aboutsummaryrefslogtreecommitdiff
path: root/src/printf.c
diff options
context:
space:
mode:
Diffstat (limited to 'src/printf.c')
-rw-r--r--src/printf.c84
1 files changed, 69 insertions, 15 deletions
diff --git a/src/printf.c b/src/printf.c
index ca8d26e4f..11d5b8282 100644
--- a/src/printf.c
+++ b/src/printf.c
@@ -624,9 +624,38 @@ void sqlite3VXPrintf(
case etCHARX:
if( bArgList ){
bufpt = getTextArg(pArgList);
- c = bufpt ? bufpt[0] : 0;
+ length = 1;
+ if( bufpt ){
+ buf[0] = c = *(bufpt++);
+ if( (c&0xc0)==0xc0 ){
+ while( length<4 && (bufpt[0]&0xc0)==0x80 ){
+ buf[length++] = *(bufpt++);
+ }
+ }
+ }else{
+ buf[0] = 0;
+ }
}else{
- c = va_arg(ap,int);
+ unsigned int ch = va_arg(ap,unsigned int);
+ if( ch<0x00080 ){
+ buf[0] = ch & 0xff;
+ length = 1;
+ }else if( ch<0x00800 ){
+ buf[0] = 0xc0 + (u8)((ch>>6)&0x1f);
+ buf[1] = 0x80 + (u8)(ch & 0x3f);
+ length = 2;
+ }else if( ch<0x10000 ){
+ buf[0] = 0xe0 + (u8)((ch>>12)&0x0f);
+ buf[1] = 0x80 + (u8)((ch>>6) & 0x3f);
+ buf[2] = 0x80 + (u8)(ch & 0x3f);
+ length = 3;
+ }else{
+ buf[0] = 0xf0 + (u8)((ch>>18) & 0x07);
+ buf[1] = 0x80 + (u8)((ch>>12) & 0x3f);
+ buf[2] = 0x80 + (u8)((ch>>6) & 0x3f);
+ buf[3] = 0x80 + (u8)(ch & 0x3f);
+ length = 4;
+ }
}
if( precision>1 ){
width -= precision-1;
@@ -634,12 +663,13 @@ void sqlite3VXPrintf(
sqlite3AppendChar(pAccum, width-1, ' ');
width = 0;
}
- sqlite3AppendChar(pAccum, precision-1, c);
+ while( precision-- > 1 ){
+ sqlite3StrAccumAppend(pAccum, buf, length);
+ }
}
- length = 1;
- buf[0] = c;
bufpt = buf;
- break;
+ flag_altform2 = 1;
+ goto adjust_width_for_utf8;
case etSTRING:
case etDYNSTRING:
if( bArgList ){
@@ -654,14 +684,30 @@ void sqlite3VXPrintf(
zExtra = bufpt;
}
if( precision>=0 ){
- for(length=0; length<precision && bufpt[length]; length++){}
+ if( flag_altform2 ){
+ /* Set length to the number of bytes needed in order to display
+ ** precision characters */
+ unsigned char *z = (unsigned char*)bufpt;
+ while( precision-- > 0 && z[0] ){
+ SQLITE_SKIP_UTF8(z);
+ }
+ length = (int)(z - (unsigned char*)bufpt);
+ }else{
+ for(length=0; length<precision && bufpt[length]; length++){}
+ }
}else{
length = 0x7fffffff & (int)strlen(bufpt);
}
+ adjust_width_for_utf8:
+ if( flag_altform2 && width>0 ){
+ /* Adjust width to account for extra bytes in UTF-8 characters */
+ int ii = length - 1;
+ while( ii>=0 ) if( (bufpt[ii--] & 0xc0)==0x80 ) width++;
+ }
break;
- case etSQLESCAPE: /* Escape ' characters */
- case etSQLESCAPE2: /* Escape ' and enclose in '...' */
- case etSQLESCAPE3: { /* Escape " characters */
+ case etSQLESCAPE: /* %q: Escape ' characters */
+ case etSQLESCAPE2: /* %Q: Escape ' and enclose in '...' */
+ case etSQLESCAPE3: { /* %w: Escape " characters */
int i, j, k, n, isnull;
int needQuote;
char ch;
@@ -675,9 +721,17 @@ void sqlite3VXPrintf(
}
isnull = escarg==0;
if( isnull ) escarg = (xtype==etSQLESCAPE2 ? "NULL" : "(NULL)");
+ /* For %q, %Q, and %w, the precision is the number of byte (or
+ ** characters if the ! flags is present) to use from the input.
+ ** Because of the extra quoting characters inserted, the number
+ ** of output characters may be larger than the precision.
+ */
k = precision;
for(i=n=0; k!=0 && (ch=escarg[i])!=0; i++, k--){
if( ch==q ) n++;
+ if( flag_altform2 && (ch&0xc0)==0xc0 ){
+ while( (escarg[i+1]&0xc0)==0x80 ){ i++; }
+ }
}
needQuote = !isnull && xtype==etSQLESCAPE2;
n += i + 3;
@@ -700,10 +754,7 @@ void sqlite3VXPrintf(
if( needQuote ) bufpt[j++] = q;
bufpt[j] = 0;
length = j;
- /* The precision in %q and %Q means how many input characters to
- ** consume, not the length of the output...
- ** if( precision>=0 && precision<length ) length = precision; */
- break;
+ goto adjust_width_for_utf8;
}
case etTOKEN: {
Token *pToken;
@@ -742,7 +793,10 @@ void sqlite3VXPrintf(
/*
** The text of the conversion is pointed to by "bufpt" and is
** "length" characters long. The field width is "width". Do
- ** the output.
+ ** the output. Both length and width are in bytes, not characters,
+ ** at this point. If the "!" flag was present on string conversions
+ ** indicating that width and precision should be expressed in characters,
+ ** then the values have been translated prior to reaching this point.
*/
width -= length;
if( width>0 ){