Good evening fellow hackers, working with sbase, we've decided a while ago to include libutf in the repo and base all tools' UTF-8-support on it instead of the insane wchar.h-interface. Now that almost all tools have UTF-8-support, we, along the way, implemented a number of utility functions which seemed necessary in some cases and I'm now proposing to add these functions back to libutf. The functions in question are:
1) chartorunearr(const char *str, Rune **r)[0] Read in str and convert it to a null-terminated Rune-array. The function takes care of proper memory allocation with exactly the right amount of space needed. This function was not added to the original libutf in Plan 9 because "We have no routines for manipulating null-terminated arrays of Runes. Although they should probably exist for completeness, we have found no need for them, for the same reason that %S and L"..." are rarely used."[1] Now, having worked on tr(1) myself, I needed a function like this to simplify the parser. In many cases you don't need a such a function, but tr(1) is an example where you do. 2) readrune(FILE *fp, Rune *r)[2] writerune(FILE *fp, Rune *r)[3] You may have noticed that the prototypes here are shorter than the ones in the given links. Of course, a library function would not call eprintf(). It would return -1 on error, 0 on EOF and > 0 the number of bytes read. readrune and writerune operate on single runes only and have proven to be very handy in many sbase tools. In case this proposal here is generally accepted, the import of the functions would happen just this way. What is left to discuss is where to put these functions. I would propose to place chartorunearr into rune.c, where already chartorune() and others reside. I would place readrune() and writerune() into a new file called runeio.c. The name doesn't really matter, because on the outside, only the utf.h- header will be exported. Let me know what you think. Cheers FRIGN [0]: http://git.2f30.org/sbase/tree/libutf/chartorunearr.c [1]: http://plan9.bell-labs.com/sys/doc/utf.html [2]: http://git.2f30.org/sbase/tree/libutf/readrune.c [3]: http://git.2f30.org/sbase/tree/libutf/writerune.c -- FRIGN <d...@frign.de>
