sasquatch does not switch to the correct endianness

[qkaiser]

Sasquatch does not switch to the correct endianness and needs it to be provided explicitly on some samples.

To Reproduce

Steps to reproduce the behavior:

1. Download this sample from Netgear
2. Launch unblob with command unblob -v -e /tmp/out -f -k WNR3500Lv2-V1.2.0.40_40.0.80.zip
3. Go to extraction directory
4. See that the squashfs filesystem was not extracted

sasquatch -no-exit-code -f -d /tmp/out/WNR3500Lv2-V1.2.0.40_40.0.80.zip_extract/WNR3500Lv2-V1.2.0.40_40.0.80.chk_extract/
kernel_extract/part1_extract/0-6053888.squashfs_v3_broadcom_extract /tmp/out/WNR3500Lv2-V1.2.0.40_40.0.80.zip_extract
/WNR3500Lv2-V1.2.0.40_40.0.80.chk_extract/kernel_extract/part1_extract/0-6053888.squashfs_v3_broadcom

SquashFS version [3.1] / inode count [1024] suggests a SquashFS image of a different endianess

Non-standard SquashFS Magic: 'shsq'

Reading a different endian SQUASHFS filesystem on /tmp/out/WNR3500Lv2-V1.2.0.40_40.0.80.zip_extract/
WNR3500Lv2-V1.2.0.40_40.0.80.chk_extract/kernel_extract/part1_extract/0-6053888.squashfs_v3_broadcom

FATAL ERROR: Block size or block_log too large.  File system is corrupt.

Expected behavior

Unblob should detect the endianness and provide it explicitly to sasquatch (either -le or -be).

Additional context

Reported by @m-1-k-3 from EMBA and validated.