No longer maintained after SFOS 2.01. The SMS character counter has been integrated into the Messages app in SFOS 2.0.2, making this patch obsolete.

A patch to show the character count when typing an SMS.

The format is 'x (y)', where y is the number of messages and x is the number of characters left before a new message is started.

The patch applies the next time you start the Messages app.

Note to users

The SMS specification allows 3 types of character encoding, each resulting in a different maximum allowed number of characters per message: 7-bit (160 characters), 8-bit (140 characters) or 16-bit (70 characters).

Some characters (e.g. the tilde '~', pipe '|' or Euro sign '€') require an escape character and are therefore counted as two 7-bit characters. 8-bit characters are used in some special national implementations. Also, the shift table for 7-bit characters is locale specific. See this link for more info.

Known issues

This patch does not recognize the national variations and assumes only the basic shift table for 7-bit characters.

A practical implication may be that some characters are not recognized as belonging to the extended 7-bit set and will be treated as full Unicode characters - i.e., the counter will assume a limit of 70 characters per message rather than 140 or 160. At least it errs on the side of caution.

Character conversion

From version 0.0.3, this patch allows to convert multibyte characters in the message into their 7-bit approximations (e.g. 'ô' to 'o'). The counter changes the colour and becomes touch sensitive. Tap and hold on it and a context menu pops with with the convert option.

This feature is under test and may have some rough edges. For example, it does not convert Cyrillic characters that look similar but sound different (e.g. Cyrillic 'В' is Latin 'V', not 'B', and the patch does not know how to convert it). It also does not convert symbols like 'dz' or 'nj' that are single characters but would need a conversion to character pairs. I have tried my best but I am by no means an expert on character encoding and all my information comes from Wikipedia.

When the patch encounters a character it cannot convert, the conversion stops and the text cursors is set to that character, so the user can substitute it manually.

Please let me know if you find any problems or inconsistencies, including wrongly converted characters, and I will try to fix it in the next update.