Error: File must be in the .h5ad format.

[royfrancis]

I got this error when loading the h5ad file into cellxgene:

[cellxgene] Starting the CLI...
[cellxgene] Loading data from anndata.h5ad, this may take a while...
Error: File must be in the .h5ad format. Please read https://github.com/theislab/scanpy_usage/blob/master/170505_seurat/info_h5ad.md to learn more about this format. You may be able to convert your file into this format using `cellxgene prepare`, please run `cellxgene prepare --help` for more information.

Turns out it was caused by a character φ like Interstitial Mφ perivascular in one of metadata cell label columns. After removing that, it worked.

v1.3.0

[fresh3nough]

Opened a fix for this in #2762. The root cause was the broad except ValueError handler in _load_data() that caught Unicode-related errors from anndata/h5py and reported them as a format issue. The fix adds Unicode-aware error detection with a meaningful message, appends the original error to the generic format message for other ValueError cases, and normalizes all string metadata to NFC form during validation to prevent encoding issues downstream.