VM_sprintf: support color codes in %s

git-svn-id: svn://svn.icculus.org/twilight/trunk/darkplaces@11612 d7cf8633-e32d-0410-b094-e92efae38249
::stable-branch::merge=a327a6465c19b19fb0172d0583da6d94a655fedf

diff --git a/dpdefs/dpextensions.qc b/dpdefs/dpextensions.qc
index 9f2911a190114f23a5fb5e59f4cb5d9f61aa9090..caa9b6286e0cb91dee5c1c9af0d4bb2d147953a4 100644 (file)
--- a/dpdefs/dpextensions.qc
+++ b/dpdefs/dpextensions.qc
@@ -1027,7 +1027,9 @@ string(string format, ...) sprintf = #627;
 //    For conversions s and c, the flag # makes precision and width interpreted
 //      as byte count, by default it is interpreted as character count in UTF-8
 //      enabled engines. No other conversions can create wide characters, and #
-//      has another meaning in these.
+//      has another meaning in these. When in character count mode, color codes
+//      are ignored. To get UTF-8 semantics WITHOUT color code parsing, use
+//      the + flag.
 
 //DP_QC_STRFTIME
 //idea: LordHavoc

diff --git a/prvm_cmds.c b/prvm_cmds.c
index 7f1900d9826f6d44d3799b78e01b1803bcd7197a..217d897d8464e8155c319d442d57199d474a07bc 100644 (file)
--- a/prvm_cmds.c
+++ b/prvm_cmds.c
@@ -6326,7 +6326,10 @@ nolength:
                                                        {
                                                                if(precision < 0) // not set
                                                                        precision = end - o - 1;
-                                                               o += u8_strpad(o, end - o, GETARG_STRING(thisarg), (flags & PRINTF_LEFT) != 0, width, precision);
+                                                               if(flags & PRINTF_SIGNPOSITIVE)
+                                                                       o += u8_strpad(o, end - o, GETARG_STRING(thisarg), (flags & PRINTF_LEFT) != 0, width, precision);
+                                                               else
+                                                                       o += u8_strpad_colorcodes(o, end - o, GETARG_STRING(thisarg), (flags & PRINTF_LEFT) != 0, width, precision);
                                                        }
                                                        break;
                                                default:

diff --git a/utf8lib.c b/utf8lib.c
index 8994bd7a00a697e124779b86318919aea0d5aaf9..76cc813766d944ed399efdb03f633524ace59187 100644 (file)
--- a/utf8lib.c
+++ b/utf8lib.c
@@ -226,6 +226,29 @@ size_t u8_strlen(const char *_s)
        return len;
 }
 
+static int colorcode_skipwidth(const unsigned char *s)
+{
+       if(*s == STRING_COLOR_TAG)
+       {
+               if(s[1] <= '9' && s[1] >= '0') // ^[0-9] found
+               {
+                       return 2;
+               }
+               else if(s[1] == STRING_COLOR_RGB_TAG_CHAR &&
+                       ((s[2] >= '0' && s[2] <= '9') || (s[2] >= 'a' && s[2] <= 'f') || (s[2] >= 'A' && s[2] <= 'F')) &&
+                       ((s[3] >= '0' && s[3] <= '9') || (s[3] >= 'a' && s[3] <= 'f') || (s[3] >= 'A' && s[3] <= 'F')) &&
+                       ((s[4] >= '0' && s[4] <= '9') || (s[4] >= 'a' && s[4] <= 'f') || (s[4] >= 'A' && s[4] <= 'F')))
+               {
+                       return 5;
+               }
+               else if(s[1] == STRING_COLOR_TAG)
+               {
+                       return 1; // special case, do NOT call colorcode_skipwidth for next char
+               }
+       }
+       return 0;
+}
+
 /** Get the number of characters in a part of an UTF-8 string.
  * @param _s    An utf-8 encoded null-terminated string.
  * @param n     The maximum number of bytes.
@@ -274,6 +297,49 @@ size_t u8_strnlen(const char *_s, size_t n)
        return len;
 }
 
+static size_t u8_strnlen_colorcodes(const char *_s, size_t n)
+{
+       size_t st, ln;
+       size_t len = 0;
+       const unsigned char *s = (const unsigned char*)_s;
+
+       while (*s && n)
+       {
+               int w = colorcode_skipwidth(s);
+               n -= w;
+               s += w;
+               if(w > 1) // == 1 means single caret
+                       continue;
+
+               // ascii char, skip u8_analyze
+               if (*s < 0x80 || !utf8_enable.integer)
+               {
+                       ++len;
+                       ++s;
+                       --n;
+                       continue;
+               }
+
+               // invalid, skip u8_analyze
+               if (*s < 0xC2)
+               {
+                       ++s;
+                       --n;
+                       continue;
+               }
+
+               if (!u8_analyze((const char*)s, &st, &ln, NULL, n))
+                       break;
+               // valid character, see if it's still inside the range specified by n:
+               if (n < st + ln)
+                       return len;
+               ++len;
+               n -= st + ln;
+               s += st + ln;
+       }
+       return len;
+}
+
 /** Get the number of bytes used in a string to represent an amount of characters.
  * @param _s    An utf-8 encoded null-terminated string.
  * @param n     The number of characters we want to know the byte-size for.
@@ -318,6 +384,46 @@ size_t u8_bytelen(const char *_s, size_t n)
        return len;
 }
 
+static size_t u8_bytelen_colorcodes(const char *_s, size_t n)
+{
+       size_t st, ln;
+       size_t len = 0;
+       const unsigned char *s = (const unsigned char*)_s;
+
+       while (*s && n)
+       {
+               int w = colorcode_skipwidth(s);
+               len += w;
+               s += w;
+               if(w > 1) // == 1 means single caret
+                       continue;
+
+               // ascii char, skip u8_analyze
+               if (*s < 0x80 || !utf8_enable.integer)
+               {
+                       ++len;
+                       ++s;
+                       --n;
+                       continue;
+               }
+
+               // invalid, skip u8_analyze
+               if (*s < 0xC2)
+               {
+                       ++s;
+                       ++len;
+                       continue;
+               }
+
+               if (!u8_analyze((const char*)s, &st, &ln, NULL, U8_ANALYZE_INFINITY))
+                       break;
+               --n;
+               s += st + ln;
+               len += st + ln;
+       }
+       return len;
+}
+
 /** Get the byte-index for a character-index.
  * @param _s      An utf-8 encoded string.
  * @param i       The character-index for which you want the byte offset.
@@ -798,6 +904,17 @@ size_t u8_strpad(char *out, size_t outsize, const char *in, qboolean leftalign,
        }
 }
 
+size_t u8_strpad_colorcodes(char *out, size_t outsize, const char *in, qboolean leftalign, size_t minwidth, size_t maxwidth)
+{
+       size_t l = u8_bytelen_colorcodes(in, maxwidth);
+       size_t actual_width = u8_strnlen_colorcodes(in, l);
+       int pad = (actual_width >= minwidth) ? 0 : (minwidth - actual_width);
+       int prec = l;
+       int lpad = leftalign ? 0 : pad;
+       int rpad = leftalign ? pad : 0;
+       return dpsnprintf(out, outsize, "%*s%.*s%*s", lpad, "", prec, in, rpad, "");
+}
+
 
 /*
 The two following functions (u8_toupper, u8_tolower) are derived from

diff --git a/utf8lib.h b/utf8lib.h
index 36c16fe1444d967a255cbe11d87f2904929f6d8b..543fbfc0ed1c886e578d61a65218cdb17cb190dd 100644 (file)
--- a/utf8lib.h
+++ b/utf8lib.h
@@ -46,6 +46,7 @@ size_t u8_COM_StringLengthNoColors(const char *s, size_t size_s, qboolean *valid
 char  *u8_encodech(Uchar ch, size_t*, char*buf16);
 
 size_t u8_strpad(char *out, size_t outsize, const char *in, qboolean leftalign, size_t minwidth, size_t maxwidth);
+size_t u8_strpad_colorcodes(char *out, size_t outsize, const char *in, qboolean leftalign, size_t minwidth, size_t maxwidth);
 
 /* Careful: if we disable utf8 but not freetype, we wish to see freetype chars
  * for normal letters. So use E000+x for special chars, but leave the freetype stuff for the