diff --git a/qim3d/io/convert.py b/qim3d/io/convert.py
index 24c41e8c72478e513956321f92432e32c9d54645..dd29d7e2d25ea8be5a0d096c1b3841e8b6159502 100644
--- a/qim3d/io/convert.py
+++ b/qim3d/io/convert.py
@@ -1,3 +1,5 @@
+import difflib
+import os
from itertools import product
import numpy as np
@@ -5,56 +7,113 @@ import tifffile as tiff
import zarr
from tqdm import tqdm
+from qim3d.utils.internal_tools import stringify_path
-def convert_tif_to_zarr(tif_path, zarr_path, chunks=(64, 64, 64)):
- """ Convert a tiff file to a zarr file
- Args:
- tif_path (str): path to the tiff file
- zarr_path (str): path to the zarr file
- chunks (tuple, optional): chunk size for the zarr file. Defaults to (64, 64, 64).
+class Convert:
- Returns:
- zarr.core.Array: zarr array containing the data from the tiff file
- """
- vol = tiff.memmap(tif_path)
- z = zarr.open(zarr_path, mode='w', shape=vol.shape, chunks=chunks, dtype=vol.dtype)
- chunk_shape = tuple((s + c - 1) // c for s, c in zip(z.shape, z.chunks))
- for chunk_indices in tqdm(product(*[range(n) for n in chunk_shape]), total=np.prod(chunk_shape)):
- slices = tuple(slice(c * i, min(c * (i + 1), s))
- for s, c, i in zip(z.shape, z.chunks, chunk_indices))
- temp_data = vol[slices]
- # The assignment takes 99% of the cpu-time
- z.blocks[chunk_indices] = temp_data
+ def __init__(self,**kwargs):
+ """ Utility class to convert files to other formats without loading the entire file into memory
- return z
+ Args:
+ chunk_shape (tuple, optional): chunk size for the zarr file. Defaults to (64, 64, 64).
+ """
+ self.chunk_shape = kwargs.get("chunk_shape", (64, 64, 64))
-def convert_npy_to_zarr(npy_path, zarr_path, shape, dtype=np.float32, chunks=(64, 64, 64)):
- """ Convert a numpy file to a zarr file
+ def convert(self, input_path, output_path):
+ # Stringify path in case it is not already a string
+ input_path = stringify_path(input_path)
- Args:
- npy_path (str): path to the numpy file
- zarr_path (str): path to the zarr file
- chunks (tuple, optional): chunk size for the zarr file. Defaults to (64, 64, 64).
+ if os.path.isfile(input_path) and os.path.isfile(output_path):
+ match input_path, output_path:
+ case (".tif", ".zarr"):
+ return self.convert_tif_to_zarr(input_path, output_path)
+ case (".npy", ".zarr"):
+ return self.convert_npy_to_zarr(input_path, output_path, shape=(64, 64, 64))
+ case (".zarr", ".tif"):
+ return self.convert_zarr_to_tif(input_path, output_path)
+ case _:
+ raise ValueError("Unsupported file format")
+ # Load a directory
+ elif os.path.isdir(input_path):
+ raise ValueError("Unsupported file format")
+ # Fail
+ else:
+ # Find the closest matching path to warn the user
+ parent_dir = os.path.dirname(input_path) or "."
+ parent_files = os.listdir(parent_dir) if os.path.isdir(parent_dir) else ""
+ valid_paths = [os.path.join(parent_dir, file) for file in parent_files]
+ similar_paths = difflib.get_close_matches(input_path, valid_paths)
+ if similar_paths:
+ suggestion = similar_paths[0] # Get the closest match
+ message = f"Invalid path. Did you mean '{suggestion}'?"
+ raise ValueError(repr(message))
+ else:
+ raise ValueError("Invalid path")
- Returns:
- zarr.core.Array: zarr array containing the data from the numpy file
- """
- vol = np.memmap(npy_path, dtype=dtype, mode='r', shape=shape)
- z = zarr.open(zarr_path, mode='w', shape=vol.shape, chunks=chunks, dtype=vol.dtype)
- z[:] = vol[:]
+ def convert_tif_to_zarr(self, tif_path, zarr_path, chunks=(64, 64, 64)):
+ """ Convert a tiff file to a zarr file
- return z
+ Args:
+ tif_path (str): path to the tiff file
+ zarr_path (str): path to the zarr file
+ chunks (tuple, optional): chunk size for the zarr file. Defaults to (64, 64, 64).
-def convert_zarr_to_tif(zarr_path, tif_path):
- """ Convert a zarr file to a tiff file
+ Returns:
+ zarr.core.Array: zarr array containing the data from the tiff file
+ """
+ vol = tiff.memmap(tif_path)
+ z = zarr.open(zarr_path, mode='w', shape=vol.shape, chunks=chunks, dtype=vol.dtype)
+ chunk_shape = tuple((s + c - 1) // c for s, c in zip(z.shape, z.chunks))
+ # ! Fastest way is z[:] = vol[:], but does not have a progress bar
+ for chunk_indices in tqdm(product(*[range(n) for n in chunk_shape]), total=np.prod(chunk_shape)):
+ slices = tuple(slice(c * i, min(c * (i + 1), s))
+ for s, c, i in zip(z.shape, z.chunks, chunk_indices))
+ temp_data = vol[slices]
+ # The assignment takes 99% of the cpu-time
+ z.blocks[chunk_indices] = temp_data
- Args:
- zarr_path (str): path to the zarr file
- tif_path (str): path to the tiff file
+ return z
+
+ def convert_npy_to_zarr(self, npy_path, zarr_path, shape, dtype=np.float32, chunks=(64, 64, 64)):
+ """ Convert a numpy file to a zarr file
+
+ Args:
+ npy_path (str): path to the numpy file
+ zarr_path (str): path to the zarr file
+ chunks (tuple, optional): chunk size for the zarr file. Defaults to (64, 64, 64).
+
+ Returns:
+ zarr.core.Array: zarr array containing the data from the numpy file
+ """
+ vol = np.memmap(npy_path, dtype=dtype, mode='r', shape=shape)
+ z = zarr.open(zarr_path, mode='w', shape=vol.shape, chunks=chunks, dtype=vol.dtype)
+ z[:] = vol[:]
- returns:
- None
+ return z
+
+ def convert_zarr_to_tif(self, zarr_path, tif_path):
+ """ Convert a zarr file to a tiff file
+
+ Args:
+ zarr_path (str): path to the zarr file
+ tif_path (str): path to the tiff file
+
+ returns:
+ None
+ """
+ z = zarr.open(zarr_path)
+ tiff.imwrite(tif_path, z)
+
+
+def convert(input_path: str, output_path: str, chunk_shape: tuple = (64, 64, 64)):
+ """ Convert a file to another format without loading the entire file into memory
+
+ Args:
+ input_path (str): path to the input file
+ output_path (str): path to the output file
+ chunk_shape (tuple, optional): chunk size for the zarr file. Defaults to (64, 64, 64).
"""
- z = zarr.open(zarr_path)
- tiff.imwrite(tif_path, z)
\ No newline at end of file
+
+ converter = Convert(chunk_shape=chunk_shape)
+ converter.convert(input_path, output_path)