ids_lv0.txt: All differences in strokes are differentiated.
ids_lv1.txt: Stroke-level differences (e.g. 丶 and 乀 in certain position) are merged.
ids_lv2.txt: Part of variants defined in UCV list, which considered as "must unifiable" are merged.
Each file contains 3 columns: first column is character, second and third column each contains a semi-colon seperated list of IDS sequences. Each sequence is followed by some comma-seperated indicators within parentheses, indicating variant of the sequence. The sequences in second column and third column are considered as "primary IDS" and "alternative IDSes" respectively.
Author: Yi Bai