def write_hesuvi_zip(): manufacturers = ManufacturerIndex() zip_object = ZipFile(os.path.join(DIR_PATH, 'hesuvi.zip'), 'w') dir_paths = [ os.path.join(DIR_PATH, 'oratory1990'), os.path.join(DIR_PATH, 'crinacle', 'gras_43ag-7_harman_over-ear_2018'), os.path.join(DIR_PATH, 'crinacle', 'ears-711_harman_over-ear_2018'), os.path.join(DIR_PATH, 'innerfidelity'), os.path.join(DIR_PATH, 'rtings'), os.path.join(DIR_PATH, 'headphonecom'), os.path.join(DIR_PATH, 'referenceaudioanalyzer'), ] zip_files = set() for dir_path in dir_paths: for fp in glob(os.path.join(dir_path, '**', '* GraphicEQ.txt'), recursive=True): _, name = os.path.split(fp) name = name.replace(' GraphicEQ.txt', '') if re.search(MOD_REGEX, name, flags=re.IGNORECASE): # Skip samples, there are averaged results available continue manufacturer, _ = manufacturers.find(name) if manufacturer is None: print(f'Manufacturer could not be found for {name}') continue name = manufacturers.model(name) arcname = f'eq/{manufacturer}/{name}.txt' if arcname in zip_files: # Skip duplicates continue with open(fp, 'r', encoding='utf-8') as fh: s = fh.read() data = np.array( [x.split() for x in s.split(': ')[1].split('; ')], dtype='float') sl = np.logical_and(data[:, 0] > 100, data[:, 0] < 10000) data[:, 1] -= np.mean(data[sl, 1]) s = 'GraphicEQ: ' s += '; '.join([f'{x[0]:.0f} {x[1]:.1f}' for x in data]) zip_object.writestr(arcname, s) zip_files.add(arcname) zip_object.close()
def write_hesuvi_index(): os.makedirs(os.path.join(DIR_PATH, 'hesuvi'), exist_ok=True) manufacturers = ManufacturerIndex() zip_object = ZipFile(os.path.join(DIR_PATH, 'hesuvi.zip'), 'w') dir_paths = [ os.path.join(DIR_PATH, 'oratory1990'), os.path.join(DIR_PATH, 'crinacle', 'harman_in-ear_2019v2'), os.path.join(DIR_PATH, 'crinacle', 'crinacl_over-ear'), os.path.join(DIR_PATH, 'innerfidelity'), os.path.join(DIR_PATH, 'rtings'), os.path.join(DIR_PATH, 'headphonecom'), ] zip_files = set() for dir_path in dir_paths: for fp in glob(os.path.join(dir_path, '**', '* GraphicEQ.txt'), recursive=True): _, name = os.path.split(fp) name = name.replace(' GraphicEQ.txt', '') if re.search(r' \(?(sample |sn)[a-zA-Z0-9]+\)?$', name, flags=re.IGNORECASE): # Skip samples, there are averaged results available continue manufacturer, _ = manufacturers.find(name) name = manufacturers.model(name) arcname = f'eq/{manufacturer}/{name}.txt' if arcname in zip_files: # Skip duplicates continue with open(fp, 'r', encoding='utf-8') as fh: s = fh.read() data = np.array( [x.split() for x in s.split(': ')[1].split('; ')], dtype='float') sl = np.logical_and(data[:, 0] > 100, data[:, 0] < 10000) data[:, 1] -= np.mean(data[sl, 1]) s = 'GraphicEQ: ' s += '; '.join([f'{x[0]:.0f} {x[1]:.1f}' for x in data]) zip_object.writestr(arcname, s) zip_files.add(arcname) zip_object.close()