Regarding the bug fix for data reading on Lemi424 and the establishment of data reading function on Lemi417

[XuBingcheng6]

1. Fixed a bug in the lemi424.py file. The original code had an issue where the first day's data was repeatedly read when reading multiple days of lemi424 data from the same location (for example, when reading data from August 8th and 9th, 2025, it incorrectly read data from August 9th to August 10th, 2025. After checking the file size, it was found that the .h5 data file, which should have contained three days' worth of data, contained three days' worth of data).

2. Established lemi417 data reading functionality. The reading method has been strictly unified with that of lemi424.
   make_mth5_lemi417_example.zip

[kujaku11]

@XuBingcheng6 Thanks for the updates. Could you please update your code to work with mt-metadata > 1.0.0.

The key changes are the imports. MTime is now from mt_metadata.common import MTime.

[kujaku11]

I had to split this PR a bit because of changes to the architecture of mth5. We split the module mth5.io into a new repository mt-io. So changes to the reader of LEMI417, LEMI424, and LEMICollection are in mt-io.lemi while adjustments to mth5.clients are still in mth5. I'm going to leave this the PR open for now, but the changes are on the patches branch.

@XuBingcheng6 Do you have any example data from a LEM417 that you could share so we can make some tests?