mirror of
https://github.com/ocornut/imgui.git
synced 2024-11-27 17:10:57 +01:00
InputText: amends: add note about STB_TEXTEDIT_GETCHAR. tweak InputTextCalcTextSize() to use similar debug-friendly logic as ImFont:CalcTextSizeA(). misc small tidying up. (#7925)
This commit is contained in:
parent
1674fe96ff
commit
e240bc151a
@ -3836,6 +3836,7 @@ static int InputTextCalcTextLenAndLineCount(const char* text_begin, const char**
|
||||
return line_count;
|
||||
}
|
||||
|
||||
// FIXME: Ideally we'd share code with ImFont::CalcTextSizeA()
|
||||
static ImVec2 InputTextCalcTextSize(ImGuiContext* ctx, const char* text_begin, const char* text_end, const char** remaining, ImVec2* out_offset, bool stop_on_new_line)
|
||||
{
|
||||
ImGuiContext& g = *ctx;
|
||||
@ -3849,8 +3850,12 @@ static ImVec2 InputTextCalcTextSize(ImGuiContext* ctx, const char* text_begin, c
|
||||
const char* s = text_begin;
|
||||
while (s < text_end)
|
||||
{
|
||||
unsigned int c;
|
||||
s += ImTextCharFromUtf8(&c, s, text_end);
|
||||
unsigned int c = (unsigned int)*s;
|
||||
if (c < 0x80)
|
||||
s += 1;
|
||||
else
|
||||
s += ImTextCharFromUtf8(&c, s, text_end);
|
||||
|
||||
if (c == '\n')
|
||||
{
|
||||
text_size.x = ImMax(text_size.x, line_width);
|
||||
@ -3863,7 +3868,7 @@ static ImVec2 InputTextCalcTextSize(ImGuiContext* ctx, const char* text_begin, c
|
||||
if (c == '\r')
|
||||
continue;
|
||||
|
||||
const float char_width = font->GetCharAdvance((ImWchar)c) * scale;
|
||||
const float char_width = ((int)c < font->IndexAdvanceX.Size ? font->IndexAdvanceX.Data[c] : font->FallbackAdvanceX) * scale;
|
||||
line_width += char_width;
|
||||
}
|
||||
|
||||
@ -3883,23 +3888,27 @@ static ImVec2 InputTextCalcTextSize(ImGuiContext* ctx, const char* text_begin, c
|
||||
}
|
||||
|
||||
// Wrapper for stb_textedit.h to edit text (our wrapper is for: statically sized buffer, single-line, wchar characters. InputText converts between UTF-8 and wchar)
|
||||
// With our UTF-8 use of stb_textedit:
|
||||
// - STB_TEXTEDIT_GETCHAR is nothing more than a a "GETBYTE". It's only used to compare to ascii or to copy blocks of text so we are fine.
|
||||
// - One exception is the STB_TEXTEDIT_IS_SPACE feature which would expect a full char in order to handle full-width space such as 0x3000 (see ImCharIsBlankW).
|
||||
// - ...but we don't use that feature.
|
||||
namespace ImStb
|
||||
{
|
||||
|
||||
static int STB_TEXTEDIT_STRINGLEN(const ImGuiInputTextState* obj) { return obj->CurLenA; }
|
||||
static char STB_TEXTEDIT_GETCHAR(const ImGuiInputTextState* obj, int idx) { IM_ASSERT(idx <= obj->CurLenA); return obj->TextA[idx]; }
|
||||
static float STB_TEXTEDIT_GETWIDTH(ImGuiInputTextState* obj, int line_start_idx, int char_idx) { unsigned int c; ImTextCharFromUtf8(&c, obj->TextA.Data + line_start_idx + char_idx, obj->TextA.Data + obj->TextA.Size); if ((ImWchar)c == '\n') return IMSTB_TEXTEDIT_GETWIDTH_NEWLINE; ImGuiContext& g = *obj->Ctx; return g.Font->GetCharAdvance((ImWchar)c) * g.FontScale; }
|
||||
static char STB_TEXTEDIT_NEWLINE = '\n';
|
||||
static void STB_TEXTEDIT_LAYOUTROW(StbTexteditRow* r, ImGuiInputTextState* obj, int line_start_idx)
|
||||
{
|
||||
const char* text = obj->TextA.Data;
|
||||
const char* text_remaining = NULL;
|
||||
const ImVec2 size = InputTextCalcTextSize(obj->Ctx, obj->TextA.Data + line_start_idx, obj->TextA.Data + obj->CurLenA, &text_remaining, NULL, true);
|
||||
const ImVec2 size = InputTextCalcTextSize(obj->Ctx, text + line_start_idx, text + obj->CurLenA, &text_remaining, NULL, true);
|
||||
r->x0 = 0.0f;
|
||||
r->x1 = size.x;
|
||||
r->baseline_y_delta = size.y;
|
||||
r->ymin = 0.0f;
|
||||
r->ymax = size.y;
|
||||
r->num_chars = (int)(text_remaining - (obj->TextA.Data + line_start_idx));
|
||||
r->num_chars = (int)(text_remaining - (text + line_start_idx));
|
||||
}
|
||||
|
||||
static bool ImCharIsSeparatorW(unsigned int c)
|
||||
@ -3922,14 +3931,15 @@ static int is_word_boundary_from_right(ImGuiInputTextState* obj, int idx)
|
||||
if ((obj->Flags & ImGuiInputTextFlags_Password) || idx <= 0)
|
||||
return 0;
|
||||
|
||||
const char* prevPtr = ImTextFindPreviousUtf8Codepoint(obj->TextA.Data, obj->TextA.Data + idx);
|
||||
unsigned int curr; ImTextCharFromUtf8(&curr, obj->TextA.Data + idx, obj->TextA.Data + obj->TextA.Size);
|
||||
unsigned int prev; ImTextCharFromUtf8(&prev, prevPtr, obj->TextA.Data + obj->TextA.Size);
|
||||
const char* curr_p = obj->TextA.Data + idx;
|
||||
const char* prev_p = ImTextFindPreviousUtf8Codepoint(obj->TextA.Data, curr_p);
|
||||
unsigned int curr_c; ImTextCharFromUtf8(&curr_c, curr_p, obj->TextA.Data + obj->TextA.Size);
|
||||
unsigned int prev_c; ImTextCharFromUtf8(&prev_c, prev_p, obj->TextA.Data + obj->TextA.Size);
|
||||
|
||||
bool prev_white = ImCharIsBlankW(prev);
|
||||
bool prev_separ = ImCharIsSeparatorW(prev);
|
||||
bool curr_white = ImCharIsBlankW(curr);
|
||||
bool curr_separ = ImCharIsSeparatorW(curr);
|
||||
bool prev_white = ImCharIsBlankW(prev_c);
|
||||
bool prev_separ = ImCharIsSeparatorW(prev_c);
|
||||
bool curr_white = ImCharIsBlankW(curr_c);
|
||||
bool curr_separ = ImCharIsSeparatorW(curr_c);
|
||||
return ((prev_white || prev_separ) && !(curr_separ || curr_white)) || (curr_separ && !prev_separ);
|
||||
}
|
||||
static int is_word_boundary_from_left(ImGuiInputTextState* obj, int idx)
|
||||
@ -3937,14 +3947,15 @@ static int is_word_boundary_from_left(ImGuiInputTextState* obj, int idx)
|
||||
if ((obj->Flags & ImGuiInputTextFlags_Password) || idx <= 0)
|
||||
return 0;
|
||||
|
||||
const char* prevPtr = ImTextFindPreviousUtf8Codepoint(obj->TextA.Data, obj->TextA.Data + idx);
|
||||
unsigned int prev; ImTextCharFromUtf8(&prev, obj->TextA.Data + idx, obj->TextA.Data + obj->TextA.Size);
|
||||
unsigned int curr; ImTextCharFromUtf8(&curr, prevPtr, obj->TextA.Data + obj->TextA.Size);
|
||||
const char* curr_p = obj->TextA.Data + idx;
|
||||
const char* prev_p = ImTextFindPreviousUtf8Codepoint(obj->TextA.Data, curr_p);
|
||||
unsigned int prev_c; ImTextCharFromUtf8(&prev_c, curr_p, obj->TextA.Data + obj->TextA.Size);
|
||||
unsigned int curr_c; ImTextCharFromUtf8(&curr_c, prev_p, obj->TextA.Data + obj->TextA.Size);
|
||||
|
||||
bool prev_white = ImCharIsBlankW(prev);
|
||||
bool prev_separ = ImCharIsSeparatorW(prev);
|
||||
bool curr_white = ImCharIsBlankW(curr);
|
||||
bool curr_separ = ImCharIsSeparatorW(curr);
|
||||
bool prev_white = ImCharIsBlankW(prev_c);
|
||||
bool prev_separ = ImCharIsSeparatorW(prev_c);
|
||||
bool curr_white = ImCharIsBlankW(curr_c);
|
||||
bool curr_separ = ImCharIsSeparatorW(curr_c);
|
||||
return ((prev_white) && !(curr_separ || curr_white)) || (curr_separ && !prev_separ);
|
||||
}
|
||||
static int STB_TEXTEDIT_MOVEWORDLEFT_IMPL(ImGuiInputTextState* obj, int idx) { idx--; while (idx >= 0 && !is_word_boundary_from_right(obj, idx)) idx--; return idx < 0 ? 0 : idx; }
|
||||
|
Loading…
Reference in New Issue
Block a user