Tamil Unicode Converter — Bamini & TSCII to Unicode

For roughly two decades before Unicode adoption, almost all digital Tamil was either typed in font-encoded legacy fonts — Bamini being the most widespread for Sri Lankan and Indian publishing — or stored in the 8-bit TSCII code page. In Bamini each Tamil glyph occupies a Latin/ASCII code position and text is stored in visual order (the order glyphs appear left-to-right on screen). On any device without that exact font, the same bytes render as Latin gibberish such as jkpo; instead of தமிழ். TSCII stores glyphs in bytes 0x80–0xFF and, when mis-decoded by a browser, appears as Windows-1252 symbols like ¾Á¢ú.

Tamil Unicode (the block U+0B80–U+0BFF) instead stores text in logical order — pronunciation/typing order — and lets the system shape it at render time. Because there is one standard code point per character, Unicode displays identically on every modern device with no special font. Conversion between the two is a deterministic, longest-match-first glyph-cluster substitution, not a guess:

1. Glyph-cluster substitution. Each legacy visual cluster — a left-side vowel sign, its consonant, and any trailing glyph — maps directly to the correct logically-ordered Unicode sequence. Longest-match-first scanning means a three-glyph cluster such as nfh → கொ is matched before the shorter nf → கெ or single f → க.
2. Visual-to-logical ordering is baked in. The left-side vowel signs ெ ே ை are drawn before their consonant in legacy fonts but must follow it in Unicode, and the two-part signs ொ ோ ௌ wrap around the consonant. Because the tables key whole clusters, the reorder and composition happen automatically — legacy visual ெ + க + ா becomes the single Unicode sign in கொ (க U+0B95 + ொ U+0BCA).

The mapping points come from the Unicode Standard Tamil chart, the TSCII 1.7 code page, and the Ezhil Language Foundation open-tamil converter tables. Every mapping in this build is checked to round-trip losslessly— converting Unicode → legacy → Unicode returns the original text — which is how the tool earns its “round-trip verified” badge. The engine deliberately uses positional longest-match-first scanning rather than the upstream library's sequential replace, because sequential replacement mis-orders a few au-clusters (a one-character rule can fire before a three-character cluster); positional scanning is order-independent and correct. Glyphs outside the verified core sets are never guessed: they pass through unchanged and are counted for you as unmapped, so a partial conversion is always visible.

• The Unicode Standard — Tamil block U+0B80–U+0BFF (code chart)
• TSCII 1.7 specification — INFITT / Tamil Virtual University
• Ezhil Language Foundation — open-tamil (bamini2utf8 / tscii2utf8 reference tables)
• Unicode Technical Note 21 — Tamil script and ordering notes

The mapping tables on this page were last cross-checked against the Unicode Tamil chart and the cited references on 2026-07-16. Coverage currently spans the Bamini and TSCII core sets; other legacy systems (STMozhi, TAB/TAM, Shreelipi-style fonts) each need their own verified table and are added only once confirmed, because a wrong mapping produces silently-incorrect text.