sparkocr.utils.svs.phi_redaction#

Functions#

calculate_actual_level_height(level_info)#

calculate_actual_level_width(level_info)#

calculate_max_group_mem_gb(mem_per_worker_gb=1.5, ram_fraction=0.75, overhead_mem_gb=2.0)#

chunk_list(lst, n)#

dataframe_to_detections(detections_df)#

detect_tile_level(folder_path)#

detections_to_dataframe(detections_dict, dir_path, svs_name, level)#

df2list(tiles_deid)#

Convert a list of dictionaries (with stringified ‘coordinates’) into a standard format. Returns a list of dicts with keys ‘file_path’ and ‘bbox’ (list of (x, y, w, h) tuples).

estimate_svs_memory_gb(svs_path)#

extract_base_name(tile_path)#

Extract base name from a tile file path, supporting both PNG and JPG (case-insensitive). Extracts everything before ‘_L<digits>_T<digits>.<ext>’

extract_svs_name_from_tiles(files_path)#

find_intersecting_tiles(x1, y1, x2, y2, level_info)#

find_thumbnail_index(tif)#

Find the thumbnail page index using Aperio SVS conventions.

get_existing_extension(dir_path, svs_name, level, tile_num)#

get_svs_info_from_images(files_path)#

get_tile_position_and_size(tile_index, level_info)#

group_svs_files_by_memory(svs_files, max_group_mem_gb=6.0)#

prepare_thumbnail_deid(only_thumbnail_deid, main_folder)#

process_file(svs_path, tiles_deid, thumbnail_deid, output_svs_path=None, create_new_svs_file=False)#

redact_phi_in_tiles(svs_path, tiles_deid, main_folder, output_svs_path=None, create_new_svs_file=False)#

Modifies tiles in an SVS file to remove sensitive information by drawing black rectangles. Can operate on a single SVS file or a directory containing multiple SVS files. The function will first scale the detections to all pyramid levels before applying deidentification. Parameters: - svs_path (str): Path to a single SVS file or a directory containing SVS files. - tiles_deid (list or DataFrame): List of dictionaries or a PySpark DataFrame containing the de-identification instructions. - main_folder (str): Path to folder containing tile directories. - output_svs_path (str, optional): Path where the new de-identified SVS file(s) will be saved. - create_new_svs_file (bool): If True, creates new SVS files instead of modifying originals. - max_group_mem_gb (float): Max memory per group for batching files.

scale_boxes_to_all_levels(detections_df, main_folder)#

setup_logging()#