Just for completeness' sake, in Unicode characters like e-acute can appear either composed (your \303\251) or decomposed (\145 \314 \201). This is actually two characters, LATIN SMALL LETTER E (the \145) and COMBINING ACUTE ACCENT (the \314 \201). If you run into this, Unicode::Normalize is your friend -- or will be once you read the file '<:encoding(utf-8)'.