127 is 'delete' -- ascii all right, but not 'printable'.

cheers Miller

On Tue, Jan 19, 2010 at 09:37:08PM -0500, Hans-Christoph Steiner wrote:

Looks good to me. One comment, shouldn't this be n<128? 127 is an ASCII char, AFAIK.

•    if (n == '\n' || (n > 31 && n < 127))
  
  

It looks worth checking to me, hopefully we can get Miller and others to weigh in on it.

.hc

On Jan 19, 2010, at 4:16 PM, Bryan Jurish wrote:

...

morning all,

attached is a UTF-8 support patch against branches/pd-gui-rewrite/0.43 revision 13051 (HEAD as of an hour or so ago). most of the bulk is new files (s_utf8.c, s_utf8.h), most other changes are in g_rtext.c. It's not too monstrous, and I've tested it again here briefly with some utf-8 test patches (see other attachment), and things appear to be working as expected. if desired, I can check this in; otherwise feel free to do it for me ;-)

2 annoying things here during testing (I don't see how my patches could have caused this, but you never know):

(1) all loaded patch windows appear at +0+0 (upper left corner), which with my wm (windowmaker) means the title bar is off the screen, and I have to resort to keyboard shortcuts to get them mouse-draggable, which is a major pain in the wazoo: is this a known bug?

(2) I can't figure out how to get at the properties dialog for number, number2, or any other gui-atom objects: should these be working already?

marmosets, Bryan

On 2010-01-18 23:09:34, Hans-Christoph Steiner hans@eds.org appears to have written:

...

Awesome! If its big and complicated, I say post it to the list first, if not too bad, then just commit.

.hc

On Jan 18, 2010, at 4:47 AM, Bryan Jurish wrote:

...

moin Hans, moin list,

I think perhaps I never actually did post the cleaned-up patch anywhere (bad programmer, no biscuit); I guess I'll check out branches/pd-gui-rewrite/0.43 and try patching my changes in; then I can either commit or just post the (updated) patch. Hopefully no major additional changes will be required, so it ought to go pretty fast.

marmosets, Bryan

On 2010-01-17 22:57:33, Hans-Christoph Steiner hans@eds.org appears to have written:

...

Hey Bryan,

I'd like to try to get your UTF-8 code into pd-gui-rewrite. You mention in this posting back in May that you had the whole thing working. I couldn't find the diff/patch for this. Is it posted anywhere? Do you want to try to check it in yourself directly to the pd-gui- rewrite/0.43 branch?

.hc

On Mar 20, 2009, at 6:16 PM, Bryan Jurish wrote:

...

morning all,

Of course I never really like to see my code wither away in the bit bucket, but I personally don't have any pressing need for UTF-8 symbols, comments, etc. in Pd -- I'm a native English speaker, after all ;-)

Also, my changes are by no means the only way to do it (or even the best way); we could gain a little speed by slapping on some more buffers (mostly and possibly only in rtext_senditup()), but since this seems to effect only GUI/editing stuff, I think we can live with a smidgeon of additional cpu time ... after all, it's all O(n) anyways.

Really I just wanted to see how easy (or difficult) it would be to get Pd to use UTF-8 as its internal encoding... turned out to be harder than I had thought, but (ever so slightly) easier than I had feared :-/

marmosets, Bryan

On 2009-03-20 18:39:06, Hans-Christoph Steiner hans@eds.org appears to have written: > >I wonder what the best approach is to getting it included. I >also >think >its a very valuable contribution. I think we need to first get >the >Tcl/Tk only changes done, since that was the mandate of the pd- >devel >0.41 effort. Then once Miller has accepted those changes, then >we can >start with the C modifications there. So how to proceed next, >I think >is based on how eager you are, Bryan, to getting this in a >regular >build. > >One option is making a pd-devel-utf8 branch, another is posting >these >patches to the patch tracker and waiting for Miller to make his >next >update with the Pd-devel Tcl-Tk code. > >Maybe we can get Miller to chime in on this topic. > >.hc > >On Mar 13, 2009, at 12:00 AM, dmotd wrote: > >>hey bryan, >> >>just a quick note of a appreciation for getting this one out.. >>i hope >>it gets >>picked up in millers build soon.. a very useful and necessary >>modification. >> >>well done! >> >>dmotd >> >>On Thursday 12 March 2009 08:07:50 Bryan Jurish wrote: >>>moin folks, >>> >>>I believe I've finally got pd-devel 0.41-4 using UTF-8 across >>>the >>>board. >>>So far, I've tested message boxes & comments (g_rtext), as >>>well as >>>symbol atoms, and all seems good. I think we can still expect >>>goofiness >>>if someone names an abstraction using a multibyte character >>>when the >>>filesystem isn't UTF-8 encoded (raw 8-bit works for me here >>>too), >>>but I >>>really don't want to open that particular can of worms. >>> >>>So I guess I have 2 questions: >>> >>>(1) what should I call the generic UTF-8 source files? (see >>>my other >>>post) >>> >>>(2) shall I commit these changes to pd-devel/0.41-4, or >>>somewhere >>>else, >>>or just post a diff (ca. 33k, ought to be easier to read now; >>>I've >>>tried >>>to follow the indentation conventions of the source files I >>>modified)? >>> >>>marmosets, >>> Bryan

-- Bryan Jurish "There is *always* one more bug." jurish@ling.uni-potsdam.de -Lubarsky's Law of Cybernetic Entomology

---

The arc of history bends towards justice. - Dr. Martin Luther King, Jr.

--

---

Bryan Jurish Deutsches Textarchiv Berlin-Brandenburgische Akademie der Wissenschaften

J?gerstr. 22/23 10117 Berlin

Tel.: +49 (0)30 20370 539 E-Mail: jurish@bbaw.de

---

---

As we enjoy great advantages from inventions of others, we should be glad of an opportunity to serve others by any invention of ours; and this we should do freely and generously. - Benjamin Franklin

-- Bryan Jurish "There is *always* one more bug." jurish@uni-potsdam.de -Lubarsky's Law of Cybernetic Entomology Index: src/Makefile.am =================================================================== --- src/Makefile.am (revision 13051) +++ src/Makefile.am (working copy) @@ -24,6 +24,7 @@ m_conf.c m_glob.c m_sched.c \ s_main.c s_inter.c s_file.c s_print.c \ s_loader.c s_path.c s_entry.c s_audio.c s_midi.c \

• s_utf8.c \ d_ugen.c d_ctl.c d_arithmetic.c d_osc.c d_filter.c d_dac.c

d_misc.c \ d_math.c d_fft.c d_array.c d_global.c \ d_delay.c d_resample.c \ Index: src/g_editor.c =================================================================== --- src/g_editor.c (revision 13051) +++ src/g_editor.c (working copy) @@ -9,6 +9,7 @@ #include "s_stuff.h" #include "g_canvas.h" #include <string.h> +#include "s_utf8.h" /*-- moo --*/

void glist_readfrombinbuf(t_glist *x, t_binbuf *b, char *filename, int selectem); @@ -1666,8 +1667,9 @@ gotkeysym = av[1].a_w.w_symbol; else if (av[1].a_type == A_FLOAT) {

•    char buf[3];
  

•    sprintf(buf, "%c", (int)(av[1].a_w.w_float));
  

•    /*-- moo: assume keynum is a Unicode codepoint; encode as  
  

UTF-8 --*/

•    char buf[UTF8_MAXBYTES1];
  

•    u8_wc_toutf8_nul(buf, (UCS4)(av[1].a_w.w_float));
   gotkeysym = gensym(buf);
  

  } else gotkeysym = gensym("?");

Index: src/s_utf8.c

--- src/s_utf8.c (revision 0) +++ src/s_utf8.c (revision 0) @@ -0,0 +1,280 @@ +/*

• Basic UTF-8 manipulation routines
• by Jeff Bezanson
• placed in the public domain Fall 2005
• This code is designed to provide the utilities you need to

manipulate

• UTF-8 as an internal string encoding. These functions do not

perform the

• error checking normally needed when handling UTF-8 data, so if

you happen

• to be from the Unicode Consortium you will want to flay me alive.
• I do this because error checking can be performed at the

boundaries (I/O),

• with these routines reserved for higher performance on data known

to be

• valid.
• modified by Bryan Jurish (moo) March 2009
• • removed some unneeded functions (escapes, printf etc), added

others +*/ +#include <stdlib.h> +#include <stdio.h> +#include <string.h> +#include <stdarg.h> +#ifdef WIN32 +#include <malloc.h> +#else +#include <alloca.h> +#endif

+#include "s_utf8.h"

+static const u_int32_t offsetsFromUTF8[6] = {

• 0x00000000UL, 0x00003080UL, 0x000E2080UL,
• 0x03C82080UL, 0xFA082080UL, 0x82082080UL

+};

+static const char trailingBytesForUTF8[256] = {

• 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
• 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
• 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
• 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
• 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
• 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
• 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
• 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 3,3,3,3,3,3,3,3,4,4,4,4,5,5,5,5

+};

+/* returns length of next utf-8 sequence */ +int u8_seqlen(char *s) +{

• return trailingBytesForUTF8[(unsigned int)(unsigned char)s[0]]
• 1;

+}

+/* conversions without error checking

• only works for valid UTF-8, i.e. no 5- or 6-byte sequences
• srcsz = source size in bytes, or -1 if 0-terminated
• sz = dest size in # of wide characters
• returns # characters converted
• dest will always be L'\0'-terminated, even if there isn't enough

room

• for all the characters.
• if sz = srcsz+1 (i.e. 4*srcsz+4 bytes), there will always be

enough space. +*/ +int u8_toucs(u_int32_t *dest, int sz, char *src, int srcsz) +{

• u_int32_t ch;
• char *src_end = src + srcsz;
• int nb;
• int i=0;
• while (i < sz-1) {

•    nb = trailingBytesForUTF8[(unsigned char)*src];
  

•    if (srcsz == -1) {
  

•        if (*src == 0)
  

•            goto done_toucs;
  

•    }
  

•    else {
  

•        if (src + nb >= src_end)
  

•            goto done_toucs;
  

•    }
  

•    ch = 0;
  

•    switch (nb) {
  

•        /* these fall through deliberately */
  

+#if UTF8_SUPPORT_FULL_UCS4

•    case 5: ch += (unsigned char)*src++; ch <<= 6;
  

•    case 4: ch += (unsigned char)*src++; ch <<= 6;
  

+#endif

•    case 3: ch += (unsigned char)*src++; ch <<= 6;
  

•    case 2: ch += (unsigned char)*src++; ch <<= 6;
  

•    case 1: ch += (unsigned char)*src++; ch <<= 6;
  

•    case 0: ch += (unsigned char)*src++;
  

•    }
  

•    ch -= offsetsFromUTF8[nb];
  

•    dest[i++] = ch;
  

• }
• done_toucs:
• dest[i] = 0;
• return i;

+}

+/* srcsz = number of source characters, or -1 if 0-terminated

• sz = size of dest buffer in bytes
• returns # characters converted
• dest will only be '\0'-terminated if there is enough space. this

is

• for consistency; imagine there are 2 bytes of space left, but

the next

• character requires 3 bytes. in this case we could NUL-terminate,

but in

• general we can't when there's insufficient space. therefore this

function

• only NUL-terminates if all the characters fit, and there's space

for

• the NUL as well.
• the destination string will never be bigger than the source

string. +*/ +int u8_toutf8(char *dest, int sz, u_int32_t *src, int srcsz) +{

• u_int32_t ch;
• int i = 0;
• char *dest_end = dest + sz;
• while (srcsz<0 ? src[i]!=0 : i < srcsz) {

•    ch = src[i];
  

•    if (ch < 0x80) {
  

•        if (dest >= dest_end)
  

•            return i;
  

•        *dest++ = (char)ch;
  

•    }
  

•    else if (ch < 0x800) {
  

•        if (dest >= dest_end-1)
  

•            return i;
  

•        *dest++ = (ch>>6) | 0xC0;
  

•        *dest++ = (ch & 0x3F) | 0x80;
  

•    }
  

•    else if (ch < 0x10000) {
  

•        if (dest >= dest_end-2)
  

•            return i;
  

•        *dest++ = (ch>>12) | 0xE0;
  

•        *dest++ = ((ch>>6) & 0x3F) | 0x80;
  

•        *dest++ = (ch & 0x3F) | 0x80;
  

•    }
  

•    else if (ch < 0x110000) {
  

•        if (dest >= dest_end-3)
  

•            return i;
  

•        *dest++ = (ch>>18) | 0xF0;
  

•        *dest++ = ((ch>>12) & 0x3F) | 0x80;
  

•        *dest++ = ((ch>>6) & 0x3F) | 0x80;
  

•        *dest++ = (ch & 0x3F) | 0x80;
  

•    }
  

•    i++;
  

• }
• if (dest < dest_end)

•    *dest = '\0';
  

• return i;

+}

+/* moo: get byte length of character number, or 0 if not supported */ +int u8_wc_nbytes(u_int32_t ch) +{

• if (ch < 0x80) return 1;
• if (ch < 0x800) return 2;
• if (ch < 0x10000) return 3;
• if (ch < 0x200000) return 4;

+#if UTF8_SUPPORT_FULL_UCS4

• /*-- moo: support full UCS-4 range? --*/
• if (ch < 0x4000000) return 5;
• if (ch < 0x7fffffffUL) return 6;

+#endif

• return 0; /*-- bad input --*/

+}

+int u8_wc_toutf8(char *dest, u_int32_t ch) +{

• if (ch < 0x80) {

•    dest[0] = (char)ch;
  

•    return 1;
  

• }
• if (ch < 0x800) {

•    dest[0] = (ch>>6) | 0xC0;
  

•    dest[1] = (ch & 0x3F) | 0x80;
  

•    return 2;
  

• }
• if (ch < 0x10000) {

•    dest[0] = (ch>>12) | 0xE0;
  

•    dest[1] = ((ch>>6) & 0x3F) | 0x80;
  

•    dest[2] = (ch & 0x3F) | 0x80;
  

•    return 3;
  

• }
• if (ch < 0x110000) {

•    dest[0] = (ch>>18) | 0xF0;
  

•    dest[1] = ((ch>>12) & 0x3F) | 0x80;
  

•    dest[2] = ((ch>>6) & 0x3F) | 0x80;
  

•    dest[3] = (ch & 0x3F) | 0x80;
  

•    return 4;
  

• }
• return 0;

+}

+/*-- moo --*/ +int u8_wc_toutf8_nul(char *dest, u_int32_t ch) +{

• int sz = u8_wc_toutf8(dest,ch);
• dest[sz] = '\0';
• return sz;

+}

+/* charnum => byte offset */ +int u8_offset(char *str, int charnum) +{

• int offs=0;
• while (charnum > 0 && str[offs]) {

•    (void)(isutf(str[++offs]) || isutf(str[++offs]) ||
  

•           isutf(str[++offs]) || ++offs);
  

•    charnum--;
  

• }
• return offs;

+}

+/* byte offset => charnum */ +int u8_charnum(char *s, int offset) +{

• int charnum = 0, offs=0;
• while (offs < offset && s[offs]) {

•    (void)(isutf(s[++offs]) || isutf(s[++offs]) ||
  

•           isutf(s[++offs]) || ++offs);
  

•    charnum++;
  

• }
• return charnum;

+}

+/* reads the next utf-8 sequence out of a string, updating an index */ +u_int32_t u8_nextchar(char *s, int *i) +{

• u_int32_t ch = 0;
• int sz = 0;
• do {

•    ch <<= 6;
  

•    ch += (unsigned char)s[(*i)++];
  

•    sz++;
  

• } while (s[*i] && !isutf(s[*i]));
• ch -= offsetsFromUTF8[sz-1];
• return ch;

+}

+/* number of characters */ +int u8_strlen(char *s) +{

• int count = 0;
• int i = 0;
• while (u8_nextchar(s, &i) != 0)

•    count++;
  

• return count;

+}

+void u8_inc(char *s, int *i) +{

• (void)(isutf(s[++(*i)]) || isutf(s[++(*i)]) ||

•       isutf(s[++(*i)]) || ++(*i));
  

+}

+void u8_dec(char *s, int *i) +{

• (void)(isutf(s[--(*i)]) || isutf(s[--(*i)]) ||

•       isutf(s[--(*i)]) || --(*i));
  

+}

+/*-- moo --*/ +void u8_inc_ptr(char **sp) +{

• (void)(isutf(*(++(*sp))) || isutf(*(++(*sp))) ||

• isutf(*(++(*sp))) || ++(*sp));
  

+}

+/*-- moo --*/ +void u8_dec_ptr(char **sp) +{

• (void)(isutf(*(--(*sp))) || isutf(*(--(*sp))) ||

• isutf(*(--(*sp))) || --(*sp));
  

+} Index: src/g_rtext.c =================================================================== --- src/g_rtext.c (revision 13051) +++ src/g_rtext.c (working copy) @@ -13,6 +13,7 @@ #include "m_pd.h" #include "s_stuff.h" #include "g_canvas.h" +#include "s_utf8.h"

#define LMARGIN 2 @@ -32,10 +33,10 @@

struct _rtext {

• char *x_buf;
• int x_bufsize;
• int x_selstart;
• int x_selend;

• char *x_buf; /*-- raw byte string, assumed UTF-8 encoded

(moo) --*/

• int x_bufsize; /*-- byte length --*/
• int x_selstart; /*-- byte offset --*/
• int x_selend; /*-- byte offset --*/ int x_active; int x_dragfrom; int x_height;

@@ -119,6 +120,15 @@

/* LATER deal with tcl-significant characters */

+/* firstone(), lastone()

• • • returns byte offset of (first|last) occurrence of 'c' in

's[0..n-1]', or

• • -1 if none was found
• • • 's' is a raw byte string
• • • 'c' is a byte value
• • • 'n' is the length (in bytes) of the prefix of 's' to be

searched.

• • • we could make these functions work on logical characters in

utf8 strings,

• • but we don't really need to...
• */

static int firstone(char *s, int c, int n) { char *s2 = s + n; @@ -155,6 +165,16 @@ of the entire text in pixels. */

• /*-- moo:
• • • some variables from the original version have been renamed
• • • variables with a "_b" suffix are raw byte strings, lengths,

or offsets

• • • variables with a "_c" suffix are logical character lengths

or offsets

• • (assuming valid UTF-8 encoded byte string in x->x_buf)
• • • a fair amount of O(n) computations required to convert

between raw byte

• • offsets (needed by the C side) and logical character

offsets (needed by

• • the GUI)
• */
• /* LATER get this and sys_vgui to work together properly, breaking up messages as needed. As of now, there's a limit of 1950 characters, imposed by sys_vgui(). */

@@ -171,14 +191,16 @@ { t_float dispx, dispy; char smallbuf[200], *tempbuf;

• int outchars = 0, nlines = 0, ncolumns = 0,

• int outchars_b = 0, nlines = 0, ncolumns = 0, pixwide, pixhigh, font, fontwidth, fontheight, findx, findy; int reportedindex = 0; t_canvas *canvas = glist_getcanvas(x->x_glist);

• int widthspec = x->x_text->te_width;
• int widthlimit = (widthspec ? widthspec : BOXWIDTH);
• int inindex = 0;
• int selstart = 0, selend = 0;

• int widthspec_c = x->x_text->te_width;
• int widthlimit_c = (widthspec_c ? widthspec_c : BOXWIDTH);
• int inindex_b = 0;
• int inindex_c = 0;
• int selstart_b = 0, selend_b = 0;
• int x_bufsize_c = u8_charnum(x->x_buf, x->x_bufsize); /* if we're a GOP (the new, "goprect" style) borrow the font

size from the inside to preserve the spacing */ if (pd_class(&x->x_text->te_pd) == canvas_class && @@ -193,65 +215,76 @@ if (x->x_bufsize >= 100) tempbuf = (char *)t_getbytes(2 * x->x_bufsize + 1); else tempbuf = smallbuf;

• while (x->x_bufsize - inindex > 0)

• while (x_bufsize_c - inindex_c > 0) {

•    int inchars = x->x_bufsize - inindex;
  

•    int maxindex = (inchars > widthlimit ? widthlimit : inchars);
  

•    int inchars_b  = x->x_bufsize - inindex_b;
  

•    int inchars_c  = x_bufsize_c  - inindex_c;
  

•    int maxindex_c = (inchars_c > widthlimit_c ? widthlimit_c :  
  

inchars_c);

•    int maxindex_b = u8_offset(x->x_buf + inindex_b, maxindex_c);
   int eatchar = 1;
  

•    int foundit = firstone(x->x_buf + inindex, '\n', maxindex);
  

•    if (foundit < 0)
  

•    int foundit_b  = firstone(x->x_buf + inindex_b, '\n',  
  

maxindex_b);

•    int foundit_c;
  

•    if (foundit_b < 0)
   {
  

•        if (inchars > widthlimit)
  

•        if (inchars_c > widthlimit_c)
       {
  

•            foundit = lastone(x->x_buf + inindex, ' ', maxindex);
  

•            if (foundit < 0)
  

•            foundit_b = lastone(x->x_buf + inindex_b, ' ',  
  

maxindex_b);

•            if (foundit_b < 0)
           {
  

•                foundit = maxindex;
  

•                foundit_b = maxindex_b;
  

•                foundit_c = maxindex_c;
               eatchar = 0;
           }
  

•            else
  

•                foundit_c = u8_charnum(x->x_buf + inindex_b,  
  

foundit_b); } else {

•            foundit = inchars;
  

•            foundit_b = inchars_b;
  

•            foundit_c = inchars_c;
           eatchar = 0;
       }
   }
  

•    else
  

•        foundit_c = u8_charnum(x->x_buf + inindex_b, foundit_b);
  

•  if (nlines == findy)
   {
       int actualx = (findx < 0 ? 0 :
  

•            (findx > foundit ? foundit : findx));
  

•        *indexp = inindex + actualx;
  

•            (findx > foundit_c ? foundit_c : findx));
  

•        *indexp = inindex_b + u8_offset(x->x_buf + inindex_b,  
  

actualx); reportedindex = 1; }

•    strncpy(tempbuf+outchars, x->x_buf + inindex, foundit);
  

•    if (x->x_selstart >= inindex &&
  

•        x->x_selstart <= inindex + foundit + eatchar)
  

•            selstart = x->x_selstart + outchars - inindex;
  

•    if (x->x_selend >= inindex &&
  

•        x->x_selend <= inindex + foundit + eatchar)
  

•            selend = x->x_selend + outchars - inindex;
  

•    outchars += foundit;
  

•    inindex += (foundit + eatchar);
  

•    if (inindex < x->x_bufsize)
  

•        tempbuf[outchars++] = '\n';
  

•    if (foundit > ncolumns)
  

•        ncolumns = foundit;
  

•    strncpy(tempbuf+outchars_b, x->x_buf + inindex_b, foundit_b);
  

•    if (x->x_selstart >= inindex_b &&
  

•        x->x_selstart <= inindex_b + foundit_b + eatchar)
  

•            selstart_b = x->x_selstart + outchars_b - inindex_b;
  

•    if (x->x_selend >= inindex_b &&
  

•        x->x_selend <= inindex_b + foundit_b + eatchar)
  

•            selend_b = x->x_selend + outchars_b - inindex_b;
  

•    outchars_b += foundit_b;
  

•    inindex_b += (foundit_b + eatchar);
  

•    inindex_c += (foundit_c + eatchar);
  

•    if (inindex_b < x->x_bufsize)
  

•        tempbuf[outchars_b++] = '\n';
  

•    if (foundit_c > ncolumns)
  

•        ncolumns = foundit_c;
   nlines++;
  

  } if (!reportedindex)

•    *indexp = outchars;
  

•    *indexp = outchars_b;
  

  dispx = text_xpix(x->x_text, x->x_glist); dispy = text_ypix(x->x_text, x->x_glist); if (nlines < 1) nlines = 1;

• if (!widthspec)

• if (!widthspec_c) { while (ncolumns < 3) {

•        tempbuf[outchars++] = ' ';
  

•        tempbuf[outchars_b++] = ' ';
       ncolumns++;
   }
  

  }

• else ncolumns = widthspec;

• else ncolumns = widthspec_c; pixwide = ncolumns * fontwidth + (LMARGIN + RMARGIN); pixhigh = nlines * fontheight + (TMARGIN + BMARGIN);

@@ -259,31 +292,32 @@ sys_vgui("pdtk_text_new .x%lx.c {%s %s text} %f %f {%.*s} %d %s\n", canvas, x->x_tag, rtext_gettype(x)->s_name, dispx + LMARGIN, dispy + TMARGIN,

•        outchars, tempbuf, sys_hostfontsize(font),
  

•        outchars_b, tempbuf, sys_hostfontsize(font),
       (glist_isselected(x->x_glist,
           &x->x_glist->gl_gobj)? "blue" : "black"));
  

  else if (action == SEND_UPDATE) { sys_vgui("pdtk_text_set .x%lx.c %s {%.*s}\n",

•        canvas, x->x_tag, outchars, tempbuf);
  

•        canvas, x->x_tag, outchars_b, tempbuf);
   if (pixwide != x->x_drawnwidth || pixhigh != x->x_drawnheight)
       text_drawborder(x->x_text, x->x_glist, x->x_tag,
           pixwide, pixhigh, 0);
   if (x->x_active)
   {
  

•        if (selend > selstart)
  

•        if (selend_b > selstart_b)
       {
           sys_vgui(".x%lx.c select from %s %d\n", canvas,
  

•                x->x_tag, selstart);
  

•                x->x_tag, u8_charnum(x->x_buf, selstart_b));
           sys_vgui(".x%lx.c select to %s %d\n", canvas,
  

•                x->x_tag, selend + (sys_oldtclversion ? 0 : -1));
  

•                x->x_tag, u8_charnum(x->x_buf, selend_b)
  

• 	      + (sys_oldtclversion ? 0 : -1));
        sys_vgui(".x%lx.c focus \"\"\n", canvas);
    }
    else
    {
        sys_vgui(".x%lx.c select clear\n", canvas);
        sys_vgui(".x%lx.c icursor %s %d\n", canvas, x->x_tag,
  

•                selstart);
  

•                u8_charnum(x->x_buf, selstart_b));
           sys_vgui(".x%lx.c focus %s\n", canvas, x->x_tag);
       }
   }
  

@@ -448,12 +482,12 @@ .... } */ if (x->x_selstart && (x->x_selstart == x->x_selend))

•            x->x_selstart--;
  

•            u8_dec(x->x_buf, &x->x_selstart);
   }
   else if (n == 127)      /* delete */
   {
       if (x->x_selend < x->x_bufsize && (x->x_selstart == x- 
  

...

x_selend))

•            x->x_selend++;
  

•            u8_inc(x->x_buf, &x->x_selend);
   }
  
   ndel = x->x_selend - x->x_selstart;
  

@@ -466,7 +500,13 @@ /* at Guenter's suggestion, use 'n>31' to test wither a character might be printable in whatever 8-bit character set we find ourselves. */

•    if (n == '\n' || (n > 31 && n != 127))
  

+/*-- moo:

• ... but test with "<" rather than "!=" in order to accomodate

unicode

• codepoints for n (which we get since Tk is sending the "%A"

substitution

• for bind <Key>), effectively reducing the coverage of this clause

to 7

• bits. Case n>127 is covered by the next clause.

+*/

•    if (n == '\n' || (n > 31 && n < 127))
   {
       newsize = x->x_bufsize+1;
       x->x_buf = resizebytes(x->x_buf, x->x_bufsize, newsize);
  

@@ -476,20 +516,39 @@ x->x_bufsize = newsize; x->x_selstart = x->x_selstart + 1; }

• /*--moo: check for unicode codepoints beyond 7-bit ASCII --*/
• else if (n > 127)

•    {
  

•        int ch_nbytes = u8_wc_nbytes(n);
  

•        newsize = x->x_bufsize + ch_nbytes;
  

•        x->x_buf = resizebytes(x->x_buf, x->x_bufsize, newsize);
  

•        for (i = x->x_bufsize; i > x->x_selstart; i--)
  

•            x->x_buf[i] = x->x_buf[i-1];
  

•        x->x_bufsize = newsize;
  

•        /*-- moo: assume canvas_key() has encoded keysym as  
  

UTF-8 */

•        strncpy(x->x_buf+x->x_selstart, keysym->s_name,  
  

ch_nbytes);

•        x->x_selstart = x->x_selstart + ch_nbytes;
  

•    }
   x->x_selend = x->x_selstart;
   x->x_glist->gl_editor->e_textdirty = 1;
  

  } else if (!strcmp(keysym->s_name, "Right")) { if (x->x_selend == x->x_selstart && x->x_selstart < x-

...

x_bufsize)

•        x->x_selend = x->x_selstart = x->x_selstart + 1;
  

•    {
  

•        u8_inc(x->x_buf, &x->x_selstart);
  

•        x->x_selend = x->x_selstart;
  

•    }
   else
       x->x_selstart = x->x_selend;
  

  } else if (!strcmp(keysym->s_name, "Left")) { if (x->x_selend == x->x_selstart && x->x_selstart > 0)

•        x->x_selend = x->x_selstart = x->x_selstart - 1;
  

•    {
  

•        u8_dec(x->x_buf, &x->x_selstart);
  

•        x->x_selend = x->x_selstart;
  

•    }
   else
       x->x_selend = x->x_selstart;
  

  }

@@ -497,18 +556,18 @@ else if (!strcmp(keysym->s_name, "Up")) { if (x->x_selstart)

•        x->x_selstart--;
  

•        u8_dec(x->x_buf, &x->x_selstart);
   while (x->x_selstart > 0 && x->x_buf[x->x_selstart] != '\n')
  

•        x->x_selstart--;
  

•        u8_dec(x->x_buf, &x->x_selstart);
   x->x_selend = x->x_selstart;
  

  } else if (!strcmp(keysym->s_name, "Down")) { while (x->x_selend < x->x_bufsize && x->x_buf[x->x_selend] != '\n')

•        x->x_selend++;
  

•        u8_inc(x->x_buf, &x->x_selend);
   if (x->x_selend < x->x_bufsize)
  

•        x->x_selend++;
  

•        u8_inc(x->x_buf, &x->x_selend);
   x->x_selstart = x->x_selend;
  

  } rtext_senditup(x, SEND_UPDATE, &w, &h, &indx);

Index: src/s_utf8.h

--- src/s_utf8.h (revision 0) +++ src/s_utf8.h (revision 0) @@ -0,0 +1,88 @@ +#ifndef S_UTF8_H +#define S_UTF8_H

+/*--moo--*/ +#ifndef u_int32_t +# define u_int32_t unsigned int +#endif

+#ifndef UCS4 +# define UCS4 u_int32_t +#endif

+/* UTF8_SUPPORT_FULL_UCS4

• • define this to support the full potential range of UCS-4

codepoints

• • (in anticipation of a future UTF-8 standard)
• */

+/*#define UTF8_SUPPORT_FULL_UCS4 1*/ +#undef UTF8_SUPPORT_FULL_UCS4

+/* UTF8_MAXBYTES

• • maximum number of bytes required to represent a single

character in UTF-8

• • UTF8_MAXBYTES1 = UTF8_MAXBYTES+1
• • maximum bytes per character including NUL terminator
• */

+#ifdef UTF8_SUPPORT_FULL_UCS4 +# ifndef UTF8_MAXBYTES +# define UTF8_MAXBYTES 6 +# endif +# ifndef UTF8_MAXBYTES1 +# define UTF8_MAXBYTES1 7 +# endif +#else +# ifndef UTF8_MAXBYTES +# define UTF8_MAXBYTES 4 +# endif +# ifndef UTF8_MAXBYTES1 +# define UTF8_MAXBYTES1 5 +# endif +#endif +/*--/moo--*/

+/* is c the start of a utf8 sequence? */ +#define isutf(c) (((c)&0xC0)!=0x80)

+/* convert UTF-8 data to wide character */ +int u8_toucs(u_int32_t *dest, int sz, char *src, int srcsz);

+/* the opposite conversion */ +int u8_toutf8(char *dest, int sz, u_int32_t *src, int srcsz);

+/* moo: get byte length of character number, or 0 if not supported */ +int u8_wc_nbytes(u_int32_t ch);

+/* moo: compute required storage for UTF-8 encoding of 's[0..n-1]' */ +int u8_wcs_nbytes(u_int32_t *ucs, int size);

+/* single character to UTF-8, no NUL termination */ +int u8_wc_toutf8(char *dest, u_int32_t ch);

+/* moo: single character to UTF-8, with NUL termination */ +int u8_wc_toutf8_nul(char *dest, u_int32_t ch);

+/* character number to byte offset */ +int u8_offset(char *str, int charnum);

+/* byte offset to character number */ +int u8_charnum(char *s, int offset);

+/* return next character, updating an index variable */ +u_int32_t u8_nextchar(char *s, int *i);

+/* move to next character */ +void u8_inc(char *s, int *i);

+/* move to previous character */ +void u8_dec(char *s, int *i);

+/* moo: move pointer to next character */ +void u8_inc_ptr(char **sp);

+/* moo: move pointer to previous character */ +void u8_dec_ptr(char **sp);

+/* returns length of next utf-8 sequence */ +int u8_seqlen(char *s);

+#endif /* S_UTF8_H */ <test-utf8.pd>

---

"[T]he greatest purveyor of violence in the world today [is] my own government." - Martin Luther King, Jr.

---

Pd-dev mailing list Pd-dev@iem.at http://lists.puredata.info/listinfo/pd-dev