Tamil Font Converter

Before Unicode was supported everywhere, computers had no agreed numbers for Tamil letters. Font makers worked around that by drawing Tamil shapes in the slots of English letters and symbols. In the Bamini font the key j draws த, k draws ம and ; draws the dot (புள்ளி). So the word தமிழ் is stored as jkpo;.

That works while everyone has the same font. Open the file on a phone, paste it into an e-mail or search it on the web, and the font is gone. What is left is the stored letters: jkpo;. TSCII, the other widely used encoding, stores Tamil in the upper half of the 8-bit range, so the same word comes out as accented Latin, ¾Á¢ú.

These fonts also store letters in the order they are drawn, not the order they are read. In கொ the sign ெ is drawn first, then க, then ா, so Bamini stores nfh. A converter has to put the pieces back in reading order, which is why simple find-and-replace tools often get vowel signs wrong.

In February 1999 the TamilNet99 conference in Chennai tried to end the font chaos, where every Tamil font used its own private layout. On 13 June 1999 the Government of Tamil Nadu adopted its recommendations in G.O. Ms. No. 17: the Tamil99 keyboard, and two glyph encodings for fonts. TAM, the monolingual scheme, was meant for desktop publishing and printing. TAB, the bilingual scheme, was meant for everyday text and e-mail. A second order, G.O. Ms. No. 28 of 9 August 1999, named Kani Tamil Sangam to certify software, and government offices were to buy only software that followed the standard.

Both are glyph encodings, like Bamini: they store the shapes in the order they are drawn, so கொ is stored as ெ, க, ா. TAB keeps English in bytes 0 to 127, which is why the same file can mix English and Tamil. TAM gives up the English letters and fills them with ready-made shapes such as கி and ம், so English typed in a TAM font shows as Tamil. The copy of the order on tamilvu.org still prints its Tamil in TAB code: îñ¤ö¢99→தமிழ்99.

Unicode Tamil replaced both, but TAB and TAM text still turns up in older government and office documents and in print files. The converter reads and writes both. The Detect setting, here and on the Tamil Unicode converter, tells them apart from TSCII by checking which reading gives real Tamil spelling; very short snippets of one or two letters can be ambiguous, so check the From font if the result looks wrong. Word documents in fonts named TAB-… or TAM-… are recognised by the font name.

1. Paste the text into the left box, or press Open file (under More on a phone) to load a Word .docx or a plain-text file. A .docx opens in its own view, where you pick the fonts to convert and download the converted document.
2. Set From to the font the text was typed in. If you don't know, choose Detect and the converter will guess from the characters.
3. Set To. Pick Unicode for anything new. Pick a legacy font such as Bamini, TSCII, TAB or TAM only when a program needs it.
4. Press Copy, or download a .txt file. Unicode downloads are UTF-8, and legacy downloads keep one byte per character so older programs open them correctly.

For a page-by-page walkthrough, see Bamini to Unicode for fixing old documents, Unicode to Bamini for print and DTP work, and the Tamil Unicode converter when you have text and no idea which font it came from.