Skip to content

Commit 534c89e

Browse files
DannyDaemonicDannyDaemonic
committed
Track character width
1 parent 8f9f962 commit 534c89e

File tree

1 file changed

+106
-51
lines changed

1 file changed

+106
-51
lines changed

‎examples/common.cpp

+106-51
Original file line numberDiff line numberDiff line change
@@ -15,29 +15,13 @@
1515
#endif
1616

1717
#if defined (_WIN32)
18+
#define WIN32_LEAN_AND_MEAN
19+
#define NOMINMAX
20+
#include <windows.h>
1821
#include <fcntl.h>
1922
#include <io.h>
20-
#pragma comment(lib,"kernel32.lib")
21-
extern "C" __declspec(dllimport) void* __stdcall GetStdHandle(unsigned long nStdHandle);
22-
extern "C" __declspec(dllimport) int __stdcall GetConsoleMode(void* hConsoleHandle, unsigned long* lpMode);
23-
extern "C" __declspec(dllimport) int __stdcall SetConsoleMode(void* hConsoleHandle, unsigned long dwMode);
24-
extern "C" __declspec(dllimport) int __stdcall SetConsoleCP(unsigned int wCodePageID);
25-
extern "C" __declspec(dllimport) int __stdcall SetConsoleOutputCP(unsigned int wCodePageID);
26-
extern "C" __declspec(dllimport) int __stdcall WideCharToMultiByte(unsigned int CodePage, unsigned long dwFlags,
27-
const wchar_t * lpWideCharStr, int cchWideChar,
28-
char * lpMultiByteStr, int cbMultiByte,
29-
const char * lpDefaultChar, bool * lpUsedDefaultChar);
30-
#define ENABLE_LINE_INPUT 0x0002
31-
#define ENABLE_ECHO_INPUT 0x0004
32-
#define CP_UTF8 65001
33-
#define CONSOLE_CHAR_TYPE wchar_t
34-
#define CONSOLE_GET_CHAR() getwchar()
35-
#define CONSOLE_EOF WEOF
3623
#else
3724
#include <unistd.h>
38-
#define CONSOLE_CHAR_TYPE char
39-
#define CONSOLE_GET_CHAR() getchar()
40-
#define CONSOLE_EOF EOF
4125
#endif
4226

4327
int32_t get_num_physical_cores() {
@@ -545,6 +529,7 @@ void console_init(console_state & con_st) {
545529
new_termios.c_cc[VTIME] = 0;
546530
tcsetattr(STDIN_FILENO, TCSANOW, &new_termios);
547531
#endif
532+
setlocale(LC_ALL, "");
548533
}
549534

550535
void console_cleanup(console_state & con_st) {
@@ -557,9 +542,80 @@ void console_cleanup(console_state & con_st) {
557542
console_set_color(con_st, CONSOLE_COLOR_DEFAULT);
558543
}
559544

545+
#if defined (_WIN32)
546+
int puts_get_width(_In_z_ CONST CHAR* lpBuffer) {
547+
DWORD nNumberOfCharsToWrite = strlen(lpBuffer);
548+
549+
HANDLE hConsole = GetStdHandle(STD_OUTPUT_HANDLE);
550+
CONSOLE_SCREEN_BUFFER_INFO bufferInfo;
551+
if (!GetConsoleScreenBufferInfo(hConsole, &bufferInfo)) {
552+
// Make a guess
553+
return 1;
554+
}
555+
COORD initialPosition = bufferInfo.dwCursorPosition;
556+
557+
DWORD written = 0;
558+
WriteConsole(hConsole, lpBuffer, nNumberOfCharsToWrite, &written, nullptr);
559+
560+
CONSOLE_SCREEN_BUFFER_INFO newBufferInfo;
561+
GetConsoleScreenBufferInfo(hConsole, &newBufferInfo);
562+
563+
int width = newBufferInfo.dwCursorPosition.X - initialPosition.X;
564+
if (newBufferInfo.dwCursorPosition.Y > initialPosition.Y) {
565+
width += (newBufferInfo.dwSize.X - initialPosition.X);
566+
}
567+
568+
return width;
569+
}
570+
#endif
571+
572+
char32_t getchar32() {
573+
wchar_t wc = getwchar();
574+
if (static_cast<wint_t>(wc) == WEOF) {
575+
return WEOF;
576+
}
577+
578+
#if WCHAR_MAX == 0xFFFF
579+
if ((wc >= 0xD800) && (wc <= 0xDBFF)) { // Check if wc is a high surrogate
580+
wchar_t low_surrogate = getwchar();
581+
if ((low_surrogate >= 0xDC00) && (low_surrogate <= 0xDFFF)) { // Check if the next wchar is a low surrogate
582+
return (static_cast<char32_t>(wc & 0x03FF) << 10) + (low_surrogate & 0x03FF) + 0x10000;
583+
}
584+
}
585+
if ((wc >= 0xD800) && (wc <= 0xDFFF)) { // Invalid surrogate pair
586+
return 0xFFFD; // Return the replacement character U+FFFD
587+
}
588+
#endif
589+
590+
return static_cast<char32_t>(wc);
591+
}
592+
593+
void append_utf8(char32_t ch, std::string & out) {
594+
if (ch <= 0x7F) {
595+
out.push_back(static_cast<unsigned char>(ch));
596+
} else if (ch <= 0x7FF) {
597+
out.push_back(static_cast<unsigned char>(0xC0 | ((ch >> 6) & 0x1F)));
598+
out.push_back(static_cast<unsigned char>(0x80 | (ch & 0x3F)));
599+
} else if (ch <= 0xFFFF) {
600+
out.push_back(static_cast<unsigned char>(0xE0 | ((ch >> 12) & 0x0F)));
601+
out.push_back(static_cast<unsigned char>(0x80 | ((ch >> 6) & 0x3F)));
602+
out.push_back(static_cast<unsigned char>(0x80 | (ch & 0x3F)));
603+
} else if (ch <= 0x10FFFF) {
604+
out.push_back(static_cast<unsigned char>(0xF0 | ((ch >> 18) & 0x07)));
605+
out.push_back(static_cast<unsigned char>(0x80 | ((ch >> 12) & 0x3F)));
606+
out.push_back(static_cast<unsigned char>(0x80 | ((ch >> 6) & 0x3F)));
607+
out.push_back(static_cast<unsigned char>(0x80 | (ch & 0x3F)));
608+
} else {
609+
// Invalid Unicode code point
610+
}
611+
}
612+
560613
// Helper function to remove the last UTF-8 character from a string
561-
void remove_last_utf8_char(std::string & line) {
562-
if (line.empty()) return;
614+
void pop_back_utf8_char(std::string & line) {
615+
if (line.empty()) {
616+
return;
617+
}
618+
563619
size_t pos = line.length() - 1;
564620

565621
// Find the start of the last UTF-8 character (checking up to 4 bytes back)
@@ -569,33 +625,24 @@ void remove_last_utf8_char(std::string & line) {
569625
line.erase(pos);
570626
}
571627

572-
#if defined (_WIN32)
573-
// Convert a wide Unicode string to an UTF8 string
574-
void win32_utf8_encode(const std::wstring & wstr, std::string & str) {
575-
int size_needed = WideCharToMultiByte(CP_UTF8, 0, &wstr[0], (int)wstr.size(), NULL, 0, NULL, NULL);
576-
std::string strTo(size_needed, 0);
577-
WideCharToMultiByte(CP_UTF8, 0, &wstr[0], (int)wstr.size(), &strTo[0], size_needed, NULL, NULL);
578-
str = strTo;
579-
}
580-
#endif
581-
582628
bool console_readline(console_state & con_st, std::string & line) {
629+
console_set_color(con_st, CONSOLE_COLOR_USER_INPUT);
630+
583631
line.clear();
632+
std::vector<int> widths;
584633
bool is_special_char = false;
585634
bool end_of_stream = false;
586635

587-
console_set_color(con_st, CONSOLE_COLOR_USER_INPUT);
588-
589-
CONSOLE_CHAR_TYPE input_char;
636+
char32_t input_char;
590637
while (true) {
591638
fflush(stdout); // Ensure all output is displayed before waiting for input
592-
input_char = CONSOLE_GET_CHAR();
639+
input_char = getchar32();
593640

594641
if (input_char == '\r' || input_char == '\n') {
595642
break;
596643
}
597644

598-
if (input_char == CONSOLE_EOF || input_char == 0x04 /* Ctrl+D*/) {
645+
if (input_char == WEOF || input_char == 0x04 /* Ctrl+D*/) {
599646
end_of_stream = true;
600647
break;
601648
}
@@ -608,31 +655,39 @@ bool console_readline(console_state & con_st, std::string & line) {
608655
}
609656

610657
if (input_char == '\033') { // Escape sequence
611-
CONSOLE_CHAR_TYPE code = CONSOLE_GET_CHAR();
612-
if (code == '[') {
658+
char32_t code = getchar32();
659+
if (code == '[' || code == 0x1B) {
613660
// Discard the rest of the escape sequence
614-
while ((code = CONSOLE_GET_CHAR()) != CONSOLE_EOF) {
661+
while ((code = getchar32()) != WEOF) {
615662
if ((code >= 'A' && code <= 'Z') || (code >= 'a' && code <= 'z') || code == '~') {
616663
break;
617664
}
618665
}
619666
}
620667
} else if (input_char == 0x08 || input_char == 0x7F) { // Backspace
621-
if (!line.empty()) {
622-
fputs("\b \b", stdout); // Move cursor back, print a space, and move cursor back again
623-
remove_last_utf8_char(line);
668+
if (!widths.empty()) {
669+
int count;
670+
do {
671+
count = widths.back();
672+
widths.pop_back();
673+
// Move cursor back, print spaces, and move cursor back again
674+
for (int i = 0; i < count; i++) {
675+
fputs("\b \b", stdout);
676+
}
677+
pop_back_utf8_char(line);
678+
} while (count == 0 && !widths.empty());
624679
}
625-
} else if (static_cast<unsigned>(input_char) < 32) {
680+
} else if (input_char < 32) {
626681
// Ignore control characters
627682
} else {
628-
#if defined(_WIN32)
629-
std::string utf8_char;
630-
win32_utf8_encode(std::wstring(1, input_char), utf8_char);
631-
line += utf8_char;
632-
fputs(utf8_char.c_str(), stdout);
683+
int offset = line.length();
684+
append_utf8(input_char, line);
685+
#if defined (_WIN32)
686+
int width = puts_get_width(line.c_str() + offset);
687+
widths.push_back(width);
633688
#else
634-
line += input_char;
635-
putchar(input_char);
689+
fputs(line.c_str() + offset, stdout);
690+
widths.push_back(wcwidth(input_char));
636691
#endif
637692
}
638693

@@ -655,7 +710,7 @@ bool console_readline(console_state & con_st, std::string & line) {
655710
putchar('\n');
656711
has_more = !has_more;
657712
} else {
658-
// llama doesn't seem to process a single space
713+
// llama will just eat the single space
659714
if (line.length() == 1 && line.back() == ' ') {
660715
line.clear();
661716
putchar('\b');

0 commit comments

Comments
 (0)