I have a table with big text column on which I want to enforce uniqueness, postgres can only do this with a btree index. I get the following error :
CREATE unique INDEX idx_protein_seq_sequence ON protein_seq USING hash (sequence);  ERROR: index row requires 8736 bytes, maximum size is 8191
Are there solutions ?
I'm thinking of using a trigger (update and insert) and find duplicates with a hash index (a hash index seems to have no limit on size of text columns);