Looks like
pd.read_clipboard(sep='s*|s*').iloc[1:,1:-1]
works pretty well. Output:
Record ID Shared On<br/>(UNIX timestamp) Share type Share To User
1 1 1611872850 shared user A
2 2 1611872851 shared user B
3 3 1611872852 shared user B
4 1 1611872853 share_removed user A
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…