The PyOpenEXR Python bindings return stale heap data when reading a specially crafted deep scanline EXR file that utilizes layer-prefixed RGB channels. When using default channel coalescing (where `separate channels` is set to `False`), the wrapper groups channels like left.R, left.G, and left.B into a single RGB sample array. However, the lane-offset calculation in the `setDeepSliceData()` function only recognizes the exact unprefixed names G, B, and A. Consequently, prefixed channels such as left.G and left.B are decoded into lane 0, leaving lanes 1 and 2 uninitialized. A Python application using the default OpenEXR.File API to read untrusted deep EXR files may expose uninitialized same-process heap contents when processing the resulting NumPy sample arrays, while also receiving incorrect green and blue channel data.