9WF3uit99olHo6g5+e2RMPOYNGIEzouOa/DwZjUcml9F4ZHIZjUcml9F4ZHIZjUcml9F4GsNlyw+0a2b5yiuvRHh4OMaNG+eUznFWVhbefPNNES5+ySWX4M4773Q4R25uLoKCghy+36bkMhqPq6srSkpKRO21D/9LQrC3Oy7uE60Ll9KDsl1T6EHpzj5U5meh/PBmVHr5ojgsHqU+QfAoKUNI8Mm6wmWHNiH/iwewyCsKd7SeimifAPw2YhLGtDjxu9d/HMYAuCMzCUU5qUj0DUNYYBgifHxPm/NkNB6ZXEbjkcllNB6ZXEbjkcllNB6ZXEbjcTSXu71x33Qyo6OjcdVVV+Hiiy9GsMVNxemOgIAA3H///dixY4d40qAHZCY4k8VlNJ69e/fiiiuuwPvvf4DPNqXivJ5R8PaoPZtsQ6H0oGzXFHpQuqsdFWmHUfj1kzhydBfeD2qPTwLaIMutKut0uJsrxkRG48yoGHRP2Y2DW37F/Ij++M3FD3Pad8XrQ8fZLNHkFhoL/9BYdD4Nz5PReGRyGY1HJpfReGRyGY1HJpfReGRyGY3H0Vx2OcssC7Vp0yZs3bpVrFfu378/Bg0aJGacx4zhs3X9wfXSP/zwA5YvXy7WSGdkZIgZQg10gp966il8++23Yto9Li4Or7/+unDwtbBxht/qBQ8PeTUsZXEZjSc/Px9///03/j2YjIScMlzQS59ZZULpQdmuKfSgdGcb/H0oXbcEuT+8jA8iemBum+lwc3PD5a3aYkBFPkz7/8WmrFz8UZCGxUnHUMnQrcgh6Bkaga/7DcOsNh2d8jwZjUcml9F4ZHIZjUcml9F4ZHIZjUcml9F4HM1ll7NMR5no2bOnyIL9zDPPYOXKlaKM1L333otZs2bhsssuQ2ys/SUt6os9e/YIp/23335DYWHhKZ/PnTtX1H1mTWhfX1/ceOONmDBhAjZs2GB2lvWEpeNuFC6j8WhYl5ALf09/DGmlXyiI0oOyXVPoQenuVJhKCkUpp837N+PGNlOxrdwF13XuhacHjjTXMsa4i3FebgZccpKRXZSPA95hCAsKQ/ugEKc+T0bjkcllNB6ZXEbjkcllNB6ZXEbjkcllNB5Hc7k25EucueXM8qhRo8Ra5gcffFD39cuTJk3CiBEj0Lt3b5tx6XTgWdaKjjLxv//9T8yEMzmZDOTl5UnhkcllNB4N6xNyMaptCDzcGiR/u6D0oGzXFHpobrqrzE5G0fLXsfX5c3HN63diyPvz0HH+ixjy5bu45KeFeObnz/Dsu/fg7CwTRsaNhykwEutmXoo3h40/6SifQH6FK9zjuyG84yAMatleN0e5rmNSPMp2Sg+n13Ukk8toPDK5jMYjk8toPI7msstbePrpp83x3ywVddFFF4nwZs4mr1u3Dtdeey3++usvNBX+/PNPFBQUCAdeA8OwW7RogWXLltVrX5WlJSjav9P8d9HB3ShNTqz6rKQYhbu3oKKg6gSUZaSicO92c9viw/tQcjxB/N9UXlbVNj+3qm1WOgp3bz3Z9ugBlByryiZuqqgQbctzs8Xf5dmZ4m9tfXVJ4iHxEm1NJpTs2y7aiLa52VVtTxTf5j65bw3kJDfBvoi25WVVbY8niD6b2+7dLo5JtC3IEzw8ZoI2oC3Mdtm/E2XpyVVtiwrEfiuLi6rapiah6MCuk20P7EJpyrEqGxYXVdmlML/KLunJKLHc76E9J+1dWlLdhrT3nm3V7Z10tJq9y/NyLOy9pZq9TWlJ4v9bk3Iw3SsF5TlZVTbMyaqyy4n1DSWJh1GccNDChltQnp1R3d7l5SftfWT/ybZ7tqEsM62avSvLSs02LD6892TbfTtQlp5S1bYwv6qtZu+UY6fYuzTt+AkbFp6wd1WERXl6yqma1eytadZs7xTBe9KGe0/au6y0ur0z06rZu5Q61DRbXm6l2Yzq9k44KOwo2lZWVrW1tremWdrbQrPUnaZZns9qmk06Wl2ztLemWc3epSVme1NP1e2dXLO9rTWbmlRds0UFJ2yYfNqPEVWa1XeMqLS2oU5jRDltaKFZntOShAMo/PYZpDw9A3P/+hPDvbvhG+9IdM1NxXn7/kHLhM3Ytn8zPty0Ea+XBCAzpjPeHnIm/uzSG31OJOCyHiPKhA0PV7e3TmNEhZW99RsjTtg77biUMYLHKWOM4Gd6jxGV1prVaYwopw4ljRFlSUekjBHlWtuCPF3HiEoLe1uPEXrdR8gaI/S/jyiqdh+h5xhRYrFfPccIjuEyxgj+JlWzt85jhJHvIyp0HCN4L2npa9g7RtiEyQ54enqazj//fFNUVJTJ1dXV5O7ubpo0aZJp0aJFpuLiYpNMPPLIIzyrpqKiIvO2l156SWzLysqq1nbAgAGmvn37mv++9NJLTc8991yt+x/UKtb024BI08SJE01Tp041/Tmuq2nvE7eacnNzTQWH9pr+6xdiOrbyR8Gf8MELpo2jWpsyMzNNhYWFpu2XjDHteeBaU05Ojqk45Zhom/jTEvFZ4oI3Tf8NjhJty8vLTTuvnmradeclom1ZbrZoe/SbBaLt8SUfi78z0lJNZWVlpl03n2PaeePZpuzsbFNFaan47MgX75oKCgpMqT99Jf5OTzhiKi0tNe25e45px1VTzLYg56GPXhFt01f+INqm7tst2u57+HrTtovHiD4RPJaDbz9tys/PN2WuXSnapmzbaCopKTEdePIO09azB5vbbprQxbT/5UdMeXl5ppwt/4q2x//7R7Q99OJDps3Teom2lZWVpi0z+pr2PX23sGHe3u2ibdJfvwrtHHnrSdPGsR1E24qKCtO284aZ9j56c5W9jx6ssuGv31fZ+6OXTRuGx5ttuH3OONPu+66ssnd6imibsOxLYcOkhe+a/hsQftLe184wbb5htum6B58yed/wWZW9F88XbZO/rfo7I/m4sPfu2y4w7bh+lrC3sGG/ENPhBW9V2fvnr8XfaYcPChvuve8K0/bLJ5rtvWForOnQBy8K3ozVP1fZe88O0Xb/ozeZtl0w4qS9x7QzHXhjnrB31vo/RdvkLf8JGx585h7TlrMGmm24eXJ30/4XHqiy97YNVfZet1rY8MCLD4rPtbZbzxpo2vfkncKG+Qd2V9l79S/ChkffeUbwmu194UjTnoevFzYsSkqosvfPS0XbxE9eE8ej2XDb5RNMu++5TLQtyUyvsve3X1TZ+8sPxN9sKzR742zTzlvOEzYsLy6q0uyXH4q2Kd8vrNLsscQqe99xkWnHNdNPanZAuOnwJ68Le6f9+l2VvQ/uq7L3A1ebtl069qS9h8ebDr77nLBhxl8rqjS7c2uVZp+4zbT1vGEnNTuuo+nAa48LG2Zv/Keq/+v+rNLsc/8TOjVrdlpv0/7n/ifa5u7cXGXDNauEvQ+//oTQv2bDrecMqXOMSE5OFm31HiNSU1OljREpf66QMkYcpL3HdTyp2bMHmfae38O05uGxplGvPijavvbhS6b03NyqMWJYnClz21+m4r3rTNsvGmnafd8Vdo0R26+ZZtp124VCs2UFebqOEfs0zeo8Rhz967cqzb7yiK5jBPVNG+64YpLuYwSPR8YYkbb2jyp7b1yr6xhxiPY+cR+h9xix/fqzpNxH7DuhWb3vI46s/Ml8H1FtjNDhPoLHzvsIvccIHo92H6HnGJG+cU21+whdx4gzTt5H6DlGbLt6qpT7iL0nNKvdR+g5RtBOlr6GXmMEj12Gr1FcXNxkvkZNY8SiN14RbW8cM8Q0ZcqUan6jBrucZRcXF/Hq1q2b6dlnnzUlJSWZmgq2nOV58+aJbTz5lhg+fLipY8eO4oS89dZbphYtWpj69Olj+uSTT2rc/5QJ402F+3aY/y48sMtUcjxB/L+iuMhUsGuzqTw/V/xdmp5iKtizrapdYaGp6NBeU3HSUfF3ZVlpVdu8nKq2mWmmgl1bzPstOrLfVJx4uKptebloW5ZTJbqyrAzxN08+UZxwULwqKipN3V/42/T2p8tEG9E2J6uqbXl5VdvEw2LfGshJboJ9EW3LSqvaJh0VfTa33bNNHJNom59ryti0ThwzQRvQFma77NthKk07XtW2MF/st6Koyv4lKcdMhft3nmy7f6epJDmxyoZFhVV2KcirskvacVPm1v9Otj24+6S9S4qr25D23r31pA1p72NHqtmbg8FJe2+ubu+EQ6a7vttpin78d1P+zk2msuyqi5Hvwi4VFSfsfchUdPSAhQ03m8qy0qvbu6zspL0P7zvZdvdWU2lGqtCDZu+K0hKzDYsO7TnZdu92U2lacpUNC/Kq2mr2Tk48xd4lqVXXXUVRwQl7F4i/c44ePFWzmr01zZrtnSx4T9pwz0l7l5ZUt3dGajV7Z+3edlKzZWVWmk2vbu+jB4QdRduKiqq21vbWNEt7W2iWutM0y/NZTbPHjlTXLO2taVazd0nxSc0e3G1l7+PV7J1/4joS9rbWbMqx6potzD9hw+P1HiO0sUnvMYI83FalWX3HiLz01OqabeQYUZx40FSWtNdUvG2V6dCPH5t+3/Gf6fdjh01/bFpnyty5qcpG+9ebDt8zwvTSvItN/u8/b+rx2Vumjf/85pAxImvP9pOa1eyt0xiRe6xKA3qPEfmZJ/SdmqTrGGHW9+F9uo8R5OJneo8R+elp1TWr0xiRm3jEfB+h9xiRva+q/3qPETwm0TY/16FjhPV9BM+RZm9LzepxHyHu8U7cR+g5Rgh9n7iP0HOM4NhgeR+h5xiRueXfk211HCM4hlvaW68xIufooer21nGMoB4sfQ29xgjyaPcReo4RhdS3la+h1xjBe0lLX8PeMWLatGkma7jwH9ixSJo1iqdNm4amxqOPPorHHnsMRUVF5sXb7777rggFZz1ly5JWgwcPFtnQGKbdmGLU9oDcISH6rXHTMPb9jXCtLMcv1wzUnUvWMcniSU9PR9+bnseA0ZOx5NoRhjgmmVxG45HJZTQeR3KZykpQ8vcXKF71MXZUuGBeaHes8I1BucvJVULMadkaZQgvysYun1Bkww1Xd+6FF4eMsVnOqbnYrrnxyOQyGo9MLqPxyOQyGo9MLqPxyOQyGk9juGz5gXaliX755ZdPC0e5JrRu3Vq8JyUlVXOW+TeTkNUHCQkJIpnY7NmzRU1elhuqqKgQGbWZPIzJxAgfHx/xTqddW8/Nz9iWJUn8/f2Rk1O17oVOPZOiaW0DAwNFRu/y8nLRlnWgs7OzzW1dXV3NGb/5Getcl5WVie3jO4Ti0RUHkJyWgUA/H9EvrtcmyFlaWipe5KMtKBbCy8tLPDjg8Vi3JSgo9oHPTlhiiy/tWP38/ERfS0pKzG15bDxm7pN91hbSsy1twD4T7AP3w7bWNuT/NbsRLB5ur72tbWhtb0sbam3XbNiChEXP4LIpo8x2YVt+X7Ohtb3ZJ0sb2mtv9pOfW9qb+7S0obW9tbbW9ra0obW9aSO203jZlp/VZEPyaefG0t7WNqzJ3kRd9ra2Ib9rqe+a7K3ZUDs2e+xtrw0t7W1pQ9qmNs3asje327JhbZrlttrs7agxgvvX7FJfzdZ3jGDftO82eIw4tAGuv76G9NxM3N9uEhaXuaG1nz/mRsVgeFkWfFL2IDUjGZvdfLHDJxwFMW0wLLYjzm/VER2CQ+Hj6mbugyPGCKI+mm3oGEE9WGtWjzGC26w1q8cYwe/Xx96NGSOI+ti7oWMEj9lS33qNEZbjt95jBHmt7a3HGGFL33rcR9jSt173Edwf+8e2eo4RlvrWc4yw1LfeY4T2Xb3HCPLY0qyjxwjy1EezjRkj2E+20zSr1xjB/XG/tuzt5cAxgqiPvR3pa9Rl7yVLlmDp0qU4frxqXb8l7JpZPp1ga2aZhmXCsRdeeEE4uMSxY8dEki8e+IwZM3SfWeaJoMD0xpakPPR+dR1+vaovzmwfqiuXrGOSxfP8l7/i7vPGYdnKvzFl1BmGOCaZXEbjkcllNB5HcJX89x12ffcSFsUNxtseETC5uOKpgSNxZaee8HRzcxiPEW3XXHlkchmNRyaX0XhkchmNRyaX0XhkchmNpzFctvxAOT12IA4erMoueODAySxsfJJx3XXX4aWXXjI/+Zg3bx46d+6MyZMnS+mX9mREb/SM8UeYjzv+OFj1FMcIxySL59+EqqdvMYHehjkmmVxG45HJZTQee7gqc9NQvGUFVq38HHf+vBDDlnyIMxa/j0lfvYNJ7z6CgWv+Qd+4CXjZLRyXduqJPedejeu79qnmKNvDY0TbKR5lO6WH0+c6ksllNB6ZXEbjkcllNB5Hc9kVhn064PHHH8fHH39sdpb79euH7t27Y+HChWjfvj2eeOIJsX3o0KFiuj4yMhK//PKL+L+RwrD53QHRXvhtXxruGxajaxg2v8vj0Ts0QuPROwx77dEqW/AYZYRhs88ywrB5bLSZ3mHY2n71DsNmW/LICsPm/2WEYWth+XqHYVu21TMM2zJ81Gxv2mXfP3D980P8k1eAmyIH4KBHACLKizGwJANupkrkuHqiwtMHHaJa4aI23XFtzwFAcQkqS0qRV1F5yhhhqW89xwhN37LCsLlPGWHY7JOMMGz2SUYYtuX4recYwX3ICMOuSd96jBE8Hh6f3mHY7LO1vvUKw7bWt7OHYVvqW88wbEt96x2Grd3j6R2GTR4eh95h2Owz+yQrDJt9UmHYHvUeIzTdNcswbL3R0DBsnkBesDLw7O/78PDvCch+ZCS8ParPwjgSso5JBk9iTjHi716IzhvexrdffIKOHTvqyidTD0Y6TzJ5ZHIZjccWl6m0GAVfPoyC7SvxXLvxeMkUiP5hEXh2wAgMNhXBpSALLm7ucA2JhWt4S7jYGR7VHGyneJTtlO5O3+tIJpfReGRyGY1HJpfReBrD1eAEXyNHjsThw4exY8cO8VRG4VTwyYQsnNkhAvf+cgTrEnIxsm2I0x+TDJ5f92UCoXFYtXIlogKrnnYaRQ9GOk8yeWRyGY3Hmqsi6zjy59+GfXk5uKvXZVhTUIRH+w3B/3oPgXsj1ycZ3XaKR9lO6e70vo5kchmNRyaX0XhkchmNx9FcdjnLDH1m+AGnvWsC1xC3a9cOzo7GhGFrGQb1DsMOKc9GkJcbft2TiiFxfrqFYTObOPn0DsNOTU0VPHqGWH6z7Tj6RvkgL/UYwv3amPerZxh2RESElDBscrJveodhJycni/3pHYbN5HzcLisMOyYmRkoYdnx8vJQwbCY2dHiIZUU5ggN8kZJbgN9Tj+FIUQE8Kk0I8fJCdOZRFG5Yim+8WuCrmH6IrgCWTz4Hff2CkZeT0+gxwlLfeo4Rmr5lhWGHh4dLCcNm8ksZYdjUnawwbA16jhHkZD/q0mxjxwhy2tK3HmNERkaG6LuMMGxrfesVhm2tbz3DsHn8MsKwNegZhm2pb73DsFNSUoS99A7Dpr7ZHxlh2GFhYdLCsPm5jDBsbpORDdvlhM71zoadmJgojkNaGPb3338v6ix/8sknNbbp1asXtmzZAmfH6V5nWeOa890RFJZViqzYevIYoe5acVkFwp9YjTkt8vHmddOxYcMG9O2rn92cpZZcc+eRyeWMPKbifJRsWIay3X8h+fBWvBrQHh8EtkORq+1nrBFe3ri+W1/c3mMAgr28m7XtThcuo/HI5DIaj0wuo/HI5DIaj0wuo/HI5DIaT5PUWeZTCs6KsGaxVprJErt378b27dvRnKE9QZPFNbx1MB799SDKKirh4ebq1MekN8/vB7JQUFqBUW1D8CaMqQfFo2znKC3w+Wnpph9R+O2zqCgpwNeth+F/raYi32TCVUE+OLciG31KslHApIaR7ZAa2xW+cV3RIyzylEzWjoDR9C2Ty2g8MrmMxiOTy2g8MrmMxiOTy2g8MrmMxuNoLruc5dtuu02EUfAmavXq1TbbaMXUFeRgeJsQMbO8KSkPA+ODlNlrwScbj6NjuC/ahlaFuikoKNiGqaIchd88hawNy/B1h3F4xycOO/NyMDWuLd4ePgEt/E4my/AsLkaMtzc6KWMqKCgoKCgoGBR2OctjxowRYdYDBw60+TnXtl5zzTUwAhqzZpkx+TLWLPPBRRs/f/i4u+K3feno4F+py5plbZ2JM69ZPpRVgq+2peD5Ca2Rl5cotXSUrDXLlmve9FyzzPVAMtYsUw9qzbL+a5aJzUW5+O3wfhzMyYKpIAPFqYdwtMKEjW3OAsqBId6++KbPMIyIjEWoX0A1zZJf04eeY0RNazodvWZZ07das3z6r1nWzrHea5Yt9a3WLKs1y9SgjDXLmu70XrOs8ag1y6f3mmXL3zU91yyznYw1y8XFxVLWLGs+jLQ1y3/++Sc6deokahfXhPvuuw9PP/00nB3OsmaZXOM/2AhXFxcsv6KPrjx6w9E8647m/L+96wCPquiiN72RQCCB0HsH6b0joFJUuqjYC/befgUEQUVQ7AUriooiggKiSJfeBASU3luAQHrP/t+ZzSwvyybZTfZNsrP3fN+yu+/NmzNz974h980ttP5YvLgZP99yiuLTsmjf053p3392iPrcHLNcOn6nkuZRyVUaeCw5OZS1bx1l7FxKP1+MpxeoAp30sdahr5WZRP74z8vfnypWqELX1L+Krq9Zj5pERhWJy53QjUcll248Krl041HJpRuPSi7deFRy6cajkks3HnfHLBepzrJ8oqEjimos46kFnsaogOSauuoIjV96iOLGmVNvWdWc3MUDVR4z7z+asemk7Vi9CiE0c0RT6lyznHiyhYy3yJ6Jp1lmoiT0gXlYdvnpQtaJPZQydxLFnT1Mj8Z0pgXBFakrpdMj2eeoX4CFgqo1ooDG3cmvejOn6x+z3vE9q1oXWO9YdiWhC6x3LDvWO++5Zx3ZgU73gm1puFpXqVJFGBrYam/UqBG99NJLtm1xb4Z0I1DJ1bd+BUrNzKF1R+NN5TEb7uJBbDIM5SnX1aOMyb0paWIv2vtUZ2EoA9BbuGqYbSiXlD4wD8vOkS6kb/iJEj+4nTZl+1GPhiNpeZmq9Hn362j1vWNpxAPvUbl73qeQax8m/5pXOW0os97xPVsSusB6x7JjvfOs+0gll248Krl043E3l1N/GR09elSU2vnss8+EfzpcWfECpkyZQq1ataLz58+TNwO77aq5roopQ9FhAfTngThTecyGO3iwqzx2yUG6qUUlerZHLZEhPCzQj3x9ffLUCx89erR411EfmIdlZ68LqUs+okvzp9D7TW6kAREtqEKZcrRlyO10V6Orip2UkfWOZadaF1jvWHYloQusdyw71jvvvmedSvD1v//9j5o3b06LFi26oj4tAsDHjh0r2syYMYO8NcEX/vDEORUJvmSQPXb3e9cpR7//F0vPdqjg9gRfksfsBF+SpzjJe/bFW+h4fDrd1DhCtLFP3oPPx44dE64V48aNMz3BF47JhEtmJ/hCv5LXzARfuA48Zif4Muq3Wcl7pAzRt5nJe6QMwWNWYg4pw7NJibQl9hQFx4dTyqU4ot1LaeuJ/fRj/eF0PNVCt9SqRx9060f+OdYEd8VdI4yJ8sxcI9Cv5DEzwRfGDh4VCb5w3szkPUb9Njt5j9RvV+RdnDUC51yRd1HXCHv9NmuNyE+/zVgjICczk/dIGTrSbzPWCEf6bdYaAWB8Zif4Muq3mWsEIHnMXiPkXM1eI/Ddkc66e41AX67obHHWCNmn2Qm+APRrdoIvPz8/l+RdnDVCjldZgi8k99qxY4cYqCOgC2TK3rx5M3k6PClmGfhqyym6a+4eOvtid4ou4173Yk+KSXz5z4P09trjdG5s93zrTm/btk1Zgi9PiMvwdh6VXGbwZF84Qdmn9tKBs8do7LkE+jnR+h+BPfpXr0PPt+xI3SpXdyu/J8uuJHlUcunGo5JLNx6VXLrxqOTSjUcll248Krl04ymRmGU8YcjPUAbwZMHb6yzLJzuquQY0iqIAXx/65u/TpvI4g/PJGbTov/O0/3yKqTyOsOC/89S/YYV8DWVv0Qfm0Vt2OfFnKXnuJIp/4waKn3ojffrLh9T2wClaEp9ItyccpE9iN9LK+G20OpJofc9r6Ozoh2nRdcPdbih7ouxKC49KLt14VHLpxqOSSzcelVy68ajk0o1HJZduPO7mcsoNG9i+fTu1bNnS4bn169fbtrcZaoHd5KHNKorEVk90rVFiDy2m/3WUXlpykFIyrTWfx11dm17uU0fJeE7Gp9G2k4n0TPeapnMxGCWF9C0LKOXXN4iyMuhE46vp4Sq9aHVKOvWvVpu+7t6PIrPSKD45lcpVqeX1Dy8ZDAaDwWAwlBnLDz30EF199dX0wAMPULdu3UTpHfh6I7532bJl9MUXX9Brr71G3oyCdt7N5hrToRr1mLGVvoRLdruqpvHkh5f+OECTVxyhe9tXpUc6V6cP1h+nicsOU0x4ED3QsZrbePLDpuPW2I4etQuup4ZM7hMmTBDvOusD8+ghu9SsTIpNTaFT8RfId933tGvvZjpUozstLlub/r54gcoGEn3bexCNqtvYZhwHB6UqM5RLs+xKM49KLt14VHLpxqOSSzcelVy68ajk0o1HJZduPO7mcspYvuOOO+jIkSM0efLkK4xixCs/+uij9Mgjj5A3Q+WOrj1Xt9rl6K62Vejuuf/S7/su0OhWlWlQk2i38zjCm6uPCkN5/NW1aXzuTvLHgxtTjoXoucX7aVDjKKpWNthU2f1zJokqhAZQTHjBMdt4yPPss88quVlLUh+Yx7Nlt+7MCXpjxyb65eh+w1F/okqdiHKIGluI3uzYi25v0JwqBIcUmUdH2XkCj0ou3XhUcunGo5JLNx6VXLrxqOTSjUcll2487uZyKsGXxOHDh2nevHmi9A4Cp+vUqUMDBw4U9ZZ1AcpgwahyNRu2zNamIhs2sjrj3ZihLjvHQtM2nqdPNp2kS2nZ9GyXyjRlUNNiZcM+deqU4HGUVfH0+Yv0zobT9Mba0/R4l2r0UpeK5OvjY8tQd/ZSEnX6fA+1qBxO3w6pQ75kyTcbdmxsrOApahbLO385RPHpFvp5RF2HMpRtL1y4IDwhhgwZYvu9zcqGjXFGR0cryYYNTozN7GzYZ86cEf2ZnQ0b+o3jKrJhQzaVK1c2PRs2jlWvXt2xvC05FHBwPWXsXkHjUnzpA/8oCsvJolsSD1PDjAQqW7k++TfrSTHlq1PD8HJUKaJsvvJG/9WqVTM1062UN/jRt9nZsI36bWY2bKnfKrJhY8xRUVGmZ8PGMfx/ZnY2bFwPvVORDdv4J4uZawQ4MY7CdLa4awQ4Hem3GX9H4P9AjN3sbNgYs71+m7FGONJvs9YI9If5m50N26jfZmbDNuq32dmwz549K+RldjZs6DfGY3Y2bIy5QoUKSrJhY5w4b3Y2bPSHY2ZnwwaknpudDfvEiRNiHkXJho1EwC4by6ivDJfrl19+WftYuKJmw4ai4IdVgYK48HNOWHqIJiw7TNsf7UAtqoS7hefYpTSatvoo/bb3PJ1KSKfU3Njkx7pUpzcHNCA/Qz1jiV/3nKMbvt4hdr0/G2p1FcX4Fvx7nt5ec4xyLBZ6e1BDqhmSVSzZNXpzHfWrX4Hevb5hge1UZsMuLfrAPKVXdrgX0jfModTfP6DTWdl0X0xnWhtUgcb4JtPkkEwKa9CBAq/qSz5BYcXiMQusdyw71jvPuY9UcunGo5JLNx6VXLrxqOTSjac4XI7sQKeMZVjseCqAHUA8JdAZRTWW5VMeFSiMKzM7h6q++hfd2qoyvTWwQbF5fthxhu79+V/KyrFQ99qR1Dg6lMqHBtC1DSpQu+plC+zjvbXH6NEF+6hWZDCNuKoSbT6eQCsOXaTqZYPofEomhQf501/3taYGFYumV+lZORQ6djl9dGMjuq9DtVJjLJcmfWCe0ic7S0YqJX37HGXtXUf/1O1KtwbWpASLD33a/VoaXqfonjqsd+77jcwE37MsO9Y7z7mPVHLpxqOSSzcelVy68RSHq8ilo2666Sa67777CjSUR48eTd4MY2HvkuZC+aRbWsbQt9vPCMO5ODwHL6TQXT/toT71ytPpF7vT73e1oumDGtLYq+sUaigDD3euTj/fepWIYX5j1VE6fDFVGLZHn+9KB5/pQtiPfvnPA0UeI8aHvhtXdH73zdv0gXlKl+ws2VmU/N0LlHFgE03v8gD1tFQmv8AQWn/DrcUylO15zAbrHcuO9c5z7iOVXLrxqOTSjUcll248Krl043E3l1PG8ocffijiTz7//HOH5/fu3UsLFiwgbwb87EsT1x1tqlBsUgb9vvdCsXjunfuvyGr9xbAmVDbY6UpjNsD1enCzisI4jn+5Jx14pguN6VhNHK8cEURPdqtBP++5QAlpRZPfvtyazg2iQqk0obTpA/OUvOzgxJOTmU7JcyfTpydPUv06w+iV0+fpjgbNadewu6hxZJRbeFSB9Y5lx3rnOfeRSi7deFRy6cajkks3HpVcuvG4m8sp6weGMgKjEXSNHWbGlVDlVuAsF2KVG0aH0sL/zhc5M/axhEzhMj17VDMqFxJAxUWEA2N75FWV6LnFB8Q4b24ZUyRjOSLIjyqWKTgTNoDkAEhGpyIbdmnTB+YpGdldSk+jT/7dTu/s3Eyn0+RTzmCi6LbUvGx5+qZ9DxpQo67H/UYquXTjUcmlG49KLt14VHLpxqOSSzcelVy68ajk0o3H3Vz+zmaIhqFcv359h+eRNXnp0qXkzUCWudLGBdfpxcXYWV56NJWC/H1pQKPi73jlh5qRIdS+WgTN2Xm2SMbywQupVC8q1KnEc02aNKE9e/YoSVJXGvWBedTIDjvIWQe30KYTB2jg/qN0yeJD1bNSaGRqLKUGhFKVuq2oX5P2NLh20fMJ5AfWO5ZdSegD6x3LjvXOs+4lvmdZdqx3bjaW+/fvL0pENWyYf7bhZs2akTcDKdZVZXhzlqtXnUj6YP0Jkb26SoS1NIUr+GnnaepbrzyVCXLd/doV9K8XTq+vOS2SdcE4dwUn4tOoeiF1nEvidyqN+sA85snOkpNNGZvmUca/f1H28V00xyeCHq7YjipkZ9BPZ9dS6/CyVGbg4+TfoDP5+Lqm466A9Y5lVxL6wHrHsmO986x7ie9Zlh3rnfNwygq69dZbbTW/8sOWLVtIB6BE1nXXXVekOss4p6LOshxDYbXP2ucm4Fq19xTd0LSSSzVUL2UQbTyRRO9cV0G0MauGKo63KO9LaVk5tP1UAjUu5+tSfcSjcSnUuUa46KegGqr4/Pfff9OAAQNo9erVVKtWLdPrLOO8ijrLaCd5zayzLOVpdp1lOTdVdZaLWx8xYe0csiyaaltDvqjWmZ4OrEodQ0NodrtOFFH+AUpITqWyNWpQook1VGWdZSkXs+ssY2zyWjPrLBv128w6y/KzqjrLZtZQNeq32TVUZZ1ls2uoyjUCMLOGqlwjMGejfptVZzk//TZjjQCvmTVUpQwd6bdZdZZdlXdR1wj0h/GZXWfZqN9mrhFG/TZ7jZDXmr1GgMeRzrp7jQCPKzpb3DrLaKeizjL6VVFn+ZIL8nanreFKnWV7OFU6qkqVKnTu3DnxwoB0RlFLR0E55cJgNpzlwk8bMX4ljbu6Nj3Tw2ocOovPNp2k++f9S2de7E7RTsQDFwcXE5Op0usbaPrABvRQp+ouXVth4ip6ulsNeqFX7ULbqiwdVRr1gXncI7vsnBzKtljIPzuDkn9+jbK2/0bZFWrQrk630AvnkmnDudM0uFYD+rHPDeSfu4us22+kkks3HpVcuvGo5NKNRyWXbjwquXTjUcmlG49KLt14isPlyA50amcZxnL79u0LNJR//PFHGjFiBHkr8CSmtHHhiU+D6FDae8719Onz95yjzjUiTDeUgdBAf2pZOZw2HU+ghzo5f11KRjbFpWRSNRfcsL1ZH5ineLI7nHCJXtryF313YI/haBhRneHWj3usJdDe6tibHm3WhvwMfev2G6nk0o1HJZduPCq5dONRyaUbj0ou3XhUcunGo5JLNx53cznV0+zZs4V1Hhsb6/A8trdfeOEF8maU1tphDaNCaW9ueSVnkZ1joVWHLlLvWuHK5tO+egRtOm51JXEWJxOsrhrVyroej+2t+sA8rssOdZE3nThIDWZ/Igzl8tnp1Dn1HDXISKAGwcEUExJG3WKq0f2NW9KJWx6kJ65ql8dQ1vE3UsmlG49KLt14VHLpxqOSSzcelVy68ajk0o1HJZduPO7mcmpnefr06cJfvHnz5iLm0x779++nI0eOuG1QDPehYXQYLTtojSVwFntikykpI5vaVQlT9lO0qxYhkpHFp2U5Xc/5ZLw1/qA07iwzPB/ZsYcp9bd3aO3hPTSkcg8KteTQqpN/Uu1yURR6/bPkX6+9qcm6GAwGg8FgMBglC6eskhUrVtB///0nPn/11VcO26gox1OaUVrT8KPWcmxSBl1KzXS6VvLaI5fIz9eHutWvRKrm06yStR7a/vMp1LZahFPXncjdWa7qZKZv1FjeuHGjePdWfWCewmWXk3COkr58jLJP76MtQeWpf5XeFEHZtKVeNapz52/kE1zG638j1ju+l0piDWK9Y9mx3nnWfaSSSzcelVy68biby6ltkauvvloYzHC3dvQ6efIkRUdHkzdDZlkrbVxwwwZciVtefjBO1D72y8kkVfOpWyHEZiw7i9MJ6WIXOjTQucLjyHiHWst491Z9YJ6CZZd18j9KeGeUMJRX1u9Ng6r3pZoR5WjfbU9Q3T53FslQ1vE3UsmlG49KLt14VHLpxqOSSzcelVy68ajk0o1HJZduPO7mcspYfuaZZ0QW4fxQuXJlev/998mbgXTrpZGrQbTVlXqfk0YoMmivRLxy3UhlcwIPdr2jwgLowAXnjeUziRlUyYUEZMeOHaPHH39cvHurPjBP/rKzpCVR0leP4yag1/s8T0OyK1BEYDCtGDiKokOK94BFt99IJZduPCq5dONRyaUbj0ou3XhUcunGo5JLNx6VXLrxuJvLKWO5Ro0aolbWhQsXRPkdCcQpnz9/XnweNmwYeTNKa4a3sEA/igzxp1O5LsuF4ejFNDqXnEmdapZTnh2vXoVQOnDBWh/OGZxJyqCYcOeNZejq559/btNZb9QH5nEsO0tWBiV9/yLNtYRS9arX0RuHDtI11WqLhF01w631yll2rHd8L5XsGqSSSzcelVy68ajk0o1HJZduPCq5dOMpkWzY2MoeM2YMVapUiQYPHmw7jgzZQ4YMEQnAvB0o4F1auWLCg+h0onPG8taT1iLebaqGK5uT5KlfIcSlneWzSa7tLKtEadYH5rmMhbEnqMevsyjw87coIrMK3VOxAyVlZ9NDTVrT4uuGU6Cfcy7+3vYbqeTSjUcll248Krl041HJpRuPSi7deFRy6cajkks3HndzOWUsoyzU999/T6NHj84TMA3j+csvvxRu2m+//TZ5My5dulRquSqHBwqXZWew7mg8VS8bJAxsVXOSPNhZdiVm+UxiuhhnaURp1gfmIVp07CD5zJhCt65YSKvPnKSs3ASFzctHU+zoR+j9rn3dmrRQt99IJZduPCq5dONRyaUbj0ou3XhUcunGo5JLNx6VXLrxuJvLqWzY8+bNow0bNlDjxo2pU6dOec7VrVtXuGlPnTpVxIN6Oo4fP07XXXcdDR06lO666y5KSkqi7Oxs8vf3F4mhEhISbLvqQGqq1W0Yic5wDm39/PyE23p8vLVucHBwsPjDW7bF0w7U/8rKyhJt8QBC/qhoC9cBWR8M57CzD997HMe1cgxBQUFiXMnJyeI7ODMyMsQLfOXKlaOLFy9S+SAfOhWfKvrAfOzbApGRkWIMv/0XSz1rRYi2kicsLEyMNT093dYWc8OcUVIMY05MTLS1hQxkYD3GgH7Q1l6G+CzlBtQpHyxcwI+eOU/lw4IKlDfkAGM5wi9btLGXt1GGUt6yL8RlQy6yLX4DKUN7eZctW9bW1ll5y3HivFHe6NMoQ8gbYwkMDBQv2dZe3kYZ2ssbMkI7yYu2OJefzoJP/jaYm9Rve511JG8pT0f6XZjOGvU7P3lLGcq5OSNvZ2VolPesf7fQo7us4SQt0+Poh9NrKKL/YxTUtK9VhsEh+eq31FlHMixojcCxguTtrjUC/Uu5uKqzaIv5FrZGSHljbPJaM9cIo347K29HMnRmjQBc0dmirhEYs73OmrFG4Ji9zpqxRuB6V+RdnDUCcEXeRVkjIEPM2ajf+elscdeI/PTbjDUCvPbyNmONcKTfZqwRjvTbrDUC/WF8aGvmGmHUbzPXCKN+m71GyGvNXiPA40hn3b1GgMcVnS3OGoFxop3UWbPWCPSHfh3JO8iNawTgirzdaWsUJu+5c+fS/Pnz6fTp02QPHwtGXAg6duwojGWgc+fOtG7dOts5kFWoUMG2kHg6rr/+evr1119dvg4KpyLLclG4nlq4jxb+d572Pt25wHYn4tOo+mtraPaoZjSyRYyyOUmeTcfjqcMHm2nbI+2pVdWC3Scys3Mo8MXl9NnQxnR3u6pO8Zw4cYKmTJlCzz33HFWrVo3MRGnWB2/ksWRnUdqqmfTn6p/ohio9KSI7g5aeWkb1wsKp3FM/kU+AuR4Kniy7kubSjUcll248Krl041HJpRuPSi7deFRy6cajkks3nuJwObIDndpZhgWPpwaw8u3xzjvvCIu8S5cu5M3A04nSylU5IkjswhaGNUesT5F61y1fJJ6iQvJUKxss3k8mpFOrQuxf7EADrsQsw0B+8803xZMsb9YHb+GBgZwy9xXK2PEHUXYWHfcPpfur9KYqWSm0PcqXIjo+R9Swu+mGsifKrjRx6cajkks3HpVcuvGo5NKNRyWXbjwquXTjUcmlG4+7uZyKWX7iiSeoZ8+eNGvWLNq+fTsdOHCA/vzzT7rjjjtEvDK298ePH0/eDOmeUBq5kDE6IT2bUjKyC2y352yyaBuda4CqmpPkgeHr5+sjjOXCII1/V4xluF2gXrh0+fBWffAGnqxj/9ClFztSxrZFwlDOIh+6rc4ASgyOoAUjHqDoEeMpqNV1lJLuXCy/N8mutHHpxqOSSzcelVy68ajk0o1HJZduPCq5dONRyaUbj7u5nDK7Bw4cKEgRkwyf74YNG4rj8OBGvPK7775Lffr0cdugGO5FTK5BCQOzToX8XRKWH4yjdtXUZaqzBwxljPVkfOHG8vncneWKLhjL+/bto2uvvZa2bt1KrVu3LtZYGaULRxPj6Y8Th+mdnZvpQd9E6rDjV6pJPhTa8w4613YwNf11NqVkZdK3va+l1lExJT1cBoPBYDAYDIYHwOk96pEjR9Lw4cNp8+bNIgkWgrpr1apFbdq0MXeEHgIEsZdWLpkxGnWJ8zOWD11IobVH4+nrEU2LzFNUGHmqRAQ5tbN8PsVqLFcIDaDSiNKsDzrx4IHd1B2b6LlNK23HHsY/Na6zfjmWRHTsG/Hx7U5X0831mhSZqzjQjUcll248Krl041HJpRuPSi7deFRy6cajkks3HpVcuvG4m8ulis3IkNahQwcaNmyYyBbNhvJlGDO9lTYulI4CCiof9eu/5ynY35eGNKtYZJ6iwshTNSKITjlhLF9IzqBAPx8KC3RPHVxv0gedeN7btdVmKD90aS8tPrmcRvlaM0FKBPn5iZrJjzVvWyyu4kA3HpVcuvGo5NKNRyWXbjwquXTjUcmlG49KLt14VHLpxuNuLnWR1poDPwoSoZVGrsiQAArw86HTBST5WnXoIl1VuUwe41PVnIw8VcsG0erDl5zaWY4KC3RrLVxv0QcdeCw52bR6/nR67Hw2heVk0b9Hf6XIq/pQ2LCP6NqAIPrOjVzugG48Krl041HJpRuPSi7deFRy6cajkks3HpVcuvGo5NKNx91cbCy7CSqNNle5fH19RCIsRzvLcGN9869jNH/POfp8aONi8RQVRp6qTrphX0jJdNkFG5nxUOZMRTa+0qwPnsyTfeEEpa3+htI3zqUnq/QmCq5Ai86sosoPfk7+1Zu5lcud0I1HJZduPCq5dONRyaUbj0ou3XhUcunGo5JLNx6VXLrxuJuLjWU3AQWxSzNX5fAry0edjE+jkd/9I2KVe9aJpDvbVik2T1Fg5EHMclxKJqVlZlNwgF+BxnKUi8byVVddRefPnycVKO364Gk8low0SvriYco6sl18f7dsQ9oWXIE+7NCDerR4zq1cZkA3HpVcuvGo5NKNRyWXbjwquXTjUcmlG49KLt14VHLpxuNuLpdilhn5A3WoSzMXSkKdtttZfn/9CWEo1ykfQl8Nb3LFUxhVczLyYGcZKCxuGdmwK4S5ntyrJOakC1dJ8qQsestmKC/o8RBNqHAV1QkvR3c3a+d2LjOgG49KLt14VHLpxqOSSzcelVy68ajk0o1HJZduPCq5dONxN5fbjOVu3bq5qyuGCYix21k+EZ9GH6w/Tve1r0oHn+1CNSNDSoXcq5YNFu+FuWIXxQ179+7dIikd3hmeg6zT+ylj48/kW7s1XdPhUbr9eKw4vuHG0RToVzoTvDEYDAaDwWAwPB9Ou2GfOHGCli5dSqdPn6bMTGvZHgmUklq/fj15M4KCgko1V4xdzPKve85RamYOTbmunlt5igIjD9ywnTGWzydnuOyGnZ6eTocPHxbv3q4PnsST+utUOhBWkdr71CU6d1oc2zviXooOCXU7l1nQjUcll248Krl041HJpRuPSi7deFRy6cajkks3HpVcuvG4m8spY/mrr76i++67j7KysvJtU1qzEgOfffYZ7d+/n5KTk+m6666jAQMGuJ0jIEBdvd+icMEN+2xSBuXkWETCrxWHLlL76hFULiSgxOdk5IkI8hMZuU+ZsLOsEqVdHzyFJ33bIvrzzAkaWrmH+N6lUlVaff0t5Oum9UZn2enCpRuPSi7deFRy6cajkks3HpVcuvGo5NKNRyWXbjzu5nLKDXvChAl055130urVq4XRid054wvHIyIiqDRi7969NGPGDJoyZQpNnz6dnnzySbp0qfDSRK4iKSnJ7X26kwtu2Fk5FopLzRQG88pDF6lXnUi38xQFRh48dBEZsePzN5ZTM7MpJTNHlI4qrSjt+lCaeRITE+mL/3aSz4wpFLxll81QHte6M6254Va3Gco6yo71jmXHeudZ9xLfsyw71jvPupf4nvU+2Tm1sxwaGkqffPJJvudr1qxJ48aNI7Nw6NAhWrRoEf3++++0cuVKunDhAgUHW2NbZfmj1157jX755RdhbFWrVo3ef/99iomJofnz51OXLl1sTxkaN25MS5YsoREjRpA3ATvLAFyxkegLCbJ61S1PpRFyF7ygXWWgNO8sM4oG3MsDVi6gDefP5Dm+5vpbqEtMNRYrg8FgMBgMBkMZnNpZ7tSpU6FtrrnmGjJzd7hFixbC2E1JSbni/KRJk+iHH36gFStW0IYNG6hSpUpiPHAbx853VFSUrS0+45i7UaZMGbf36U4uxCwDSPK14mAcBfr5UOeaZd3OUxTY81QsU4ixnFw0Y7levXq0cOFC8W42Srs+lEaeHIuF2vw8UxjKwTnZtP747xTfqzdl3fOMaYayLrJTzaOSSzcelVy68ajk0o1HJZduPCq5dONRyaUbj0ou3XjczeWUsYx45WnTpuV7Picnh66//noyC4gz7t69O7Vs2fKKcwkJCcLF+vHHHxc74MALL7xAO3fupO+///4Kn3XsPAcGut99NyMjf+OuNHBVCrcGup9JyqDNJxKoddUICimgjnFReYoCe55KZQIptgBj+XzuznKUi6WjECrQs2dPJSEDpV0fSiPPxK1r6e8LZ8Xn42f+pPbPzqGI+u3Iz9e8Cne6yE41j0ou3XhUcunGo5JLNx6VXLrxqOTSjUcll248Krl043E3l1Nu2DBG4c78008/2QxSI06dOmXKbq0z+Ouvv0Tirg4dOtiOwQ27atWqYhexffv2Ilu3xLlz56hu3br59oe2MM4Bf39/Gjx4sHjhM+YO4xwICbGWWkpNTbU9MMjOzhYvPz8/8UQjPj5enIPLOIx02RbGGnbIsfONtuHh4bY4arT19fW17aDjXFpamshAjuO49vz580IJkOkN48L8AXDiOF7gQ0FuWWcMbcOD/OjQ2Ut09EIyVY8IFtdJZYqMjBRjgBssHibgJXnCwsLEWGUWabTF3DBnPIzAmBFnCqAtZIAxAxgDZIa29jLEZxyXPGXLlhUxBuF+2WIHHP04kvfRWOucAnPS6eLFdIfyNspQyhsZ3RG//swzz9j6Q1tcL2VoL2+MyShDZ+WNcUKGMmYCbdGnUYb28pZt7eVtlKG9vCFDfJa/I9riO2TnSGfBJ38bKW9HOmuv35AhxguewuTtSGeN+u1I3sf3bqQJ27aIY/8cXUh+wydSml8o+WdkFChvZ2VolLdRhpANvuens47kjeOOZFjQGoFj+J6fvN21RqB/jEG2dXWNwHyNOivbOtJZ8MtzZq4RRv12Vt6OZFjYGiH121mdLc4aIcds9hqBYxiPUWfNWCNwPTidlXdR1gi0hQwxPqkPzsi7qGsEOI36bdYakZ9+m7FGIIQN7Y3yNmONkGMuTN7FXSMc6bdZawT6Aw/amrlGGPXbXmfduUYY9dteZ929Rsi/8VyRd1HWCOg32tvrrLvXCIwZ7Z3V2eKsEfJvLqmzZq0R6A9zciTvIDeuEQA4nJV3cdYIqXeFyfvXX3+lWbNmiXOQE6o+2cPHghEXglq1agnhOzKUgbi4OBuxmXj55ZdFsjGMRcYsv/322/TEE0+IHwrCkoCRjPHAPfumm26iLVu2CKE1b96ctm/fbruJ7YEdcgjOVYAfP6wKFJWrwbR1dH3jKFr433nq3zCK3hrYwBQeV2HP88nGE/Tg/P8oY/LV5Od7ZTIn1Id+YuE+Sp/U26Us7Nu2bRN1lrdu3UqtW7cmM+EJ+lAaeLD8pK/5jq7buoVWhsbQkpPLqG3ZSIp88gdSAU+WXUnyqOTSjUcll248Krl041HJpRuPSi7deFRy6cajkks3nuJwObIDndpZHjRoEL333nv5nofxisRZJQH5VMS+nhaM6bNnz4r41AcffJAeeugh8VQBib/yM5SLA1U/fnG4ELd8KC6VDl5IpfpRoaVmTvY8cMPOsRDFpWRSdG6stRFIToYay6W5XJkn6ENJ8WT8+xclz3zC9v2YfyitrDGA7vVNpt4Pf0p+0TVJFTxNdqWFRyWXbjwquXTjUcmlG49KLt14VHLpxqOSSzcelVy68bibyyljeezYsQWeh/FZlN1Yd0Am78K2vdEIxi5yxYoVxee77rpLvJyBdMMeOnSouMZZdx7skKGNCjdsjBHHXXWNqBgWIHaVUUKqTeXQQt2w4YoAHrPdsOEaDx7pGhGcY+U5lZBK/pnJV8j7RFyiMJbB6ayLJT5LXsxRysUsN2z0Ad1U4YaNfmQcvplu2Hj4hOPFcZ/y372EshZOzXPPvRJt3eV/6Np7yC+6Bh07dsxp/S6uGzZeyJpvths2rkd4iNlu2JAVQlBUuGFjrBhbfvJ21xph1G8z3bClfqtww4YsKlSoYLobNl5IeGm2Gzb6g96pcMMGpEOcmWsEjmNshelscdcIyMiRfpuxRsALEBxmu2Fjfvb6bcYa4Ui/zVojwIH5mO2GbdRvM92wjfpttht2bGyskInZbtjQb7Qz2w0bsihfvrwSN2zMH3My2w0b44GszHbD9jHouNlu2CdPnhRzdlbec+fOFSHHRXbDlsAPgBhhGGu4CerUqUMdO3YkVXDkho0yUMh8vXv3bmrSpImtbY0aNUQyp6+//tolDp3dsB/9dS+9t+44hQb4UvzLPcnfz7dUumUcvJBC9aauo6X3tKar611Z3uqm7/6hc8kZtOzeNi7xsBu2e3+nwnA2JZle3rqG5h3ZTy+26kT3Zpyh1B/Hi3PB/R6ggDpt6JuENLprwyrqW7UWLRkwskg8xYFurkcsO5Yd651n3Ut8z7LsWO88617ie1Zv2RXZDRv44osv6Omnn7Y9wZDA02S4NpuZDbsg9OjRQzxRQMkoaSzjaQIMeuwOq4IZGbbdzdUw2up63aZqRKGGcnF4XIU9D9ywgfwyYsNQjgpzfWy4aUaOHKnkRvUEfTCTJy4tlWJmvW/7/ui6pfQoEbWr0ovuadySatTvRLP276ZvD+wR5+f2vbFIPDrKzhN4VHLpxqOSSzcelVy68ajk0o1HJZduPCq5dONRyaUbj7u5nDKW582bR/fcc49IjvXAAw8IAxlb13B52Lx5M40aNUpknu7VqxeZiUOHDon3gwcPUtOmTW1b/mPGjKHp06cLQwhb85MnT6ZGjRpR//79XeYoqhu2PKfCDVtmoXXVNaJjNWvNsRrhfqK/glwsjdluzXbDljzSNQJcIf6+dDohzTZ+o7zPJqRRo6hQl92wMZYPPvhAuM+Y7YaNa2V2YrPdsCFvyWumG7acqzPuU5O2rBGfK2el0h0JB+m18s3E983BUbT58Amiw3Ns99z2QbdQVnIKXUpJFeN3Rb+L64aNNiqyYQudVpANG2NTlQ0b18trzV4jJI+ZbthyrircsDEmwGw3bLRRkQ0bPKqyYWPuZrpYyjUCYzDqt1lrRH76bcYaIX8rs92w8VvZ67cZa4Qj/TZrjZCZy812wzbqt5lu2Eb9NtsNW/6fbrYbNn5XFdmwZTZnFW7YMnO52W7Y+I5+zXbDLlOmjEvyLs4aIfVOmRs2XK1hLOPlCOvWrRNxzcuWLSMzMHHiRJo5c6bNWMaP0axZM5o9e7ZI4AXhgn/x4sVCoIhV/vjjj6l69eouc+nshg18t/0M9alXnio6SJzlTh5X4Iin9pQ1dFOLGHrt2npXtI+ZtJoe7FiNxvWp4xIPbhS46+NBi3TjNwueog9m8Gw5dZjaLfyR6mQm0trjSyisfjs6U70lfZYVQm+cOGlr1ySyAi0fMIoqhYYViUdH2XkKj0ou3XhUcunGo5JLNx6VXLrxqOTSjUcll248Krl04ykRN2w8pcjPUAY6d+5ss9jNwLhx48QrP+DJwGuvvSZejIJxc8sYjxARXLEduWFnZedQbHIGVYnIm/3cGezZs4fatm2rpHSUtyInJYFmf/MyUWQT+ih2E5Xtex+FXH0PhaNeOxG9jjp+OTmibWBucigGg8FgMBgMBqM0wumY5cIgt+u9FXAJ0I2rJHmw833WgbGMY/CFqFwEY1klvEUf4JiS/N3zZElPpbCbXqF9s1+mNyObUM3MJOo5+EkKbHltnvZw2SnMSPYW2Xkyj0ou3XhUcunGo5JLNx6VXLrxqOTSjUcll248Krl043E3l1PGMlydkYl6/PjxV9S2hX/7Cy+8QDVrqquNaiaKGrMs4wpUxCxfuHBBtDMzHhEvyWN2PCJ40YcxjqB8kA/tPndlzPLek+fFe8VQ/1JdOgrzAq+KmGWMR96XZsYs4xz6zNP2p5eIDm4UbeInXk3bQysTxcTQu+16UHLNNpQaH+9SPCLG74p+FzdmGdwocWJ2zDLao8SJ2THLsmSeiphlzEP2a+YaYdRvM2OWpX6riFnG9eAwO2YZPMjRYHbMMtpD71TELEMO8pyZawRkZtRvs9YIyE/ymB2zjDGC3+yYZYzRXr/NWCMc6bdZawT6wtjNjlk26reZMcviQXcuj9kxy5g7ZGR2zDLGDE6zY5ahP+BUEbMMoK3ZMcvghUzMjlkOCgoSnCpiluXfkspilhHnCVdrEHTp0sVWlxSG5dq1awXp6tWrhYurp0P3mGVP4XnxjwM06+8zdPT5rnmO/7rnHN3w9Q46/WI3igl3bXeZS0e593fC0nHphXbWL/5BRFnpdF2VXrQxOIoy73mG/HNrOBaXx0x4w73k6Vy68ajk0o1HJZduPCq5dONRyaUbj0ou3XhUcunGUyIxy0iItGbNGnrsscfoxx9/tBVMBzp06CAyUetgKDNKV8yy1eXakseb4WRCOvn5+lB0EUpHMdyL1MXvivdLw16hfys2oPjEi7Rx7TLqU7VmkQ1lBoPBYDAYDAajtMCpnWUjzp8/T0eOHBFb2nXq1KGoqCjSCUXdWWa4Fz/sOEM3fb+LLr3ck8oGX36m89ive+mP/Rfov6c6s8hLGImfPkDbExKoe/hVeY5/33sQ3VTPWvOcwWAwGAwGg8HwBBR5Z9kIGMeODOQ333yTnnrqKfLWmGUAMQEqYpZPnDgh+jc7ZvnMmTOindkxy3gAAx5jHEEZX2sd0v0nz1Hd8sE2ea89HEdNKgSJNq7EI8q2eK9SpYqp8YgAZIVYWBUxy+hHFl83M2b57Nmz4jjaBl88SmkHt1D3OsPy3D+vtexMQ2vWz7denzMxy7gHndXv4sYsoz1iic2OWUYb1Kc3O2bZXr/NXCPAiXHkJ293rRHoB324Im9HMixsjYiNjbUdNztmGXNArKXZMctoj1his2OW0QZ6pyJmGS/Mx1l5F3WNwHGjfpu1RuCzI/02Y42A7uE6s2OWIQt7/TZjjXCk32atEVIvzI5ZNuq3mTHLks+Rzrp7jTh37pw4bnbMMmSMz2bHLEM2+KwiZhn8mJPZMcty/mbHLPv6+gouFTHLp06dEvymxSyjXjICo0eMGCG+Hzt2jArCyZMnacCAARQXF0eeDo5ZLh0xDHvOJlHT6RvorzFtqWutcuLY6kMXqceMrTRrZFO6pVVll3n27t1Lt956K82aNYsaNmxI3h6XURyelMXv0X27d9O3EbVFveR/ht1NvnbJ/9zBYzZ0i9Nh2bHsWO88617ie5Zlx3rnWfcS37N6y87pneUhQ4YIq7tbt25UuXJlEbMsnz4wHEM+GdaJqyR5ELMMyFrLKRnZNPTbndSjdjkacVWlIvHgqdiWLVtsT8fMhK76cD4thQ4e2U2P7ztKGyJqi+M/XH2D2wxlyaMK3nAveTqXbjwquXTjUcmlG49KLt14VHLpxqOSSzcelVy68biby6Gx/MQTTwiXSxjKMokXrPNmzZo57ARb3V988QV5M+AeoBtXSfJEhgSIRF6y1vK2U4l0PjmTpg9sQAF+pT95lI76sOFiLPVdPMf6JSRavN3T6CpqVt762V3QUXa68ajk0o1HJZduPCq5dONRyaUbj0ou3XhUcunGo5JLNx53czk0llFT2YgbbriBBg8eTNWqVcu3I8QkeDNkDINOXCXJ4+vrQ9UigujAeatHw4qDcRQR5EfNY8qQJ0A3fUC0xutb/spzrHpYOE3t0MvtXLrJTkcelVy68ajk0o1HJZduPCq5dONRyaUbj0ou3XhUcunG424upxJ8PfLII4XGgv7888+kA4qa4AvB4zinIsGXHIPZCb4kj9kJvuR3+6D7TjUiaM3hC2IOi/49R73rlKPEhPgiJe/BZ8mDOUq5mJXgC781zqtI8IV2ktesBF8bE+No2bkzNP7CTnq8XW8K6T5aXGtJSaXUHIvbkvfI8Tsr7+Im+IJszEzeI2WIY2Yn74EM0b+Ui9lrBMYmrzVzjTDqt5kJvuRnFQm+MGYzk/cY9dvs5D1Sv81O3iPXCMDM5D1yjcCcjfpt1hqRn36bsUaA18zkPVKGjvTbjDXCkX6btUagP4zP7ARfRv02c40w6rfZa4S81uw1AjyOdNbda4RMyqciwRfGiXZmJ/hCf/klZg1y4xoBuCJvd9oabk/wZY8+ffrQ0qVL8z2/c+dOWrBgAb344ovkrQm+oAgyI7HZUMVV0jwv/XGAvt52mv5+tANFT1pNnw5pTHe3q1pkHiSgw2+L3xiZOs2ETvqAJaLhj5/S/viLFO9/kiLustZXNgs6yU5XHpVcuvGo5NKNRyWXbjwquXTjUcmlG49KLt14CuJasWIFTZs2jZYvX56nKo8r+OOPP+jtt9+m1atXC2MXPOPGjRMGY9u2bU0Jr80oBbIrih3oVPBnYcm9mjdvLoxlbwaeUOjGVdI81coG06nEDPpjfxzhkc51DYtX0xsG8siRI003lHXTh/8uXRCGMhDcZiCZDZ1kpyuPSi7deFRy6cajkks3HpVcuvGo5NKNRyWXbjwFcfXq1UuEyMrd0qLgmmuuoRtvvFHYeJJn4sSJYtdWli/TUXZudcNG2SjUvwX27NlDvXv3dtgOW+ZHjx5VKoDSCCisdDnRhaukeaqXC6bsHAutOnSRYsIDqUpEULF4EFf/1Vdf0R133EHR0e5NSqWrPmBX+YeD/4nPX51dTwFNnySzoYvsdOZRyaUbj0ou3XhUcunGo5JLNx6VXLrxqOTSjacwLnfs0MJt2p5HHtNZdm4zlr/55huaMWOGSPYFwsOHDztsBz/y+vXr0/PPP+8yOYNREJDgC9hyMoFicktJFTce/dlnn6Wrr77adGPZU5F98TT5RVqz4OdcOkNbMnJowra14nuflNPk468u7T+DwWAwGIySB8p3/neuaGU3ExJSKCLFz63jaRQdRqGBrvWJONXnnntOhI6i4k9MTAx98sknVLt2bXrmmWfov//+o2PHjlGVKlWEDYSkxtgwGDt2rNhoGTNmjDAk4fqMPuCu++GHH4q43RMnToh2cGGeOXMm/fXXX7R582Y6cOCA2K1FDiQJxN3CZlq/fr34jnhiuD8PGjRIfH/nnXfovffeo379+gn7au3atbRx40bq2LGjGJd9lmeEyX777be0bt06ESMM9+kWLVqIczt27KCXXnpJcB45coRuuukmev31193wC3gXCo1ZPnTokIhF/v7778kbUNSYZYhRJmcwG6q4SponLiWTKkxcJT5f06AC/X5Xq2LxbNu2jdq0aUNbt26l1q1bk5ko7fpwKjmRooJDKdDv8n82aet/pNRf3iAKCCbKziTKyaavyjWgJ8u3oHfPbaZbE49Q5OtbyGyUdtkxD8uO9aFk7iOVXLrxqOTSjUclV2nl2XYygdq8t4lKC7Y+0p5aV41wek44h/xLXbp0EcYrjMd69epRq1atqFOnTvTTTz/RP//8I5JLoUzutddeKwxWCWk4b9q0SRjTMEJxLeJ+YdQCMIxh3MK9GUYzjFrEBP/vf/8TSaOQWAruzT179qSoqCj68ccfRYIpGNcwpjds2EDt2rUTfXXt2pW2bNlCf//9NzVu3JjOnDkjjHoY0vfdd59oAwP+zjvvpDlz5tCwYcOEhy8qF8GQ37dvn9h5xjUPP/wwPfXUU8LIx2YR+sXfwsY+MC4pO4yvVq1a4pw33rPXO7ADC82GXadOHa8xlIuTDVs+HVKRDRtPsNC/2dmwcXOindkZ6uDuDx77DHURBveJlpVC82QELO3ZsCGrChUqKMmGjX6kO05hWSxPJCfS+7u20jt7d9CA6nXpm059IBDy2buaLDCUgUzrb5fi4ycMZeCWxCOUdvO7Yn5mZrrF+F3R7+Jmw0b7SpUqmZ4NG22qVq1qejZsHMN/5CqyYYMT48hP3u5aI9AP+nBF3kVZI2JjY23Hzc6GjTkgd4LZ2bDRvmLFiqZnw0Yb6J2KbNh4yXg6M9cIHDfqt1lrhMysW5i83bFGQPdwndnZsCELe/02Y41wpN9mrRFSL8zOhm3UbzOzYRuzbzuTDbtSQCatuL2xGL+FLJSRnmEbU0pqCuVk55Cvny+FhoTaxiDdeS9euiiOY0zgQL/2bQODAsnXx9c2htCwUMGBOfj4+lCZsDBKTLzctk7ZAJtMjfqNz46yYS9btkwksoIRC9mj388//1xcBxnCyMRxyBC7srt27RIykjLE7wibAHPA3ygwtOGhiN1hGLhS3sBDDz0keKFLHTp0EPqGcNZGjRrRqlWrxK4zdoulvo8ePZomT54sjPjvvvtOyBsYMmSIMKoxRvTXpEkT2r59u63yjtRjGMCYL/QQYYawXxYtWiSMPuxGY3y4Bg8K5MYR+sKcpB7KBF/QLzmPovztG1TIGgFO6KuKbNinTp0S/O7Iho0BuwUjRoyw6IBBgwYV6bq4uDi3j6WkuUoDDz33p3glpGUWm2fr1q3wohDvZqO06gN98nqeV+QXb1kunT9piXuuje2VNHusJeXwDsvgmVNEm9rvv2Q9dy7WogKlVXbMw7JjfSjZ+0gll248Krl041HJpRuPSq6CeKZPn26JiYkp8Ppz585ZZs2aZWnevLmlR48eec7VrFnTMn78+Dw8xmPA4cOHxd+YK1asyPfY1KlTxfeLFy/m6X/o0KGWBg0a2L6D//bbb88zJ3lM4ssvvxR9GbFr1y5x7IsvvshzfMeOHZZp06aJc7jOvo+CeLxN7xzZgQ53lidNmiQs6/fff188Nfj666+pIMB6/+WXX8ibgacTunGVBp6l97QWib3Cg4o/FjxxRKI6vJsNT9GHi5kZdPfMyfRp7vfwMZ+Rf62WNHTJPJqXZo3Q+KNWRYq4aT6lBOaNk/F22Xkzj0ou3XhUcunGo5JLNx6VXLrxqOTSjUclV0E82K28cOGC2GGUHiPG5K+PPvqo8A6AyzLcq5G4uCg8hUF6N8BjAbuhxj4duQy7yiV3m7ErC8BNHDYddplvvvlmevrpp0vtb1SauRz2NHXqVLFFjWB1BMAjKRKUqaDwZlU+6KUV0m1CJ67SwHN1PfeVeUIsyZIlS8SCqLs+ZJ3aR5bEcxTQ0Op2A0zabI3/rpqVQvNPraJ2Na4T3+eFVacJfjuoZtNutCGgHN0/5zPac/EChfj507e9B1H92g2sPIoy3pe07JiHZcf6UDrvI5VcuvGo5NKNRyWXbjwquQrigQs0XIzhniwTaQGIPb7nnntEbK8MOf3444+LzFMYEA8NwC0boSsSMM5hpBeXC0nKYCTC5RpzGzBggIi/7d+/f6n/jUozl8M6ywgWh988DGUAQn/ttddEEWxHL2Rhg/+4N0PGxOjEpRsP4hMQa6KizFlJ60PiR3dR0pePUdq6H2zHxv69QbxPOf831c1KotN0RHzP8fGhZjUHURu/2tRt0Q/CUL65XhNKuutJGpxrKOfHo2o+ns6lG49KLt14VHLpxqOSSzcelVy68ajk0o1HJVdBPEjYhQ0TxB0jozRif5EJG7bO3r17bZuBiK2FwYm/E/FCDC12o7GBKPuXsbE4JuPWjfzGccjPsl2PHj1E8q5XX33VFkOLrNlI5GXc9cV18mU8ZuST2L17t223GgnAkMwL8dQHDx4Uc5BzQzZsAMdkXLCMQz558qStPxyTscHuRoKH6p3DnWWZ2U0Cmdfat29v29Z3hIJcFhiM0gA8ZVOVDbskkX3uiC1JV+qvUym480hKWfwutU87T6f9QqhfymkK7vcARfa+m95as4ie3LNLtN0Xb03Q0DY6hr7q2Z98vdxbhMFgMBgMhucDiafgWfjggw/SAw88IL7ffffdwmitUaOGKN0EQxbJvVBeCZ61jzzyiHhHdmu4cH/wwQci4dSTTz5J3bp1E8c++ugjYYzimpEjRwouGORI1lWzZk26//77xTHsXsN4u+222+jnn38WBi24KleuLDxz4S4tM1T37dtXGM8w6MGHcSNRmDyGzNfI3o0x4DPGh2TM6AfJwjBuAMmK8ULSL8wN7bGrDu9hGPvYaUe1I5kBe/bs2eI7smUjCdjAgQNp4cKFJfabeVTpqIKALMaov4y05sg45s2lo/CUxsxC3iXBpRuPytJRJakPqX98QGkrvrR9Lzd5A+0f24N6VetDg5JP0FtNmlNIn/vIJyiUNseepvbz8+YkyLznGfK3i+lxxGMW+F5i2bHeeda9xPcsy471zrPuJb5nWXasd87bgQ7dsO2BpzAIDL/33nttx1D0GnXHkJYcT09ggHgzZKp/nbh04/EeffChI/5h9O/A58W3tNXf0PDK3eiUfygFhZaj0AGPC0MZaFexMs3sOcB25YutOjk0lB3zmAO+l1h2rHeedS/xPcuyY73zrHuJ71mWHeud83AqVRg2n+EqALcCAIbxmDFjRB1ZuCPAD/65556jpUuXkrfWWZYuDSrqLKMeKN7NrrMsecyusyx5zKyhqrrOMsaJtirqLGPeRnmnJ1ygQdX60Mk9B+mm6Hb04R8f0LGaN4jzXSpVs41RyntQdFX665rhdM2y+XRn9frivCN5x8XFCT6z6ywjmSCOq6izDLlibTO7zjKOVa9e3fQ6y+gfDzFV1FkGv5SD2XWWza6hatRvFXWWMWbUzzS7zjKOIfeI2XWWcT30TkWdZeP4zFwjwGHUb7PWiPz024w1Am6jGLvZdZYxZnv9NmONcKTfZq0RMn7V7DrLRv02s86yUb+dqbNcnDVC/o3niryLskZAvzEeR3WW3blGYMywf5zV2eKsEXKcUmfNWiNkjLYjeQe5cY0AIE8VdZal3rmjzrJTbtjwq0e8p0Tnzp1p48aNosC3LHJt38bb3LChKAXFdLsTqrh041Hphl1S+pCTcJ7Wvn83dY/qbDsfd2gOtap+HQ0LD6Gptz5HPr5Fywaumz6o5NKNRyWXbjwquXTjUcmlG49KLt14VHLpxqOSSzcelVy68RSHy5Ed6NTOsnw6BCDz9YYNG2j48OE2Q1lXV1pXgCcVunHpxtO8eXM6c+YMlS/vvnJUpU0f0tf/QJ8HVspz/rxvICX6BlBknZZFNpTtecwE30ssO9Y7z7qX+J5l2bHeeda9xPcsy471znk4FbMMy/ytt96imTNnioLd2P5+44038pSaqlQp7x/o3gbpRqATl248cOPAgx+866oPltREis4txP5O56vFe4NaN1CcXxCFl6voNh4zwfcSy471zrPuJb5nWXasd551L/E9y7JjvXOzsQxD+cMPP6Q777xT+L2jcDeSeiH9eO/evUW6dBW1a0szVM5fFZduPKg5B13Fu076cCktlY4nWeNbNiSl0rTQmtQsMooebmotQyDh7+PU7e41+qCSSzcelVy68ajk0o1HJZduPCq5dONRyaUbj0ou3XhUcunG424up9ywURrqwIEDIvEOdpkREA2MGDFCvBhkk4lOXLrxIFHA77//bkuGoMOcLqanUc35X4nPyQNupMcuZREFEl3MSBN1klcMHEW9Fn4vzt/V8KpicemmDyq5dONRyaUbj0ou3XhUcunGo5JLNx6VXLrxqOTSjUcll2487uZyqSdkNVu+fLnIGA13VhTBRukohjW7nW5cuvGohKo5/XjwP9vnsEXziQIjxOf43MyD7StWFu8wnIOLuXDoqA+6zYllx7JjvfOse4nvWZYd651n3Ut8z3qf7Jz2y0Rd5apVq4qySvfddx+NHj1aJPiqUaNGkbJH6waZ5l0nLt14VELVnI5dOk/lci6n5JdY0t/q8RHqH0CPNmtDv/QbUmwuHfVBtzmx7Fh2rHeedS/xPcuyY73zrHuJ71nvk51TW03z5s2je+65R2QTfuCBB4TRjO1t1KXcvHkzjRo1ihYuXEi9evVy28AYDEbBuJSeRnP3/0ON0hPoywoB1DjFmkWzZpkI6lSpqq3dO537sCgZDAaDwWAwGAwzjOUpU6bQjBkzhMHsCOvWraOxY8d6tbFsLK+lC5duPHjI8/rrr4t3Heb09f5dtDc9g35LOUyNHvuJaMYUcfxobrIvd0M3fVDJpRuPSi7deFRy6cajkks3HpVcuvGo5NKNRyWXbjwquXTjcTeXU8ZyampqvoYy0LlzZ0pLSyMdgHhsuJoPHTqU7rrrLpFeHxnVsJMO/3e5rS9/BMgGCAoKEufQFhnDEd8tE0kFBweTj4+PrW1ERASlpKRQVlaWaBseHk6XLl2ytfX19RXnAZyDbDMzM8VxXIu24MML40pOThZtwZmRkSFe4CtXrpwoyi3Hh5JJslyAsS2AxG3o12KxUGBgoHhJHpQKw1jT09NtbTE31NZGnxhzYmKiOIe2kIHUB4wBckFbexniM46jL/Cg7p+z8raXob28jTKUbTEn/KZRUVE2uaAtrpcytJc3xmSUobPyhhzRr1He6NMoQ3t5y7b28jbK0Cjv4xcvUjVfC3VEQq+LFylu+H1Ufs4M6lO5uvjuSIbgk7+NUd72MnQkb7TF98LknZ/OFiZvKUP0Cx5n5O2sDI3yNsoQ51F3Oz+dtddvqbOOZFiQzoKnYsWK+crbXWsEXpCbbGvmGoF3ea2ZawTaSbmYuUZARrjWFZ0t6hoBOaEfo86asUbgPPox6iza4lx+MizKGgGe6Ohop+VdnDUC512Rd1HXCMCo32atEeB3pN9mrBHoG5xGeZuxRoDLXr/NWCMc6bdZawTmgvGhrZlrhFG/7XXWnWuEUb/tddbda4T8W9IVeRdljcA4cM5eZ929RoAL/Tirs8VZIzB3o86atUbI+9qRvIPcuEZAJ1yRd3HWCKl3zsp77ty5NH/+fDp9+jTZw8eCEReCFi1a0I4dOwps07RpU9q9ezd5Oq6//voixWBDUfDDqoAqLh15cCPceOONpvOpmNMT65bR7/+spXX+5yny7nfFsZSsTArw9aUAXz+38+mmDyq5dONRyaUbj0ou3XhUcunGo5JLNx6VXLrxqOTSjae4XDNnzqR3331XGJLLli0zjacgHDlyhF555RXxtzeMUdiTpV12juxApxJ81atXj15++WXxJMAeeArx9NNPi7rLDEZpxuHDh8XOMt5LC+IPbaej635y/bqMdNp1+giFpicRNb0ck4yEXo9lCnoAAGpYSURBVGYYygwGg8FgMBgMz8Dtt99OtWrVUlrb2B7gf/zxx0WOK0+GU27YEydOFK7WyIiNDNgxMTFC+HBZXrt2rdjOXr16NXkz4LKgG5duPCrh7JwGLviG1oRUpCU1DlOfqrWES4sRmfvWU9IXj1DZccvJN/Ryn6OXL6SlF85TJx+i8DbXkgroqA+6zYllx7JjvfOse4nvWZYd651n3UuedM9Kl3uzeZwdgyfJzuWdZbhYr1mzRuww//jjj/TOO+/Q+++/T7/88gvVrVtX1F5u27YteTOk379OXLrxqISzc9oUXEG89/vtR/L99A16cu6HNPb9x+nvtXOFJ0fmf2vE+cT3R5Ml/XKflzKssRe7AspSapo1vsNs6KgPus2JZceyY73zrHuJ71mWnSfonSUjjbJO/lekV/LB7UW+Nr8XxuPqnLCx99BDD1GPHj2oYcOGogSujJfdt28fDRkyRJyrXr063XzzzbZz+Fts+vTp1KZNG+rYsSM1adKE3njjDZu37eLFi8XxVq1a0ZIlS+i2226j+vXri03Gc+fO0aRJk2jw4MFUuXJleuKJJwocI2Jy77zzTlGWd/bs2TR+/HgRYztnzhxxHhWIrr32WurevTs1aNBAbGIasWrVKrr66qupa9eu1KlTJ9q0aZPt3KFDh6h///4idnnlypXi2IIFC8TYEQMMl2mJguSRmZlJL730EnXr1o2aNWtGAwcOpLNnz9quRcz0s88+K+xCtDHmvPLU9c6pnWUAZaNgFJ8/f14IFMHSderUEcmSGCSC0nXj0o1HJZydU63sNNrnH2b7Pv1CIlFgZZq0+wCdPPMahWz+WRzPiTtJGdt/p6AOQyj7wgmqemo3kW9ZejXjuHa/E99LLDvWO8+6l/ieZdmx3unNk33uCCW+d2uR+axpmdyH8EdmkX/VRi7NacSIEcK4g0EJg65ly5b0v//9T1TzgfH57bffCkMTBmaHDh3oqquuoueff16ch+GKyj9I1AmPWiQCRiIshKHiM45NnjxZGJRff/216B/hqTDK//rrL2FcYtMRxuMtt9yS7wYjDFkY5oi1/emnn0R7GL3ghbHbr18/MX6MDXHIffv2FXHAMORXrFghjHIYwpgbxoJzErDZ3nzzTWHsSwwaNEgkuYIRLQEDvyB5PPbYY8KtGh7FeGDQu3dvuvfee0WcL2xDxPyi6szGjRtFYrEPP/xQyMeZ38idcCeX08YygCcHUJYLFy6IDLL4MdlYtgIKoRuXbjzIwoebXUXqevs55SReoPS131Payq8oZMDjFNzN+p9OLZ9s2pdPH23PptNu+z42/EQp81+n5IqdqJdvKt3mkwAyUgHd9EEll248Krl041HJpRuPSi7deFRy6cajkqu08vhF1xIGalGQnJxEYWFlinRtQeNxZU7//vuv2AH+6KOPxHf8HYid5VdffZUqVaokPGdhGALt27enzz77TOwUY6f3rbfeErvDMFjl5uGoUaPEtTCWARicMI6xGyv7h6EM71t45wLYwQX2799foDcuEnIBMOyxywzjFHjuueeEEYy/YwGMFwbwp59+KoxlnMcOMNrITNQwto07xo7+/rU/hlLB+ckjPj6ePv/8c7GDjrBBvLBzDFkii/Tff/9NS5cuFfai/D2Mhrin3rNOGctI+Q3Xha+++ko8NTACQvz444/zPL3wRkApdePSjadx48biRpap+FXOKX3LL8JQBlKXfmozlsP9/YnyyUd/2j80z/e0pZ+I90ei2tKCMtVoQPJJykk4SxGa/U58L7HsWO88617ie5Zlx3qnN49PYPAVO7nOIjwnp0T+7jJCVuuBMSljaLHzKneK4WpsxN133y3et2zZInaJYZQaeWAww6iEgQj3ameMNRjUgCyNtH37dpH8SuKOO+4QL2NyLOOcMAfsGvfs2TNPn3A3xgsu2nABLy7QT0HyyMjIEHKEqzmAkkx4UABZYHwVKlSwPVjQ5f8Kp4zlhx9+WKQgR8kdbMVjVxk+6ydPnhSu2fBph7LJpyfeCDxtUZUOXRWXbjwquex5ciIq0o7ActQi4xJRurVWHZDp45uvsZwfvo2oLd4DLDlkSYnX7nfyBn3wdB6VXLrxqOTSjUcll248Krl041HJpRuPSi5neJB3CS7CRsCNGq7HjiB3XY3xr+CRhq99UlZXgB1gGTvs7Jz69OlDf/755xXtTp06Jd6dqAZcKGDMy/jkguTYpUuXK45j87SgMXiq3jn1qAd+87///rsIMIfLAcrv3H///SJLNnzwP/jgA+HHzmCUZmBXGZnc8a4aX8ReoF7V+tIrkc3oeO6OcfqWXyktseB0+l+G16HFoZUpdOhL5BNuzQ8QlGMtA7AvINx2jMFgMBgMBoNxJaTr8vfff39FLd5GjRqJDT97gxl/K8KNOjg4mPbs2ZPn3NGjR6lKlSrChVsVMAfYXCdOnMhzHBuX0dHRYpwHDx4ssA/ERAPYLc8PBcmjYcOGFBQUdIUcsbuMOGbsSEOmeOkEp4xlBIMjgDs/IFjd02toFRdQUt24dOPB0y64j7jjyZurc8rKtiYamB7ZmIbFdBOfU36aSBnYWc7FuUNz6I6EvAvdU9Ft6JaYrnRHgi+FPL+QDt4/k2plWZ/4HQ0oQxGPztLud+J7iWXHeudZ9xLfsyw71jvPupe87Z5F5mh4xyLR1rRp04QL9Pz58+mBBx4Qrs8IMcUOM+Kat27dSs8884yILUafTz31lIjThYEsd5m/++47ESMsd5YTEhJEfLMxVBX5nWJjY23f5WckSi4IclcX1xvnNGbMGAoMDBTZsBcuXChcoqdOnUo//PCDcC0fOnSoSMolY5Tx9y5co411luEiDcMaXsGyDa4xuoffd999+cojPDxcyAxJu5Acbdu2bWIzFe7f+Nt6+PDhwuUe8dwScjwYh6fqnVPGcuvWrSk5+bLrqKMf1tGPj2Bwb4GKeAzVXLrxqIT9nMr6XDbQDwWUIUtGKq2NrEMrQ2NoaNIxOnp4HiG65b3mrWhwpDW5gxGzD/5LoV+8Re3+XEh7A8tScE42/Ri7jnzDo7T7nfheYtmx3nnWvcT3LMuO9c6z7iVvvGdhFMIQRNknhI8iDxMMZ+Rc+u2334Sxh2zSMDqxi4vs2cDLL78srrvmmmtESSach6H46KOPivMwwJE4DJuGiGU+c+aMsH927doljE1ksD5w4IAIY5X9IeO1I2AHW9pOL774It1www22cygnBbdtxAOPHDlSjBXGrixHhZK+yGLdrl07kWgMHr+Imz5+/LjI5i0Bb2AY+0j+hQcBmBfieyELuHMjv09B8pg2bRqNGzeOvvnmG8EHgx0lrmCIY1cafc+dO1fIFcm/kBVb7kYbk415kt75WJzYZsNE8SMgLbo9ELuMpx2o0SWfJKDLnTt3iqc18kmFpwApz5H+3FXA5UCVH74qLt148AQMGQPxlAwPgFTO6eNfP6YHzsTbvmfdOILKzZtNST6+NCTpGH0Wu1EcLzt2Ke04cZBar1paYP93xR+gaRf+psjXt2j3O/G9xLJjvfOse4nvWZYd651n3Ut8z7LsWO+ctwOdSvCFJzAwivG0wBGkvY3i1qURWBTgMoCYa6Q3x5MUBkMlMnLdsCWyT++jKqJslC9tCypPflUaUvapveQTFEbVK1/OumiPOuHl6FDiJapYJoKC2z6oYOQMBoPBYDAYDIZ3wiljGeWhsF0PX3lnAOMZQeb5GdeqAR97uEwg7bpZ8aoRERGm9FuSXLrxwLVkx44dtnT3KueUmW2NY2kSWYH2XLxA8Yd3Uj3KpH0UQNGValLEsFfIkp1FPn7+FBURSQdvup/qzraWipKw3PeceI9NTabIoGAK8PXT8nfie4llx3rnWfcS37MsO9Y7z7qX+J5l2bHeudlYfvDBB+mmm24iV3Ho0CGn2y1atEgEicMf3z6oHQbua6+9Rr/88osIpq9WrZpwC0dmY2cgU7w7a+wXBQj4l1nmzIYqLt14UAKgdu3aDouymzknS2YapcfHUoQlhD7tdi11+fVbOpScRNlkofq+ObRowM2iHQxliVrhZfPtu2JIWL5cZkI3HpVcuvGo5NKNRyWXbjwquXTjUcmlG49KLt14VHLpxqOSSzced3M5Ff1cFEMZQOpxZ7B3714RCI5sbsZaZhKTJk0S2d5WrFhBGzZsEKnaEZCelZXXtbUkoXIsqrh040EmQzz4kRkNVc0pbeVMSjt3RNRFrlfWGot0IDWVsi1ETfwsFB1iLSVlhK9d7b49w+92istM6Majkks3HpVcuvGo5NKNRyWXbjwquXTjUcmlG49KLt14VHLpxuNurlKRhhjpyZFRDQW67YF07FOmTKHHH3+cQkOtRsULL7wgEoghsxpSp+M6Ry+ZqU4FvDGzoKfxwGNh1qxZ4l3lnDalZdEr5ZtTQE42RQeHUjkfC+3LyKIsi4X8naxn3zgyymt+J76XWHasd551L/E9y7JjvfOse4nvWZYd652b3bBLEn/99ZcoWyVTrgNww65ataqoM4aEXaiXVtLwxPgPC+rBZaWTT2CI18TOqALmlJaVRYF+fvRKfKY4dj4gRIQRVPf3pVNpOZTtS+RvqLNcHC4V0I1HJZduPCq5dONRyaUbj0ou3XhUcunGo5JLNx6VXLrxqOTSjcfdXKXeWEYRbKBKlSp5juM76pa5AmeSe6EeGXa6Zawzaozhhc/Y2cZONyDjXlNTU8U7CnjDjRxFt1HXDDXL4uOtpYIQfw0jSbbFDwh3c7gIoC186i9dumRriyd+0h0d59LS0kQ2chzHtceOHRPvqFuGcSUf2UW0exmVGfi4qLmGF/jKlSsnMoEDaIvxyWLnGF/KHx9RzpqviZ77U5QqwBggI8R244V6a+BBWTCMVZYBQ1vMTc4ZY0YxdgBtIQOMGcAYIDO0tZchPuM4CrWDp2zZsmJ8uL4wedvL0F7eRhnKtrIvzFHKBW1xvawjbi9vjMkoQyHv3LbgdCRv9H/b2j9o0elj9ED95tTQ10LL4BJiyS3X4OtHF7ItlOlP5JOTI/qDvOVvI+W9pt8wSs7KpJZVqucrb8gIfWJsUt44l58MMTb52xjlbS9DR/JG7UD0V5i8HemsUb/zk7eUIe5BHHdG3vY6ay9Do87KtkYZQjaVK1fOV2cdyRvHHcmwIJ3FserVq+crb3etEegfDxNlW2d1Nr81Qra1lyFkDX6ZD8LMNcKo32auEVK/XdHZoq4RGHNUVFQenUVb9OlIZwvT7/x0FseQ28Oos2asEbgeeuesvIuzRhj/HzdzjQAnxlGYzhZ3jQCnI/02Y42AVxXGbpS3GWsExmyv32asEY7026w1Av1h/mhr5hph1G97nXXnGmHUb3uddfcacfbsWSEvV+RdlDUC+o3x2Ousu9cIjBk1hZ3V2eKsERgnzkudNWuNQH845kjeQW5cIwCp587IuzhrxIkTJ8Q8CpM3ykTB6xSAnE6fPm0bq0t1llUBhbonTJggfmiZ4Au1m1GYGz+oMTET3LZxAyLeuTDgB0bxbMQ+o5g3CnhjR1qXOsvxU2+knAsnqNxrm4USAlk5OfTB7m30cNPW5JePW0/C26Mo+8x+KvfKGvIJCC6Uxyyo4nn/t1/pkQE30NqNG6lz+/b5tsvIzqbxy3+ml3reQGEBrieFQ7bqSt+8Lz43LxNONwdl0wsXUmwZrYf9+CGdP3WAUvyDqHHZSJo5+tlizEq/34nrP7LsWO88617ie5Zlx3rnWfcS37MsO9Y7ctoOLBUxywUBTycB+aRBAk8pYPg6AzyBGDNmjHjKsG3btnwN5eLAmL3bbNhzpaen0VH/UCJDLd+v9+2ix9cvI//PptLyk1cmtEr9/QNKPXOAjvmHkiUlwSkes6CK54cDW4mu7UmBoQUbwIu3LaPXDx+i95bOLhLPv7EnbZ+bxR2ibLi7GxAZFEKX/ALFcX83xCfp9juV5L3EPCw71ofSex+p5NKNRyWXbjwquXTjUcmlG49KLt143M1VqBv2nDlzRCZqJNI6d+6c2KKOjo6mTp060ahRo2jAgAFkJmrVqiXe4RKMbXYJfO/Zs6fb+aQb9tChQ+muu+5y2p0H33FOhRs23BBwDO4OPmf30ZPBtejbirUpPTWZMn0DxIOEM/HW/oBFB/+jjpHReVwjfLb8Ss9HtaKvIurShTPHqXzZile4okges92wJY/ZbtjxgT5Eg6+lMuXCCnTD9k+zcpyOixXvrrpYrtz/D4XkZFGDzETKCAqltEyr+8n33fuL8xH+ARTnG0iRORkFumEX5mIpZQh+M10spQzxWbpumemGbdRvs92wwYvxmO2GDR7Mx2w3bOlKpcIN23hvmLlGYF5mu1ga9VuFGzauxXzNdsOWYzTbDRs80jXfbDds6ZrvrLyLukZgXEb9NmuNyE+/zVgj8B2yMdsNG9/t9duMNcKRfpu1Rsjxme2GbdRvM92wjfptthu2/D/dbDdsXIv5mu2GjWvl/7Vmu2HjO9qZ7YYt/z4x2w07LCzMJXkXZ42QeuesvOfOnUvz5893zQ0bvv833ngjrVu3Lt9YXwipd+/ewqA2GrLudMOG0BCT8uabbwrjFTh58qSIj8KkbrjhBnInPM0N++Lzball9f50LCCMHg/zoWmjnhZu16+u+4Ne3GVNfPZ6yw70XPu8Dxb+/OwJ6pdjrVMd36s3RdRv5/KcoBfS7dtd8zETMW+/SGePnyOqWY0O3PEI1Y1wzLly+0rqtWkj3RFC9OXo51zmGfPrl7T26D6qlpVCPmUiqW14OM24lEJnxowV56et+4Oe2bWdYrJSaXC5cPrw5qeKNS/d3LbYPYxlx3rnWfcS37MsO9Y7z7qX+J5l2bHeFdMNG0YQDGXUSe7WrZuI90UyLTxVwFMH7DCvXr1alHNCtuoRI0aQO3Do0CHxfvDgQdsxPKWAC/X06dNtTzUmT55MjRo1ov79+7uF19ORnptR+e1kC508d1x8Tos9YjsfcmTrFdcMzKlk+zxh/17qv3iOS5wZ23+nSy+0I0u69TfxBFw4F0/01qdEsecpNvVyPe/Mw3+TJSc7T8wyEJdZtBptJ5MSqIIlkyLKx1BSVraIHze6cFQoY32wdMY/hHzTrE/UGAwGg8FgMBgMRumCQzfsn3/+WRjKb7/9tsNaxcgC17VrV/G69dZb6eqrrxZWOKzxomDixIk0c+ZMm7Hcpk0batasGc2ePZvq1atHr7zyijjepUsXsRWPWOUlS5aIz6XFDRsuAarcsMFhzBCb5mPNaAjEX4qji4ERlGKM8c7JEf0ZXSMqkIVOk3VX+K0T1hhbe9cIyePINSJ522LxOelCLIVUrF4sN2zJY7YbdpahTFNmSorgTI8/SY1/X0jzqq6nFt1GCXknpFgfAMTlFjQv1J3nlR5EnW+hyOufoBPnYmlhUhpRcDQ9GBJK25JSKCMzg3zJYnMRCvG9rLe+yReL7YaNMckxmumGDR0Gj9lu2Eb9NtsNG1zSpcpMN2ycx3Vmu2FjfnipcMPGOFS4YeN6yWPmGiH1W4UbNsZvf9+b4YYNHrQx2w0b5/FS4YaN7yrcsDE/FW7Y+em3GWsE+kV/ZrthO9JvM9YIR/pt1hoh3TrNdsM26reZbthG/TbbDVv+n262GzZ40MZsN2zwoI073LBRDQaes7B/9u3bJ+ycd955x2bfYDwq3LDxW6lwww7PHYMKN2ypd6a5YWOnGAP58ssvyRl89NFHtHz5cuGO7ekoqhs2BA+lUAEj19jXb6VJ5Zvbzm3r3ptq+ftQpWV/UqbccY4Ko8eGPEzpf/9GfuWrkn/NFtTq00m03ZL3YQMyNTs7p8QvH6OsvWvF57Iv/Um+ZYruNqRKdj4vPUo0+T2iFx+hzY+9QG2jK9OOfVup5cql1Ns/m5bc8bxwYf9p3UIavms3VbJk0b47n6aIQOtDCSOyY49Q+vofKeT6Z2jPS92oQmgExbz4G51MTqRq335I/bITaFijVnTf/oM0KvEwbShThQ7dP05cu+7MCery67fi83O169DrfYcXa16q5Kcbj0ou3XhUcunGo5JLNx6VXLrxqOTSjUcll248Krk8jQcG3COPPEKvvfaaME7j4uKENy/KWS5dutStXIVBN57icDntho2M0XCxdhZI9IVrvBl4GlMSXEZDGUhOT6H/rfnDZigb3bRTfhhHiR/dLT5Xslw53kyDK7I9zxXIvnwuO/ZwoWO2ZKSSJSuzSLLLjjtJOZfOUHGwOfZUnu/p546J96zcMS3P8qOpOzdZz+VmFT/r409t5828oq+cpIuU8M5Nwli2pCZQs5qD6MFyzcS5tNxrH7ZcojbV64rP34fXpst7/0QVgi+XQMsJt2Z79wTd041HJZduPCq5dONRyaUbj0ou3XhUcunGo5JLNx6VXJ7Gg9xQTz75pDCUgfLly9Pdd99Ny5YtE6Vx3clVGHTjcTeXQzdsbLW3aNHC6U6w/S1dgj0dRXXDBlS5YUuXBUcyP3vpAsVnpOf5aZ87l0gdjh2iq3K/wy2hDIxdP7tY2/PnqWxAoFOuEVlpl2N+c/wCC3RXi9+9hiyzniCq3IgiHp55hRu25MnXfeoNaxK3kJfX5JFhcnISZSfHk19YuUJdLG/6cz6qjROVixDviUmJdPF8LF1YPxcFykS7nbGnxTyTDL/p/viLdDEujigzjYLCy1ldUT57yFamK+nkAfG+wS/cKn+0JaIAvwBqWKU+ES0R3/0sl92w/dIvF2c/fCnOLdmwVbhhy89mu2Eb9dtsN2y0N9PFUsoQbcx2sYQMcUzKxWw3bMhBVTZsFW7YGI8qN2zMwUwXS6N+m+1iKfXbbBdLuUYYs/iauUbY67eZ2bBVuWGjbzNdLKUMHem3GWuEI/02a42AXqhwwzbqt5lrBK5xxQ37TFwc7U+8JMaPa+VvjjFBRrgeY0db+79RhSvsxfNiTNJd2r4txo5jcgzybxvMAXIpUyaMEhMvt21SPoosuX9LGfXbkc5KeeP8q6++Sv/8848wRlu3bk2ff/654Ni/f78I+YQ8kTcJFX9mzJghflv0DQ9ahKhijGhz22230fPPPy/6/PPPP4UrNdqNGzdOeNlu3bpVGL+zZs0SeZ927txJGzZsoMGDB4sQV6nLaGMM0ZLOvjCkMU+MTYUbNnhVuGH75spPhRu27Nc0N+w6derY4oedRbt27Wjz5s3krW7YJQWfGVOuODY88SjNCa+Z59gb0WXono1Wt/pyr22m6997ihYGX07yBbzcpguNb9O1UE7xn8YHt1P2iT100i+EGt3/IfnXyLvDbQQydktEvr7FqXk5ut7+2tQlH1Pa8s/EfArLyt3t01dojSWQnr64h6ZFNqHlHbtQlwALvTvvfXomqrVo81LcP3RXiy5U53jeutNprZtQyk8TqOy4ZZRz/hglfnin7ZzfqNeo7MZtVMeSQQfvH0tbz50Ru9FrQxOp862T6Orv3qPlSSlUNyuFDjw4QVyD+sqhX7xFGTnZ1KFiZdpw420uy4TBYDAYDAZDBbadP0Ntfr7S066ksHXI7dQ6ylrRxVkgKTDyIb3xxhvCoGzZsiVdc801NHbsWGrevDl9++23IgfTpk2bqEOHDsI9GgbxSy+9JHIooToQcibt2rWLOnfuLAzjp59+WvSNNkg+DINryJAhov+aNWsK4w+JkJs2bUpr1qwRbtawldq2vfx3sREwwrdv3y6Ma0bpsQP98zOG/v3333xLRjkCFMKbUZrS8BsTfkmsPLKXRvj4U4Qli0Z8NfkKQxl4eetaeqlVZxG3WxBP0tJP6Z+zJ+iOatfSgcBw+vrEcbq1erMCDdYO1a6hTmnnSS610K29G+ZTw/aD6FJCIt284U+6vmZ9eqBJK4fXJ/v4UfkZU2hF/xHUs1ptcSzzYO7Dmax0ooD8i48nZ2ZQ9awUIr9AGlExhqZlIuFEFq1PS7UZykCQJZt+272BKKJJnuuzjlkXrZxLZ/MYykD8Bat7d2ju1FNzd9z9kq0G9xNte9DylYsp3uAWD/mm3f0U+X76BjUuV4GKC91KTZSme4l5WHasD6XnPlLJpRuPSi7deFRylVaeRuUqCAO1KEhISKSICKv3nbuA8bgyJ9g0ixcvFjvE0stg9OjRYqe5UqVKIpkwDGWgffv29Nlnn1GrVq3ETuZbb71FkyZNEoYyULVqVRF+imulsYwdShjHMJRl/w0bNqS6desKQxno2LGjeMcutiNjGZ6tMLbnzZvn1JzcCd143M3l0Fg+evSoePrC8CwEWnIow8eXTvtfjomV+C2sKo2s3I0Wn1pBP2Xmn0U8JSuTwh0ktBLnFrxJAY260lP/7qZPq19jO37brn/IN7om3VLfuiDYY3lIJdofGCFeMJYzzxykaV+8SP+LakVfX/iCBnYZTr8fPyxe0ljOiY8ln/AK5ONrNfwPB1iD9L/8az71HPWE+OwTYB2nJT2VfAKCRQKzrMN/U9iQF/PwD/ltNi3xK0e1juyngZ//SjTmFspuk53HlTw4J5uyyJeSM6xuGsCDl/bS7PBa5BNcVnzP3LOS1gdHUfXMZKqWbXVzib9ojaUO8bVay6mpViM5uLr1/okuZ3XxjvcNzDMmPFg4evMDFG2IX2YwGAwGg8EobQj1D3B5J1fiol+QMgMpP+zevVu833zzzbZM03DFhQGM6j/Vq1fP0x6xw8CWLVvELjE8bo3ATjTctOGyi4RcjgDXZyNgUAPSzdgIuATfcccdYie7X79+xZorw/1waCwD3bt3F09JCnNvhR/4iRMnaNWqVeTNUBmznR9X57IRtDIhiU74hzo8vy2ofJ7vlUPL0OmUvHV+408foPCaVqM33pJNved+SX/0H0EVQ8Iofe334vVH9SvrW68/e4pG1WtCvnb6gh3kYZW75zk2c9OfwlAGHjl9kTrnxv/KKy3ZWRT/Wn8K7nUXhVzzoDh2MndOXydmUPMNf9LTHfuSj3+usZyRStsmj6ZKSeco1JJNGZvnU7nJGyjr4GYKqN+RNpw+TuTrT/4WX4qNPYdViTKzs8gn83JscnhOJk2ocJV4j8xOp59O/0U7g8rRRd9Aygm2PhFNWzqDBtQZLtocPTJfHLt4aBtR2dYUmrsb/9+lC+K9TOV64j0q1xg2JlyTqFHGmtTBU3RPNx6VXLrxqOTSjUcll248Krl041HJpRuPSi7deFRyOcPz448/ip1hI5Cn6Ny5cw7byzh3GdsreaThW5iN5CwefPBBatCggXAJL62y82a9c2gs9+jRg1asWOFSR1A2b07whe8qE3zhyVRAvLU+ssSAqGhhLJ8x7Cz7W3JoaNIx+iG8ljDYGtUYZDvXO6YafXvovzx9/DxrPN1111QKiqpGP+7fQ9svxNKvh/bRzbUb2tpkO1gcPtizjSr4+dOLbbrkCbovWyav4Y5zpw0JruItRPV+/sIql8Agir8YRznHdtJ7ZRvQsKO7qfKKWXTUP5Reibzs6fDx3+voiXa9xU4wkBR3jtpEd6N+oado9tm1sNAp4eN7RT9081sUk51KCb7hlGRwT0dysJwEq8x/6DWQnvzzR/E50TeAmqdfpFYZF+lk8+vIcjGTTqdnixRgJ/xCbG0SfPzFbv3U0AbiWI6FaMPZk/TodmtctZ+vv/jN/FIvP0GUCb7clbzHUQIkMxN84TqRoExBgi/wqEjwBV5wmJ3gCzzox+wEX+gbvCoSfAEqEnyhX1UJvjB+FQm+ZE1ssxN8iTXIz8/0BF/gwdxVJPjCeFUk+DImWjIzwVd++m3GGoHvkKPZCb7w3V6/zVgjHOm3WWuEHJ/ZCb6M+m1mgi+jfpudBFD+n252EkBcCzk6SvCFzT8ACbcefvhhm7wxl9q1a4tSuUeOHBGGtJQh3KXhng0+VPzBrrTUb5yrUqWKkDm4ZHIu499imCfa2tdkxpygg1KG06ZNE/KBuzfaQG5ffPEFDR8+XIxdRYIvyElFgq+wsDClCb7Ql2kJvn766ScaNmwYuQKQwMD01gRfJeGHj8RX5etYa/TeHl2BPu3YgwIX/JynbZmcTLot4RB9WO6ysSuxe/jd1HTO51ccP9C3P9Wt3ZwmbFhJL+/cSHP63EDD6jSyJdpqWmOgQ1fvATXq0oJrhuZ50mZJSyLfrz+wfT9+8wNU67uPKG+RKitqhIXT3qqhFP/7BxRTZxj18EmneQd/pZExXenP0MtuLrWzU+nQAy9T8g9jKePvxRT+4FcU8PsicW7WmbV0Xcop2y51FvnQsMrdaHVIJQo5eoxSX/1Q1Fl+qE55qpyTQS/llKWLtz9GTT9/jU7l7l53Tj1HC0+vpL19HqVOh07To+XL0NpTR2lr8OUYmXLZGXTJL69rtRFnu/akik06iAUBscmO6li7C7rFmnhqTIs38ajk0o1HJZduPCq5dONRyaUbj0ou3XhUchXGg0zUiFtG/HGfPn2EcYzEXS+88IKIU4ZrNTJiwzUbx5G4eMSIESJ5F7JmI5s1jG4kQO7Vqxc99dRT9Oijj4q+UQJq5syZYocaxiaA/uDe/dtvv4nvsbGxIj566tSptljn6dOnizG98847NuP9l19+EUYc2pQW2XmT3jmd4MtVQxnQwVD2NEwy7LY2Dgsj/+i8GbCBN89vo8P+YVccHxyQRU0iHdf4Tc3JIUtODqXEWZNXDV/6C21uvIusVYOJsm2maF4sOnaQnt+0iqZ06Cm+46ZPeG80UVlrUgOg+nfW5AqOUDk1jnJSsyjJ1xpPku5rVc9MOz4bf25SL0v6ZVfyW2O6iPe4Q3PEe8U6l3XZ6Aj9QRaMfavBH+znnycpWkSOtTZbi/JRVHHfYXoX1aAMhjJQkKEshpYbT+0uFx0Gg8FgMBgMRtGAbNeICUY2bBjFMHjff/99qlatmjBon332WWFQx8TEiPMwlIGXX35Z7Mwic3ZUVJTYrfzf//5H999/vzgPYxqJw7BzCQMZdZLhoYqs2XghBvnDDz+kvn375unvqquuEgY3/lZu0iRvYtnly5fzz+wJMcv5Ye/evfTuu++KJzJQMGST69q18HJDugOuBqq53opsbDsWVKkO+fjnNeBqZybR8KRj9HbZK3eVA3OffDmExULHV35Db5ywJq8CZmz4nWSRqpwC7L9v9u+m11u2o+S5kyj0hufo5zQLkTU/VqGoHn+KfBvVoWWh1iQSWEC+K1OTVobGUOsyYZQSf47+8wu17Up/me5Hj9QZThc+f5god4ddAjvuZw/9lOfY92XiKPvnH2nw8X/zysLPL4+xHO7nS/61W5N/WFm6Ifk4fVoW9ZKdx4yzGyi07E2km+7pxqOSSzcelVy68ajk0o1HJZduPCq5dONRyaUbj0quwnjgbgv7BS97IBM2aiM7AnZ5J0yYIF4AXJVlkjAAO9V4GfH3339f0c/hw4evOAb3YE+QnbfrnUNjGYowceJEWrRokfADx1OT3r17i9pjUCj4rcMvHdvbX331Fb344oui3pg3xyzjPD6riFmGm4fkl8jMyBRtjEC8sniPrHLFPAN8fMXvCAWwL/p1MTGB3juOeGi/PJm2e1TtQ8m+/nTeL/8yTT6ZaXRpfA9rP5E16a5KnZyUPFFgQCBtv3iJ7q/YwdpXThY9XLG9+Nw+PIL+vXSaEGGNWUHOM3PzLVSwM5QlKhl2ldumXaCudRpR3FVXEZ27XEM8kCyUlZlJKbm72EBE+aqUNeBuSk5PpspZl5OAOYuOaecpMSWNfDMzbbEYZsYs4zrp9mNmzDJcWsBvdsyy1G8VMcuYH54Umx2zjBeeVpsds4z5ws1LRcwyxib/ozczZhljlx4aZsYsS/1WEbOMMeI6s2OWwVOhQgXTY5bxgt6piFk26qiZawTmbNRvs9YIyMSRfpuxRoAXczE7Zhm89vptxhrhSL/NWiNwHhxmxywb5WtmzDLeZVuzY5bj4uLEZ7NjljFGjNVRzLI71wiMEeNxVmeLs0ZgXPhudswyzmOeZscsBwYGin5UxCzLvyVNi1mGKwICzeFOIOrh7t1LCxYsEG4HO3bsEMHoY8aMEYOHMGEso+aYDjvMnhSz7DND7vUSfdz1Grq/SUta9PcqGrh5gzjWMCOe1seuoc+b3kDPxOc1ie8O86fPbnmKTiUn0tMfP0vfh9eynVvZsTN9//dq+iT9sgGJON51IdF5+vi8Sx+6e+3SPMeqZSbTzuPW+AyfXndT5FHrDV8YOqaeo9rhZWlkVHm68UzeDN3A/+rUpd0HttEvvmUpOjudYh8YR12/mkJrL+cKK7j/9Diac/0DNGXaVHq/WlmiSOt2d4SPheLvfT6PLJ8qX4amDXtIfP70+yl0n/XedBr/Hv2VKj38PUVWriG+N5/zuVhQdg67i8yAbrEmnhDT4u08Krl041HJpRuPSi7deFRy6cajkks3HpVcuvGo5NKNx90xyw59cefMmUNLliwRbgTbt28XLxjEcFGAq8Ejjzxic0GANQ+ffyQF82aojE11xOWfW+c3ODdLrThmsVDI1XdTgP+VdZUD/aw/fZWwcPrg3OY859IzM8gnd1dawt5Q7hIWQnc2aX1Fv0aFykizPoVyBv4hZWhfji9tTnVs/ZYNDKLbLFar9ZxfEK06dYwC8omddoQA/0CKvRBH77//IVHiZWM80IEskZVbonJQ/jWp8wN24WU8NfDP8LtNM5RV6p5uPCq5dONRyaUbj0ou3XhUcunGo5JLNx6VXLrxqOTSjUcll2487uZyaCzDnQpu1xINGzYU6daxTY9aYI4GtGfPHvJm4KFBSXL557rhhvhedp329/On4B63C/dmewT45R+uvi8pUcQtF4RuFaIcKqIvXb4uI/1yXbrC4OfjQ1tz/GlyvGNjOTwomPoGZNMdCQfF954Lv6eg3AcEzvXv+Djik4EZdLnGXkTQZRf3qrl1kl1BgCWHIqMqkm66pxuPSi7deFRy6cajkks3HpVcuvGo5NKNRyWXbjwquXTjUcmlG4+7uRway0iNDl9uI+rXr0833HCD8BO3B3ac//rrL/JmSH99s5F1dCdd3PIHnUnJ66rs75NrLIdfztockGu4tiwTfkU/gXbJwIx4ZO9e2phRsLEcEeY4a9dx/1DBiprG1S85VxB88cnl5Jc7/vwQiuzSfgFUNjdTNebryp7v5UcIeRGR+9DgnnunUu2s3PhFg4FcPcv53XHjzrIqfQBUcenGo5JLNx6VXLrxqOTSjUcll248Krl041HJpRuPSi7deFRy6cbjbi6HFgrqhSG9un2WNrhnOwLaqsxw5s1I/Oguop9epMqzLtcuNu4sh5a5/CTFP3fnt32LHnRv/P487e1ds389tZKGJR61fT+cU7DxGl7W6pZt32+2j69IuLU70PknOhEDnxE7yxLRliwaWDVvGayQwGDyCUImbGu7MEsWFVy8KS/y20eP8LeewS55Sq7BHhFyWZcjOwwpsN/Do8ZccUw+pGAwGAwGg8FgMBieC4c2RIcOHUSGMAQ433jjjYV20qBBA5FB2tuzYeOc2dmw80NaSopoE5JjyfPjiicrvmEU2bg70anLGd78LBaRkU5mqOuado66pJ2jn8KtRmpWIb7+/jnWvqd1uoY+3XM5u7TEdVUvu/EXhpysHGSas30PIgv1qlyTFp68bLwH+PhRTlAZqphtPRadkUz+2VlEfs4F7/uRRfxGN998M30XFmo7Hu4fIGQAWaTnPjvy8/ETcxNZLCvVoVaR0fT3xctu2kZEG7JoS/hc85jQA/weZma6BWSWP/kEzcxs2Bg/eMzOhi05VWXDNjPTrZShzH6sIhs25qgiGzbGIK81Mxs2+CSPmdmwpX6ryoZtZqZbo36bnelW6rfZmW7lGgE5mJnpVq4R9vpt1hqRn36bsUbgWjMz3UoZOtJvs7JhF3VNdnWNwHmMz+xs2Eb9NnONwHglj4o1wlV5F2WNAI8jnXX3GgEeV3S2uNmw5dzMXCNwHv2qyIZ9yQV5F2eNkHIwLRu2N6Oo2bDt666ZhYvPt7XVETZifr8hdEMtaz3gJj/MoH/jL1KPnCRaOeYVcWzCktn08pHLxueEmAo07vp78vQZfM2DFLr/slHYgDJpXz7Ozj/3HUyDazcgS1YG+X4xvVhz+rfH1fT0uj9pUabVWK1HmbT33hcp8pPJlOBr5V/WbzB1Pr6Jzv/2HjWuOYh6pMZSuewM+jaitlMc11uS6Jf7XxG/U+CXb9mO3xwTQ99ef7v4HP7xJEryDaDlPa+hXg1a2tpk5eRQ5GdvUJKDhGKW+57Lk0lbtB88knLKVVWiDyp1TzcelVy68ajk0o1HJZduPCq5dONRyaUbj0ou3XhUcunGo5JLN57icDmdDZvhOoz1dM3EMxVaUaMag6443rz85WzVOwfdJMpGvWCJsx0L9MsbtetvSAQmEdz9tjzf8zOUgXCZNMwu1njRqRXkKrIyssjPsGse4kPk6+NDIZbLcfMhQcEU1PkmKhMRRT1Tz9KisKpOG8qCg6xP2zZv3kzTTllLawHlDMm8MnNvh2CDG7Z0cS/jSlI9H19l+gCo4tKNRyWXbjwquXTjUcmlG49KLt14VHLpxqOSSzcelVy68ajk0o3H3VxsLHsYPi9bj2L9L5clAt6oU4vqRFyOEfbzD6D1J5ZQZ19rMiwg0C779ejul93rwx/+hiKenkfkwIDOD2Xk0xq7a8pnW90nCsI7Ydl52sH9wxizHOST13gFwhCz7OdPgS36UUhO3uRz9vi+Y/crjqX6+NK///5LXbp0oTZH9tKA5JPieLmQywnrMnMN/+Cgy27aEsEuZN62f4DAYDAYDAaDwWAwPA/8V72bUJIJzgLsjdzc7/41W9gOBRraPBuQStUjLsf6+ldrTH5R1V2qSSZ3lu2vCcvJctj+h259bZ/vDrFQkGHXuGK5SMoxGJjBuX1mGvoOkYm4QsIp2HCtPaqFhdPwajUpOssamyCRYnShDggmv9w60pGhl3+3nFy+YAeltuSYCsK1yadodMIhYSyr1AdVXLrxqOTSjUcll248Krl041HJpRuPSi7deFRy6cajkks3HpVcuvG4m4uNZTf6xpcU7F2qffwDKeKZ+RTcb4zDnWU/uyznRUF4oOOyUEbXaYnX2/egoXUaXT5gsVBwrrEad2gOBfv6U7rBmJWGKVynbf3mjj+4680UlHutI5xLSyHf8Aq0P2VHnuOplst9hQ590dZzWQclsIIc1KAuaGcZ8dsfN2tO351dS++c30o+vr5K9UEVl248Krl041HJpRuPSi7deFRy6cajkks3HpVcuvGo5NKNRyWXbjzu5mJj2U2Q2dtKAv4OjDu/CtXIx2BEBxhilv3IOWP5q7PrCo9ZdsJYHlK7AfkGBNFd8QfogUv7yMcvgIYmHbedz8jMogyjsZwbvyzdokW/uaWufAKCKa1sTL7jSs/OJp+AIIp48qc8x615+Kzwr1ibUn2s8ogMyuvSLvjt4ruBgALqUiPR2d216l4+4OOrVB9UcenGo5JLNx6VXLrxqOTSjUcll248Krl041HJpRuPSq6S4EEiqB49eoiqQMXFkiVLqFevXlStWjVq1aoVHTly5Io52bfZuHGjyO48b968PO2WL1/udtm1bt2ahg0bRsXBmTNn6PHHH6fKlSvTV199RZ6iD2wsawB/B8adPQIMBnWOkwnQ+1WsUnjMsh0cuUjDTRyG+7QLf9PkuB1EgcH0QsJeOnV4rrWBxZJ3ZznXWM4yGMuhuW7YwCA/xzdA2cAg+l/LTuIzdneN6JCTbCvLgPf0qk3E8XKBwU7tLE/sfmVStfg7Hs+zm3/5C99WDAaDwWAwGDoDmZPbtm1rK41UnGRUI0eOpClTptD27dtFv7L0VkFtUA6pXr16VL58eVu7w4cP02uvvUbuRoMGDahmTWt52aIC5aBeeeUVYTR7fJ1lb0ZR6yzjyU5J1VnOyrDWBy2o9llO6OUY5YysLNGfo/qIfVNO05+hlcWxoD4PEi37zSFnSmKSrfbZ/FMr6cYqPa3X9L6X6Ih1LhLpycl0MSuHfHveTTkrP6f0bNiTfhScnUn+1z9PWeFRlGpwDQ+wWGy11GzIzKK0rGwhw75ZibT72DpqWjOvAXvo+ttscoasquSkizFOO7+NeoeHUXCtWhQbGyvkn3Dgb+t4LWSrs2yTZ1oaXUxNy1MfsUN4JH3d7Vq67a/fRZsu0ZUpxMf3cq3qZOtvC8QnJFBgZCVldZbxXUWdZfSnos4yAB4VdZbF/ZCRYXqdZQDXmV1nGcdV1VnGeRV1lo36bWadZanfKuosA2hjdp1lAG1U1FBFGxV1liEXFXWWwauizjJequosA+jP7DrLjvTbrDXCXr/NXCNU1Fk26reZdZaN+m12nWUAXGbXWQbQxuw6ywDaSJ3F3KVeFXWN2LRpk+gbMgMPdl+lDGWd5S1btog2cv5PPfWUmM/SpUtt/JjHmDFjxGf7NbmwNaKwOssfffSR+G60H4qyRkhDGXI1/p3s7jVC6h3XWTYBRa2zLG9QszF41ps0PyVvEq1vWrejW9v2LvC6uYf20rCl88Xnpy0Xaer9rztsd/b5dhRTx+pmEXftECr/+88O26G+sATqNMu6zzm3P0y+M9/P0/bSHY+LXd+c+LMU/9oACn/0W0r85D5Y0VRu4l8Un5JOPRd8RTtSrTfIPeGB9OmoJ/LULzbyJc64nxIP/01Vaw+1Hfu+XSe6qVXeLNip6WmUdeRvyvryEfKLqUcRj8+2/U6Nf/yU/rsUR/tG3kv1y1qfyEm+zHueEeWi7IF6y9P/2UxjGreksIBAUd5KQs4NKDt2KSVkWm9eFVCle7rxqOTSjUcll248Krl041HJpRuPSi7deFRylWae06dPi5cRMGhq164tjJY9e/Y4dN0F19mzZ/NsSgC1atUSO6Lnzp0TG1VGwHirX7++MIh27MibgwYuvHi5OicYZc899xzt2rVLGGzt27cXBiAMusWLF4sdT8zj9ddfp2+//Va4OMNo/P7776lu3cuhdmvWrBFt4uLixPjQL+YGt+n8gP5fffVV0f78+fPUs2dPmjp1qjCUMYYHH3xQtMPOLeQye/ZsiomJsc3JUZvJkyfTO++8Q8uWLaO5c+eKPn/44Qd67LHHxDxatmxJffv2pRdffNE2Dsypf//+YuzR0dH0yy+/iJKqzzzzjDBo27VrJwz3AwcOiP5OnjxJ9913H1WsWJF+/PFHUVHmiy++EIbp2LFjhWwwL/x+GzZsoH/++YceeOABcU4Cc37jjTfo559/FkYvHgDATfzLL7+kO+64Q7QB98SJE+m3334Thj3G/+STT9Kdd95JL7zwgthNx4MA8EP+nTp1EnzoD3LEnKdPn07/+9//hAG/YMECatOmTZHuJUd2IO8suwn4EVVglv8l6pKeSTuCLu8UB/gVXnTbaPxlFpAgK5AM83Ci3yvgoPyUTM7lW7YSRb6+RXxGGSjB5GPdCUu3FBwzbIQlO5NC7OYwpHqtK3mDgikrOJTwHMqSky1KR910001iEbqhZn3679JGig6+skyUI0NZHn+mRQfHgzLKyseHLAXI2FN1TzcelVy68ajk0o1HJZduPCq5dONRyaUbj0qu0szzySef0IQJE/Icu+WWW2jWrFl04sQJYZw44sELRhGMGyO++eYbuvXWW4UR9vDDD+c5169fP/rjjz+EEWrf7/jx4+nll192eU4jRoygZs2a0apVq8TOK4xJGFfvvvuu8Chdu3atMECxszlz5kxh5HXo0EEYft99953oA39Hoi2MU4wRO6MwKu0fBNjHGd944420evVq0R92SxHnjLnDqINx2bhxYxGLvHLlSmEI288pvzYVKlSgOXPm2NrDTRuGOQx3tLMH+Ldt20aNGjUSRjCMTrxgdMIY//zzz0U7uHZDLjCmP/nkE3Hszz//FDu3AB4ivPnmm6LN+++/L7hgmM6fP5+GDBkiZN2wYUPRFjwwpiF3yHb9+vVXxFTDk/fgwYNCRnh4sWLFCurTp494aAKXcjyUwYMOGee8bt06cR4PE2AoA0888YS4Du+Qk9ztdgfYWHYT4LqhAsG976b2MyflMZYdJfiyh7GOcYaTa2RA6OUaxBIPNmlFr7Xvkf9FvleOJcCR8Snb+fgK2cUEBtB/adad5eDC5pN9ZXkqPwe8eYzYHKsL986dO8U75vBE83ZUzkGCr6LAGLMMF/PAQH/tdE83HpVcuvGo5NKNRyWXbjwquXTjUcmlG49KrtLMc//994tdNyNgxABIOLV169Z8uWDkONpZBmBYwWAzAkaSdLe179fRrnJhc4KRCyNSuhJjd3P06NFiVxS7s9iNhPsudm0HDbKG+GGHs3fv3rRo0aI8hnrnzp2FoQzAyMTubUHJqrBjfc011whDFcBuMgy/m2++WfxNetVVVxX5d4JrsqvAHLFxhN1z7ObC7RrjA2B4N2/eXHzG7ioMz/y4IC8ALuNyBxdygYGPecFYhrcBdqKxow13aJzDzrYR2MXGWPAAQpZ7grELYxi/D5KKwZCfNGkSff3112LM+L2aNm1K7733njCysfOPHf4LFy6Ia919L7Gx7GELnF/FWuRfvyNR7IXLx/LZCc1vtzS2fA2nuAJDytDa62+hLr9+azv2SttuFOGgbFT9jASql5nocGfZYf1maRD7Wo3lN+vUpqu3bqVLfoE2Y3n21dfTTcscuMRnXZkO3ief3WibEWtnYGNMlRw8DCgy7HaW+T/xooNlx7IrCX1gvWPZsd551r3E96xa2eXn/iwNKbhc58cldxkdAe7AeDkCDNb8+nXEkx92794t3mGgwsAFYFzBvRhu4Hh3BBh4RiMfO6jYEXYFf//9N91zzz15jkmDFOMqjrFcVGCnGgY+HiAMGDBA7Ahj1x076ti1x6YSdoOx++4KZKywlJnc3W7SpEm+80GyMqBOnTpXyOiDDz4QBjaMaMj9s88+o3HjxglDHbvnMNLhzg1DGsY2PB3MkB2n7XUTZIC6CtiXSTaWhcoP/sYyTNFXuiw7QnJqGnWOqZbnWGA+XBtP/EHfnl3n2DB2AFtZKx9fIbvmFSrSg/H78tQ0Hlm3sYgfTrnrybwX547h2kqGbN25paCuQK7hDTdsU5HHWLbOSRVUcenGo5JLNx6VXLrxqOTSjUcll248Krl041HJpRuPSi5neODyDQMOLyTMgsGVn6HsCHD7dbUkEXaxZSIt+13Zwv5mNkt2iNfGQ4iPP/5YfEfcM2KKjx07ZvuMBwvFxaXcJGHSndzRfGQiuMJkBAMfMfNwDYdLf8eOHYX7Pnab4R6OMRsfZLhTdmwseyCM9ZMBfwe7ufbwyzVAryofTe926VNg23mnVtF3Z9Y4PBfoBFfX1NhC28gdaHkT+Javais7FWEwyLEjLmss2y4tW0m8L+ppdRsR/TgoASXgl/tkyWRjOU+pKi4dxWAwGAwGg1EqIHdvkZDKCMS1yuzozqB69eoFJvJyBOzY2ic/O3r0qHhHrWR3w9lNK7g2I6EW8vggcRYSqnXr1k3sLsMFG67axUX16tXFO1yl8wN2nTFmRzIy7mxjp7lr165idxljxG49YuGxAw6DuVKlSiKW2gywsewmIK5CFQJDwlw2ln1z6xhfW70ORRYSp9sjLZauTTntcE4O44/t8OvpVYW2MbprgwcGcFCusRziVzBH6I0vUFD30eQTHkWPXfqPWqRfJApyLH8faWjnZImMjUjSgHdT4eOrVB9UcenGo5JLNx6VXLrxqOTSjUcll248Krl041HJpRuPSq6CeFAjGEm2XnrpJZo2bZpw/YXrMXYr4eoNIMMzdiNlEit5DOWK5LFRo0aJRFdIkgXgOHZjZfkoR0A2aiSk+v33323XIHZ68ODBNvd0ZMgG7OsPG+fkqI0spWTcRUVMNFzLjS7OjoC5IDYcJaikmziyT8+bN4+qVKlyhTwTExPz8MhyYvLd+FnuKCM+GQYssodDRuhTPmyQMsPf5NjFfvvtt0WWbgBtECuOTNhG4PfCzjdKTMEjALHK3bt3p0cffZTuvffefGVXXLCx7Cbgh1OGgLwxwwFhhadGl2WOclzIgOhoTs4+sZI4fetDDo+HDnyS/Ko0tPFgZza45bWSpcA+fSOiKLT/Y+Ka8XH/0IqTS/Mfl3SPzs4SiShuuOEGW0IK0+Drq1QfVHHpxqOSSzcelVy68ajk0o1HJZduPCq5dONRyaUbj0quwniQRAqZmVHGCNmoEbMLwxmxrTCiYcAiQRR2MA8fPiwMsw8//FAYcNgdhgGKeFm4+iIh1sCBA0UmZpQ9wu70jBkzHI4Biaqwow3DD6WZkCAMRh7GAyAj99133y0+X3vttXnim2V/jtqghBLijWU2abghA4jbxZgxxn37rOGN+cUX33777SK7t0ysNXz4cPH3NPozonfv3uIBAQxpGNSQT9u2bW0PA5AxHInYZHw5vqPcE7J146EEynUhm7ccJwxoGMNIBAYg4zY4kOgNxi9kjN/HPqEckn0hThnjlsB4IE/I1hV9cAWc4MtNQAyDDGw3Gzk+AS7HLEtj0mIsDaVgTjGh1hvQHgENO4uXkUc+uXGl6FLoiAmUfeyffM/7BFv5LWlJos7fp59+Kp4+wV3DNPj4KtUHVVy68ajk0o1HJZduPCq5dONRyaUbj0ou3XhUcunGo5KrMB6cQ6kjvOyBBFF4GQHjWWbPNgIlk5Dh2RWgpBNejoCMz3gVNKf82mBX2x4wGu3rYecHZAKX7ujggdHsKNZ3uV2pJ8CRIY7M1vZAdmoYy5IHG1aooWy/C+zod7EHHmzExuYN9YQrtqzXbJbe8c6yB8InN/7YUfKu/GAzRItRw+/nPjeSmWhRzxqb0LROC6evCWo9gEJvfD7f86jnLIHi6njahXcz4eruO4PBYDAYDAaDwSh94J1lOyBQHC4JQ4cOFe4CeMICv3pkZcMTCumPL7O3ycQAcCnAObRF/AOezqDwuEypDwPK2BZZ3+AigLaIGZD+/WiLmmcyKxzOpaWlicLnOC6C1+3s3ZSkJLoYcFFwwh0EL/DBVUEW5c5Iz7C2TUsTx4xtATzpkWOQQPY6Y1HvLhEVbE+FMDfEXcgU/BKyP2MfkAva2ssQn3Fc8nSp14IOR1alsv4BIjaiMHkbZWgvb6MMA7vcTFkVatn6Ms4LbXE90tz/M+Bm8g8JFr+5lDdiP2TboKAgMQeZEt9e3hJoD7mgD/mEDm3xXWZRlPLGWPCkDC/ZFk/YMC/ZFr+jlCH6xZhlnApkhHHJMaItzuWns+CDPgGYm9Rvexk6kjfagKcweTvSWaN+S3nbt5U6K3+fwuSNts7K0ChvowzxHf1BBlIuhckbxx3JsKA1Ajy4Lj95u2uNQB/gkm2dkaFRvzFfo846WiOkvCFjea0jeRvXCKMM0TY/eTtaI3Ct5HFW3q6uEWiL/sDjis66ukZIeUOGaGPUWTPWCFyLNkadNWONwDVo46y8i7tGuCLvoq4R6Muo32atEfnptxlrBOaG/ozyNmuNsNdvM9YIR/pt1hoBGWJ8aKtijQDsddada4RRv+111t1rhPw/3VV5u7pGyP/7VKwRaOOszhZnjcB1aCd11qw1Qu4oO5J3kBvXCFflXZw1Quqds/JG6Sm4jDvalfexyHzeDAH4xyMLnKvAD4ofQAWe/GsJTf/3b9v3ncPuoublHdeok9gYe4o6zv+GHm3Wht7pnH827IvPW2MQAN8Xlok5+cyYIr7n3Pusw13TzH0bKOmLh8XnyNe3iD7K1xkuvlvue67UyA7xFm3atMkTV+FOSNlBBir1QRWXbjwquXTjUcmlG49KLt14VHLpxqOSSzcelVy68ajk0o1HJZduPMXhcmQHshu2m2DMnmc2ouzKJPk74fYrs2G74oZtP6f83IsDGnR0uk9neHSAyjmp4tKNRyWXbjwquXTjUcmlG49KLt14VHLpxqOSSzcelVy68ajk0o3H3VxsLLsJ9u7IZuLBxi2pQVDg5YTPzsQsu5gNu+zzC4s+Jx9f2nV0AW079lupkh2eMMHFXsVTLZX6oIpLNx6VXLrxqOTSjUcll248Krl041HJpRuPSi7deFRy6cajkks3HndzsbHsJsCXXhXKhIbS9eUuF972s0v45QhyU7hQYzm3/rFvuZiiz8nHh6pkp1GtLGt8Q2mRHVLLw7UC7zrpgyou3XhUcunGo5JLNx6VXLrxqOTSjUcll248Krl041HJpRuPSi7deNzNxcaymyCDzkuCS7pYFwTZprB95cCW1+XL4zSc2OkuCdkhcQFqw+HdDAQ072OrgV2S+sA8LDvWh9J7H6nk0o1HJZduPCq5dONRyaUbj0ou3XhUcunG424uNpY9FT5XulgXBD9f537q0GFjqdzEv4ozMpeNZVX4559/qEGDBuLdDISNfIXKTVhtSt8MBoPBYDAYDAZDLbh0lJugqri7Iy5njOWmkVH0YqtO9ETzdgW284EbdmBI8ebkYp1hlbIzEz7+AaVCH5iHZcf6UHrvI5VcuvGo5NKNRyWXbjwquXTjUcmlG49KLt143M3FxrLHZnjzcclYRptJ7boXgafoO8shA580l6cUw1Mz/nkTj0ou3XhUcunGo5JLNx6VXLrxqOTSjUcll248Krl041HJpRsPwNmwSyFksWtVXMacXsF+/qVrTrnGe3DXm83lKcVQrQ/Mw7JjvfOM+0gll248Krl041HJpRuPSi7deFRy6cajkks3Hndzlc7gUkahCM6NQX4mKoIqBFvdpksL/Ks0LOkhMBgMBoPBYDAYDEaxwMaym1CuXDmlXNJYbhUSaCpPUVDm9rco/LHvTedxFS1atKBLly6Jd930gXlYdqx3nnEfqeTSjUcll248Krl041HJpRuPSq7CeA4dOkTvvfceDRgwgMLCwhzuPK5du5Z69OhBvXr1onbt2tHixYtd5tFRdu6Cp8qOjWUPTYcenOvqnJZjMZWnKPAJLkP+leubzuMq/Pz8yMfHR7ybDU9Nj+9NPCq5dONRyaUbj0ou3XhUcunGo5JLNx6VXLrxqOQqjGfv3r1isyQgIIBSUlKuOP/ff//RddddR6+88gqtWLFCGNaDBw8WBrQrPDrKzl3wVNl5jbH8xhtvUIcOHah27dr0xx9/uL3/7Oxst/dZEFdQ7s5yuonB8qrmpIpn//79dOONN4p3s6FaH5iHZcd65xn3kUou3XhUcunGo5JLNx6VXLrxqOQqjAeGcPfu3ally5YOz0+ePJlatWol2gAdO3akrl270rhx41zi0VF27oKnys4rjOWNGzcK5cf7jBkzaOTIkW7/wfz91SUWB5d0wzZzZ9mlOWHHtnID83mK+ZQJTwtVPNlSrQ/Mw7JjvfOM+0gll248Krl041HJpRuPSq7SzJN5/gylHthj+5566D/KOHNCfM5JT6OU/3ZQdrL176vMC7GUsm+XjSvtyH5KP31cfLdkZVrbJiVY2148Tyn/7bT1m3bsIKWfPGptm50t2mYlXBLfsy7FiXG4a05GYCMNm2pGwGBevXp1nr8bJc+sWbOoefPmdP3119Ps2bNp1KhRVL16dfH9/Pnz9MILLwgDPSYmhqZNm6adPnjTPVsqjOXC4ggsFgu9+uqrQomhuMOGDaMzZxzfLI4grwM6deok3C/cbSyrrh1WPlcJgk38BY1zKpOTWWDbcpPWU/jDXxebRxd4ai05b+JRyaUbj0ou3XhUcunGo5JLNx6VXLrxqOQqzTzn5n5F+x8dbvt++H/30Jlv3hefM2JP0b+39qKUf3eI7xcWzaZ9Y663cR2Z8BCd/myq+J516YJom7R9g/h+8c/59N+d/Wz9Hnv9aTr5wSvic05qsmibuGml+H7pr9/FONw1Jwnkszl37hxVqVIlz3F8z8rKoiNHjlzBc+uttwq37gULFlDVqlXp+++/p82bN4s4Z9gdDz74oPgMQ/n555+ns2fPaqUP3nTPlgpjubA4gkmTJtEPP/wgdgU3bNhAlSpVomuuuUYosKvYunUrDRo0iAID3ZsYKyHB+oRMBcB1XbkI+jh2I40ua57iyTn9c80N9G+vvgW29fHzF6/i8OgE1frAPCw71jvPuI9UcunGo5JLNx6VXLrxqOQqzTzRQ++g+u/OsX2v/epnFDP6YfE5sGIVajxrBYU2tiZQrTDgJmrw8a82rlrjP6DK9zwjvvuXqyDalmlp3cSK7HsjNfpyia3fGs9Po6oPjRWffUPCRNvw9j3F93LdrhXjcNecJKTtERQUlOd4cHCweE9KSnLIg91LJATr1q2b+I5dZLxGjx4tdpkBbNZhgw4bgzrpgzfds+r2wwsA3BSA5cuX0y+//HLFZKdMmSJ2nuVTArg2QAnxFKdx48Z0zz33OOwXrtfvvvuu7TuU9f33389zzFOBRFUjko6RX26iLzPRrGYj0zkYDAaDwWAwGKUTAVEx4iURUufy34a+QcEU2uhypZGAChXFSyK41uWkrz7+AXnbRkaJl61tjbqX2/r55WnrX648mYGoKCt/enp6nuMZGRnivWLFy3MpDPZJZKU7sH3fDM9BqTCWC8Jff/1FycnJeeIIqlWrJlweFi5cKJ7ebN++3am+xo4dK1whcK27ERKirtax4JJGsnkhy8rmpIoHD1imT59ue9pnJpTrA/Ow7FjvPOI+UsmlG49KLt14VHLpxqOSSzcelVzF4YG3aeXKlenUqVN5juM7zhntBpZd0eGpsiv1xrLMXOwojuDAgQNO9wMjCe7Xbdq0oX379gkDe8SIEVe0O378uG2nG0+DkDYeL3zGzrbc1pc/QmpqqnjHzYRz2L3GU6UyZcpQfHy8zY0DO8GybUREhHD5gBs52oaHh4t4CdnW19fX5hKCc4jhzszMFMdxLdoGJSeL89lZWXTx4kXxGZx4CoYX+FBjTJ6Dawnc3KUribEtEBkZKfpFfDjmgpfgCQoSceQYq3wqhraYW05OjugTY5bJD9AWMpBx5xgD5IK29jLEZxxHX+ApW7asGB+uL0ze9jK0l7dRhrIt+rz55pvFE0QpF7TF9Xgg40jeGJNRhuhDti1I3pgX+jXKG30aZWgvb9nWXt5GGdrLGzKCTKRc0Bbn8pMh+ORvY5S3vQwdyRtt8b0weeens4XJW8pQ6p0z8nZWhkZ5G2WI8+XLl89XZx3JG8cdybAgnQUPnkznJ293rRF4QW6yrSs66+oaATnI8Zq5Rhj128w1Quq3Kzpb1DUCckI/Rp01Y43AefRj1Fkz1gjwREdHOy3v4qwRkLEr8i7qGoF3o36btUagnSP9NmONQN/gNMrbjDUCXPb6bcYa4Ui/zVojMEaMD23NXCOM+m2vs+5cI4z6ba+z7l4j5P/prsi7KGsExoFz9jrrSN5yXHCZljLs378/rV+/3nYtZIiyUX379hWyQt9Sv9EPvqMvzAfvUmchS/Qn9VtyQ6bo25U1Ajph1Fmz1gjwop0jeQe5cY2QMnR2TS7OGiH1rjB5//rrryJZGwA5nT59mq6ApRRh/Pjx2Ce1pKam2o5NnjxZHEtJScnTtlu3bpYGDRo41e/3339v8fX1tQQFBYmXj4+P5Z9//nHYdtCgQUUae1xcXJGuKypX+s4/LXHPtbGkrpxpKo8KqOK5cOGC5eOPPxbvZkO1PjAPy471zjPuI5VcuvGo5NKNRyWXbjwquXTjUcnlLM/o0aOFXbFr1648x3fu3ClshJUrV4rv69ats/j7+1tWrVqVL8/gwYMtLVu2tH3PzMy0REZGWh544AHbsY0bNwq+OXPmmDan4kI3nuJwObIDS/3OsjGOwLiljqcUzsYQ3HTTTeLlDOTO8tChQ+muu+5y+gklnlqo2lkWY8ix/nTZIWVN21mWczV7Z1l+N3tneefOnTRmzBhq27at7WmdWTvLGCfOq9hZRjvJa+bOspSn2TvLcm4qdpYhGzN3jaQMcczsXSPIEP1Lubiqs66uERib8Qm8WWuEUb/NXiMAFTvLGLOZu0ZG/TZ710jqt9m7RnKNAFTsLGPORv02a43IT7/NWCPAa+aukZShI/02Y41wpN9mrRHoT8XOslG/zVwjjPpt9hohrzV7jQCPI52V8n7jjTfoxx9/tCXagqdps2bNREnZ2rVrU506dWju3Ln0xBNPCA5c89VXX4nyUOjPqN+Q4cMPPyzyLEF2+LsS5aOwO422n332mWh3yy23CFsCuO+++8T1CB91VmehFyq8T9Af+jV7ZxlQtbNsb2sUJm/89vPnz3e4s+wDi5lKCV5++WWaMGGCzd0BWLJkich8vXv3bmrSpImtbY0aNahnz5709ddFK1eUH1AfDVvyrgI/BBRMBSRX5v6N5F+vvW2BNYvHbKji2bZtm1gckRG9devWpnKVhD4wD8uO9a7030cquXTjUcmlG49KLt14VHLpxqOSSzcelVy68RSHy5EdWCpKRxUEpGTHEwOUjJI4efKk2AHG7m9pgTGtvCqugPodTDOUjTw6yU5nfWAelh3rXem/j1Ry6cajkks3HpVcuvGo5NKNRyWXbjwquXTjcTdXqXLDlq4RBw8epKZNm9q29OE6iwRdI0eOFFvvkydPpkaNGgl3B3fDE9yw4e4AHjNdLPGSPGa7Rkges10sZV+Yo9kJvjBOjFmFGzbmBpmZ7YYt+1WR4MtZ/XaHGzY+q3DDlm75ZrthG9uauUYY3UfNXCOM+m3mGiH1W5UbNvpU4YaNMalww8aYVLhhG9dvM9cI9KHCDTs//TZjjcB8MD+z3bAxZnv9NssN216/Pd0N26jfZrphG/XbbDds+Tee2W7Y4ME8nEnwVZw1AmPGmJzV2eKsERgnxsRu2AEurxFS77Rxw544cSLNnDnTZixDuRFHAP//evXqCWVF2afFixcLpUOs8scff2xKCaCiumHLmC0VUMWlG8/evXtFrMg333xDDRs2NJWL9YFlVxL6wHrHsmO986x7ie9Zlh3rnWfdS3zP6i07R3ZgqTCWSxM8KWaZeVh2rA98L/E65Fnrqkou3XhUcunGo5JLNx6VXLrxqOTSjUcll248Xhez7CmQriQ6cenGo5KL58SyY73zrHuJ71mWHeudZ91LfM+y7FjvPOteivfQv/dLVcxyaYAnxCzLMZgds6xb6Shkwe7Vqxdt2bJFlAiQbbl0FJeO4tJRXDqKS0dx6SguHcWlo0o6Zhng0lHmlI5yV8yyLB2lKmaZS0flcOkoXdywocyy3JXZUMWlG4/K0lGsDyy7ktAH1juWHeudZ91LfM+y7FjvPOte4ntWb9mxGzaDwWAwGAwGg8FgMBhOgGOW3QTpHqwTl248KsH6wLIrCX1gvWPZsd551r3E9yzLjvXOs+4lvme9T3Ycs2wHjlnWN2ZZdZ1lWVfX7DrLxnq3ZtZZlvI0u86ynJuqOstm1lCVMsQxs2uoyjrLUi5m5zXA2FTUWTbqt9lrBKCqzrKZNVSN+m12DVWp32bXUJVrBKCizjLmrKLOcn76bcYaAV57eZuxRjjSb7PqLBd1TXZ1jeCY5aKvEVL+Zq8RHLNc9DUCvx9+R0fyDnLjGgG4sia709bw+DrLOsQsy4VCBVRx6caDG+Xo0aNUs2ZN02MmWB9YdiWhD6x3LDvWO8+6l/ieZdmx3nnWvcT3rN6y45hlEyGfzujEpRsPDOQqVaooSS7A+sCyKwl9YL1j2bHeeda9xPcsy471zrPuJb5nvU92HLPsJsB1QDcu3XgOHz5Md999t3g3G6wPLLuS0AfWO5Yd651n3Ut8z7LsWO88617ie9b7ZMfGspugyq0AmDdvnlZzUsWDmIo5c+bYYivMBOuDZ8hONx1XtTboKDsd9U63/ytUcunGo6M+sN4VHfx/BetdSeiDn4f+X8EJvtyU4AuB8zhnZvIeGXS/Zs0aGjx4sKnJe/ACJ643O+he8uiU4AtykQmXzE7w9ccffwh9MDvBl5Sb2Qm+ZOIoFQm+IFP0Z3aCL1yDl9kJvlatWkV33nmnkgRfaK8iwReOqUjwBYBHRYIvzBVtzE7whf8rbrnlFtMTfKF/tFGR4AttVST4wjhUJPjKT7/NWCOgD/jbxuwEXxi/vX6bsUbgO/pTkeAL5zA+tDVzjTDqt5lJABcsWGD7u8HsJIDybzwVSQDRxl5n3b1GyGRYzupscdYIXId2UmfNWiOgDyNGjDA9wVfZsmWVJfiSeueOBF8YMMOAQYMGFUken376qTI5XnvttUp4VM1JFc/WrVuRzE68mw3WB8+QnW46rmpt0FF2Ouqdbv9XqOTSjUdHfWC9Kzr4/wrWu5LQh0894P8KR3Ygu2G7CXgioRtUzYll5xnQUR90nJMq6CY7HfVOFXSUnW48KqGj7HSckyroJjsd9U4V5nqo7Lh0lB2aNm1KdevWdVmQJ0+epKpVq7rrdykVXLrxqOTiObHsWO88617ie5Zlx3rnWfcS37MsO9Y7z7qXTnrA3/sHDx6k3bt35znGxjKDwWAwGAwGg8FgMBh2YDdsBoPBYDAYDAaDwWAw7MDGMoPBYDAYDAaDwWAwGHZgY5nBYDAYDAaDwWAwGAw7sLHsIUBNsldffZU6dOhAHTt2pGHDhtGZM2dKeliMEsSECRNEjTj5Qh262NhY/k00x6FDh+i9996jAQMGiFqDsnagBK8V3oXC9AHgtcJ7cPbsWbr//vvp2muvpVq1atFtt91GFy5csJ3n9cG7UJg+ALw+eAdQV3j8+PE0cOBAkcj4lltusdUtBnhtyB9sLHsIJk2aRD/88AOtWLGCNmzYQJUqVaJrrrlGFPBmeCf+++8/2r59u+21Y8cOqlixYkkPi2Ey9u7dSy1atBAPR1JSUq44z2uFd6EwfQB4rfAO5OTk0MSJE2nq1Kn0+++/07Zt22jr1q00cuRIWxteH7wHzugDwOuDd+D1118XG20LFy4UevDXX38J/ZDgtaEAuKHuM8NkxMfHW8LCwixffPGF7djx48ct+Pm+/vprlr8XIiMjwzJq1KiSHgajBDF+/HixBqSmptqO8VrhvXCkDwCvFd6D2NhYy4EDB/Ice/PNN4VenDlzhtcHL0Nh+gDw+uA9OH36dJ7vI0aMsAwYMEB85r8dCgbvLHsA8PQnOTlZuGBLVKtWTdQPwxMihvcBT4jhZTB8+HDx1PjIkSMlPSRGKQCvFQx78FrhPYiOjhbulUYEBQXZXDB5ffAuFKYPAK8P3oOYmJg83//991/hkg3w2lAw2Fj2AOzfv1+8V6lSJc9xfD9w4EAJjYpRkoD7fdu2bYX71LPPPkuNGjWir776in8ULwevFQx78Frh3di4cSM1b95c/L3A6wPDqA8Arw/eB/zmL730Eo0YMYLGjBkjjvHaUDDYWPYAyDg0+URQIjg4mJKSkkpoVIySRJcuXWjBggX0zz//iEWudevW9OCDD3LSNy8HrxUMe/Ba4b04fvw4zZ07l6ZNmya+8/rg3bDXB4DXB+8C8ttgc2Xy5Mn0zTff0C+//CKO89pQMNhY9gBERUWJ9/T09DzHMzIyOKETg+rVq0ezZ88WWXC3bNnCEvFi8FrBKAi8VngP4GZ7xx130PPPP0/9+vUTx3h98F440gd78PqgP1q2bCk8Unft2iXWg6FDh4pkX7w2FAw2lj0ASPcPnDp1Ks9xfK9du3YJjYpRmlCjRg3x4MT+gQrDu8BrBaMw8FrhHYCnUYMGDWjs2LG2Y7w+eC8c6YMj8PrgHWjatKnYVfb19aVly5bx2lAI2Fj2APTo0YPKlSsnSkZJnDx5UrjU4KkQg4EYFLjRtG/fnoXhxeC1glEYeK3QH4hHRM3UDz/8UHzH52+//ZbXBy9FfvrgCLw+eA+wmxwZGSkMZf7boWCwsewBQKwygvCnT58usmIDiDdA3EH//v1LeniMEgAKy69atcr2/Z133hE6Ur16df49vASHDh0S7wcPHrQd47XCe+FIHwBeK7wL+Dth06ZN9MQTT4ga3Mh4O2XKFDp9+jSvD16IgvQB4PXBezBr1ixRb1tiyZIlwvV+yJAhvDYUAh/UjyqsEaPkgad9cJ9ZvHgxBQQECJfbjz/+mI0jL0ROTg716dOH1q1bJ+JP8NDkxhtvFC+G/pg4cSLNnDnTZhzBQG7WrJmIW8d/fLxWeBcK0oc6derwWuFFgDtl3759xc6hPZYvX069evXi9cGLUJg+YDeR/5bwHtx22230ww8/CHd8JIVFVvSHHnqIQkJCxHn+2yF/sLHMYDAYDAaDwWAwGAyGHdgNm8FgMBgMBoPBYDAYDDuwscxgMBgMBoPBYDAYDIYd2FhmMBgMBoPBYDAYDAbDDmwsMxgMBoPBYDAYDAaDYQc2lhkMBoPBYDAYDAaDwbADG8sMBoPBYDAYDAaDwWDYgY1lBoPBYDAYDAaDwWAw7MDGMoPBYDAYDAaDwWAwGHZgY5nBYDAYDAaDwWAwGAw2lhkMBoPhzdi1axcNHjyYgoODycfHh6Kjo6l+/fpUpUoV8erXrx+99957lJiYSJ6EnTt3UoUKFeiee+4p6aF4BP7991968MEHqXnz5tS7d++SHg6DwWAwSiF4Z5nBYDAYXoVmzZrRvHnz6Pnnnxffx48fT/v376dTp07R+vXrqWvXruJc06ZNafPmzeQpgHF/6dIlOnHiREkPxSNQvXp1euCBB+jo0aOUk5NT0sNhMBgMRikEG8sMBoPB8ErUrFnT4bFx48bRsmXLKDY2lq677jphRHsCunTpIsb666+/lvRQPAJlypQRu8rly5cv6aEwGAwGo5SCjWUGg8FgeCXggp0fOnbsSM8++yxduHCBXnnlFfIUVKpUiQIDA0t6GAwGg8FgaAE2lhkMBoPBcID77rtPvP/000953HR/+OEH6tOnD7Vr107sRD/66KOUkZEhzk2aNIn8/PyEIY6dy3feeUccj4uLoxo1aojj2MlE3PS6deuoe/fuwjDHDufDDz9ML730Ur6/xcmTJ2n48OGCt0WLFnTLLbfQqFGjxLlNmzaJz+B4+eWXbdd88skn1LZtWwoJCaHDhw/T2LFjqW/fvhQVFSWuT0lJuYJn5syZtnEhhhtySE5OzjOO2267jTp37ky1a9cWskC8dGFYuHChiA3u0KGDkNvjjz9Oqamp4tz27dtF/HBkZKSQ4VtvvUVDhgwRMdgYB84bkZaWJmQFV3nIo0GDBuLhhv180tPTxZw7depEV111FdWtW1f07QgrV66kO++8k1q2bEkVK1YUsmMwGAyGl8PCYDAYDIYX4ssvv7Tgv8H33nsv3zbVqlUTbQ4ePCi+z507V3zft29fnu8fffSR7ZoJEyaIY88//3yevuLj4y2BgYGWY8eOWRISEixly5a1/Pjjj+Jcenq6Zfjw4Zbnnnsu37H07t3b8tBDD9m+T58+3dKhQwdb37NmzRK848ePz3Pd008/LY537drVcubMGXFs7dq14tjkyZPztH3llVcs/fr1syQmJorvr7/+umj3yCOPiO8XLlyw1KhRw7Jo0SLxPTMzU7SvWLGi5eLFi/mOffbs2ZZmzZpZzp07J77v2rXLEh4ebrn33nttbfbu3Su4ateubdmyZYs4hj5btmxpiYqKssTFxYlj2dnZQhY4Lvs7fPiwpUqVKpYePXpYsrKyxLGcnBxL3759Lf/73//EZ+C6664THPjdJGrWrCmOTZs2TfQN3HfffRY/Pz/L8ePH850Tg8FgMPQH7ywzGAwGg5EPKleuLN7PnTsn3hHLjF1a7FAC2P0MCwsTu5ISjzzyCIWGhtLixYvz9IU2AwYMEImlkIk5Pj5e7PACcJ1+/fXXC3QNR/Ix7LRKPPbYY1SrVi3xOSIiQsQsOwLGB3z66afCTRvArnCdOnXo999/t7U7fvw4TZw4kd59912xKw5g17xNmzbk7+8vvr/22mti7v379xffcfyFF14Q8d2ff/65Q/7s7Gwx1meeecY2X+wIQ3afffaZTbbSfXz06NGCEyhXrpxwgz9//jzNmDFDHJs1axYtX75cyEv2Bzk8+eSTtGrVKvr+++/FsR9//FHIecKECTa5TpkyhapVq3ZFQi/s1D/11FPk62v9s2jkyJFi3EuWLMn392AwGAyG/mBjmcFgMBiMfCANOLhWA0888YQwwqRRBcCAhbEoAVdiuDjv2LFDuFpL/PLLL3TzzTeLz3BDhqF566232gxWGK8wRvNDvXr1hLH39ttvU1ZWljAAZ8+e7fRvh1JZ9g8CTp8+bfuOxGBo07BhQ9sxPBjYsmWLzXV5wYIFtGLFCuEaLV8wgqtWrUpnz551yIuM4jgHd27jdXAdh5s3slEbYf/AAG7bwJo1a8T7nDlzxDtcxY0YNGiQzd1bus/DpVoa+gDc3fFQYNiwYXmuhVFuhHyocObMGYdzYjAYDIZ3gI1lBoPBYDDyAUoxAajFLA3WgQMHUmZmpth1xM4rdkbtdyqxuwx89NFH4h3nV69eLa6Vxtj7779vy7iNHWeUryoI2IUNDw8XBjvib//8889i/W4wIjEPCfDbG9T2gKGJslobNmywvWAMo1zVG2+8ke81AB4eGK/bs2ePuA4x1QVBZqtGsjXg0KFDYicdhrwRMNiNfJiPfRtnIXfjZSw6g8FgMLwTbCwzGAwGg+EAMJQOHjxIMTExecpMvffee8JtF7vGSEqFZFD2wA4mdj6xCwoXYrRF/WajMXr//feLXdurr76afvvtN7ELivf80L59e9q9ezfde++9tG/fPurXr1+eZF7FBZJmYayOkn4Zd32PHDniUr9yp9jV6ySkQS+NZjx4gIu0PeQOsjSQMR/7XWsGg8FgMFwBG8sMBoPBYDjAH3/8IQwumXEaQMZlvH7++WdhqDZq1Chf2WF3GdmYv/jiizwu2DJ+Ge7cMLqXLl0q+oNRiSzTxt1eI6RhjthdZIdGBmjE4/79999u+f2wM2uxWES2b3ts27ZNnENsMNzGkd3bHtgtdgQZV/3tt99ecQ5uzsjSXRCQfVs+gJDu6sjOffHixTzt5JhwXs4HDyPwwMMeW7duLZCTwWAwGAyAjWUGg8FgeCXsXaeNSEhIEKWNsKv83HPP2Y7DrRpGW0FGssSNN94okknNmzePNm7cKMomGWFMADZ48GBhKMPV2JEhKo13GKxAs2bNbPWfscvsDlxzzTXiHSWZjKWaEPuLWGkY8xgnXNNRwkrGO2MHHjKCYeoISNaFklbffPMNvfnmmyLeGoARi36ky7OE/cMCJPNCjDjiwI2xyUi2ZoTkHzp0qG0++I3vueceW0y5LCWFuGsGg8FgMAoDG8sMBoPB8ErIHUsjYEwtWrRIJJWCoQWDViZ7AvAZsbBwVwZgBMPVF9cBMrOzdAtGUivsuLZu3dqWJEwCWZsRxwzACEasLTJNO3LrlrG6U6dOtX0HLzhQZxiQxuupU6cczlPG8gKYG9rBOJdu13DzRuIrHG/VqpUwcJFtGkb8uHHjRBsk88LOLQxYnEedZSQ4w+64rEttDxjZyLANg/fpp58W7XEddsZvuOGGK+b7wQcf2BKjYSyTJ08WGbfxgAC46667xJxh1MsEXElJSaIdjHlp9GMnHjzYxcdDC4wb3NjRR01r+Xvj4QQSkMkHEUaX8cJ2vRkMBoOhOUq6dhWDwWAwGCrxzz//WIYNG2YJCwsT9XVr1aplad++vaVJkyai5u7VV19teffdd221ho1YtmyZpX79+pbKlStb+vfvb3njjTcsL7zwgiU4ONgyatQoUSvYCNT/BcfGjRvzHF+xYoVl4MCBlo4dO1quuuoqS+PGjS2DBg264nojWrRoIa5p1KiRqDGMGsuLFy8W57766itL+fLlBZePj4+t/jLa+/r6iuMRERGWV199VdSIrlOnjjiGV9WqVS0rV64U7dPS0kRdZtQsxpzatWtnWbVqVZ5xoFbz7bffLmofh4SEWIYMGWKr31wQ/vjjD0vnzp0toaGhlpiYGFHXWNY/Nsqqbdu2lqZNm4rfBPwzZsy4oq9Lly5ZHnvsMfFbYK6tW7e2TJw4UdR9NgI1rVG/OjIy0lKmTBnL9ddfL44B27dvF3OXckBfqKf94YcfClnJ4xgDg8FgMLwTPvinpA12BoPBYDB0BHZzkcDLXa7SOgO7udgJRr3jadOmlfRwGAwGg8FgN2wGg8FgMMzC119/LdyGGQwGg8FgeB6sdRYYDAaDwWC4FUgShtrI+SW+YuSFdHRjhzcGg8FglBawscxgMBgMhhtdiZH1GsmrkCUau8pIKsUoHEhgBuzatYvFxWAwGIxSAc6GzWAwGAyGmxAUFCRe69evp5YtW9KLL77IsnUC06dPF5mxgSVLlojazMhKzmAwGAxGSYITfDEYDAaDwWAwGAwGg2EH3llmMBgMBoPBYDAYDAbDDmwsMxgMBoPBYDAYDAaDYQc2lhkMBoPBYDAYDAaDwaC8+D+YA1OMZIKM8AAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(10, 5))\n",
    "for name, (err, _) in results.items():\n",
    "    ax.semilogy(tdays, err, linewidth=1, label=name)\n",
    "ax.axvline(FIT_DAYS, color=\"k\", linestyle=\"--\", linewidth=1, label=\"end of fit window\")\n",
    "ax.axhline(10, color=\"#CC3311\", linestyle=\":\", linewidth=1, label=\"10 m\")\n",
    "ax.set_xlabel(\"Days since epoch\")\n",
    "ax.set_ylabel(\"3D position error vs IGS final (m)\")\n",
    "ax.set_title(f\"GPS G{PRN:02d} from {START}: Propagation Error with Fitted SRP\")\n",
    "ax.grid(True, which=\"both\", alpha=0.3)\n",
    "ax.legend(loc=\"lower right\")\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0fdf565",
   "metadata": {},
   "source": [
    "## Benchmark: 24-hour predictions across the constellation\n",
    "\n",
    "The IGS *ultra-rapid* product is the operational reference for short predictions: its predicted half is accurate to ~5 cm (1D RMS), and Duan & Hugentobler (2021) report 8–10 cm 3D for 24-hour predictions from 3-day arcs with CODE's full force model. To compare like with like, fit 2 days and predict the following 24 h for ten satellites across blocks, including two that cross Earth's shadow in this window (G08, G03 — propagated with Gauss–Jackson 8)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "fce6db31",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-29T14:37:34.804524Z",
     "iopub.status.busy": "2026-08-29T14:37:34.804432Z",
     "iopub.status.idle": "2026-08-29T14:37:50.911437Z",
     "shell.execute_reply": "2026-08-29T14:37:50.911192Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " PRN   beta shadow        cannonball      ecom-reduced             ecom2\n",
      "      (deg)    (%)    fit / 24 h (m)    fit / 24 h (m)    fit / 24 h (m)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G20   17.7    0.0    1.669 / 12.318    0.049 /  0.055    0.047 /  0.057\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G05   14.2    0.0    0.780 /  5.808    0.048 /  0.062    0.044 /  0.062\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G01   33.6    0.0    0.151 /  0.617    0.053 /  0.069    0.050 /  0.069\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G14  -40.0    0.0    1.881 / 13.673    0.034 /  0.050    0.034 /  0.050\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G18   32.6    0.0    1.774 / 13.014    0.048 /  0.067    0.046 /  0.067\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G11   32.2    0.0    1.545 / 11.080    0.047 /  0.075    0.044 /  0.070\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G25  -42.0    0.0    0.146 /  0.507    0.041 /  0.057    0.040 /  0.053\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G30  -75.9    0.0    0.162 /  0.857    0.043 /  0.102    0.042 /  0.108\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G08    3.9    7.6    0.203 /  0.581    0.047 /  0.056    0.041 /  0.053\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G03   12.9    0.8    0.182 /  0.522    0.042 /  0.220    0.041 /  0.219\n",
      "       median 24 h             3.333             0.065             0.065\n"
     ]
    }
   ],
   "source": [
    "def beta_angle(pos, vel, t):\n",
    "    h = np.cross(pos, vel); h /= np.linalg.norm(h)\n",
    "    s = sk.sun.pos_gcrf(t); s /= np.linalg.norm(s)\n",
    "    return np.degrees(np.arcsin(h @ s))\n",
    "\n",
    "\n",
    "def in_shadow(t, p):\n",
    "    s_hat = sk.sun.pos_gcrf(t); s_hat = s_hat / np.linalg.norm(s_hat)\n",
    "    along = p @ s_hat\n",
    "    return along < 0 and np.linalg.norm(p - along * s_hat) < sk.consts.earth_radius\n",
    "\n",
    "\n",
    "gj8 = sk.propsettings()\n",
    "for attr in (\"gravity_degree\", \"gravity_order\", \"tide_model\", \"use_relativistic_correction\", \"use_spaceweather\", \"abs_error\", \"rel_error\"):\n",
    "    setattr(gj8, attr, getattr(settings, attr))\n",
    "gj8.enable_interp = True\n",
    "gj8.integrator = sk.integrator.gauss_jackson8   # immune to the shadow-boundary step-rejection abort\n",
    "gj8.gj_step_seconds = 60.0\n",
    "\n",
    "BENCH_FIT_DAYS, BENCH_PRNS = 2, [20, 5, 1, 14, 18, 11, 25, 30, 8, 3]\n",
    "print(f\"{'PRN':>4} {'beta':>6} {'shadow':>6}   {'cannonball':>15}   {'ecom-reduced':>15}   {'ecom2':>15}\")\n",
    "print(f\"{'':>4} {'(deg)':>6} {'(%)':>6}   \" + \"   \".join([f\"{'fit / 24 h (m)':>15}\"] * 3))\n",
    "bench = {name: [] for name in VARIANTS}\n",
    "for prn in BENCH_PRNS:\n",
    "    bt, bp = load_truth(START, BENCH_FIT_DAYS + 2, prn)\n",
    "    nf = sum(1 for t in bt if t <= bt[0] + sk.duration.from_days(BENCH_FIT_DAYS))\n",
    "    shadow = np.mean([in_shadow(t, p) for t, p in zip(bt, bp)])\n",
    "    use = gj8 if shadow > 0 else settings\n",
    "    row = []\n",
    "    for name in VARIANTS:\n",
    "        c0, cstep, build = VARIANTS[name]\n",
    "        x0 = np.concatenate((bp[0] / 1e3, initial_velocity(bt, bp), c0))\n",
    "        steps = np.array([1e-3] * 3 + [1e-4] * 3 + cstep)\n",
    "        def resid(x, build=build):\n",
    "            st = np.concatenate((x[:3] * 1e3, x[3:6]))\n",
    "            res = sk.propagate(st, bt[0], bt[nf - 1], propsettings=use, satproperties=build(x[6:]))\n",
    "            return (np.array(res.interp(bt[:nf]))[:, :3] - bp[:nf]).ravel()\n",
    "        sol = least_squares(resid, x0, jac=lambda x, r=resid: approx_derivative(r, x, abs_step=steps),\n",
    "                            x_scale=\"jac\", ftol=1e-12, xtol=1e-12, gtol=1e-12, max_nfev=400)\n",
    "        st = np.concatenate((sol.x[:3] * 1e3, sol.x[3:6]))\n",
    "        res = sk.propagate(st, bt[0], bt[-1], propsettings=use, satproperties=build(sol.x[6:]))\n",
    "        d = np.array(res.interp(bt))[:, :3] - bp\n",
    "        days = np.array([(t - bt[0]).days for t in bt])\n",
    "        m = (days >= BENCH_FIT_DAYS) & (days < BENCH_FIT_DAYS + 1)\n",
    "        fit_rms = np.sqrt(np.mean(np.sum(sol.fun.reshape(-1, 3) ** 2, axis=1)))\n",
    "        pred = np.sqrt(np.mean(np.sum(d[m] ** 2, axis=1)))\n",
    "        bench[name].append(pred)\n",
    "        row.append(f\"{fit_rms:6.3f} / {pred:6.3f}\")\n",
    "    print(f\"G{prn:02d} {beta_angle(bp[0], initial_velocity(bt, bp), bt[0]):>6.1f} {shadow * 100:>6.1f}   \" + \"   \".join(row))\n",
    "print(f\"{'median 24 h':>18}   \" + \"   \".join(f\"{'':>8} {np.median(bench[n]):6.3f}\" for n in VARIANTS))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e15b3fcf",
   "metadata": {},
   "source": [
    "## What this shows\n",
    "\n",
    "* **Inside the fit arc, ECOM tracks the IGS orbit at the level of the orbit's own accuracy** (~5 cm 3D over 3 days, ~8 cm over 7); the cannonball cannot do better than a few metres.\n",
    "* **24-hour predictions are at the 6–7 cm level** (median over ten satellites, 2-day fits), against ~5 cm for the IGS ultra-rapid predicted product and 8–10 cm reported by Duan & Hugentobler (2021) from 3-day arcs with a full analysis-centre force model. G08, which spends 8% of the window in umbra (propagated with Gauss–Jackson 8), is no worse than the others; G03, which only grazes the penumbra, is the outlier at 22 cm.\n",
    "* **Beyond a few days the error is almost entirely along-track** (the T column) and grows roughly as $t^2$ — the signature of a small, slowly changing acceleration bias. From a 3-day fit the 10 m line is crossed after about 12 days and the 30-day error is ~100 m; from a 7-day fit it is ~15 days and ~55–65 m.\n",
    "* **Constant coefficients do not hold for a month.** The in-window fit RMS grows with arc length (5 → 8 cm for 3 → 7 days) and the B terms change between windows: the true SRP coefficients drift with the Sun elevation angle above the orbit plane, $\\beta$, over weeks. This is why analysis centres re-estimate ECOM daily, and why ECOM2's extra harmonics do not help — the limitation is temporal, not spectral.\n",
    "\n",
    "For prediction, treat ECOM as a **short-arc model**: fit on the most recent few days, expect centimetre-level accuracy for a day, metre-level for a week, and refit as new orbits arrive. A target such as \"< 10 m over a month\" is not reachable with fixed coefficients (nor is it with any operational product — the IGS ultra-rapid predictions are re-issued four times a day for the same reason). Rust users can supply per-arc coefficients through `SatProperties::srp_ecom`; from Python, run one `propagate` per arc.\n",
    "\n",
    "An earlier version of this tutorial read the SP3 epochs as UTC; that 18 s error produced a 13 cm fit residual and 30 cm / 9 m errors at 1 / 7 days, which were (wrongly) attributed to model mismatch.\n",
    "\n",
    "The script `python/examples/ecom_gps_validation.py` in the repository runs both the single-satellite analysis and the multi-satellite benchmark for any PRN, window and fit length from the command line."
   ]
  }
 ],
 "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": "3.14.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}