fix inconsistencies btwn xoff_to_index & rendering

5 files changed, 80 insertions, 104 deletions

diff --git a/buffer.c b/buffer.c
index e2d32ea..2bc0efe 100644
--- a/buffer.c
+++ b/buffer.c
@@ -297,7 +297,9 @@ Settings *buffer_settings(TextBuffer *buffer) {
 }
 
 u8 buffer_tab_width(TextBuffer *buffer) {
-	return buffer_settings(buffer)->tab_width;
+	if (!buffer->tab_width)
+		buffer->tab_width = buffer_settings(buffer)->tab_width;
+	return buffer->tab_width;
 }
 
 bool buffer_indent_with_spaces(TextBuffer *buffer) {
@@ -780,6 +782,29 @@ static void buffer_print(TextBuffer const *buffer) {
 	fflush(stdout);
 }
 
+static void buffer_render_char(TextBuffer *buffer, Font *font, TextRenderState *state, char32_t c) {
+	switch (c) {
+	case '\n': assert(0); break;
+	case '\r': break; // for CRLF line endings
+	case '\t': {
+		u16 tab_width = buffer_tab_width(buffer);
+		double x = state->x;
+		double space_size = text_font_char_width(font, ' ');
+		double tab_width_px = tab_width * space_size;
+		double tabs = x / tab_width_px;
+		double tab_stop = (1.0 + floor(tabs)) * tab_width_px;
+		if (tab_stop - x < space_size * 0.5) {
+			// tab shouldn't be less than half a space
+			tab_stop += tab_width_px;
+		}
+		state->x = tab_stop;
+	} break;
+	default:
+		text_char_with_state(font, state, c); 
+		break;
+	}
+}
+
 // convert line character index to offset in pixels
 static double buffer_index_to_xoff(TextBuffer *buffer, u32 line_number, u32 index) {
 	if (line_number >= buffer->nlines) {
@@ -788,41 +813,15 @@ static double buffer_index_to_xoff(TextBuffer *buffer, u32 line_number, u32 inde
 	}
 	Line *line = &buffer->lines[line_number];
 	char32_t *str = line->str;
-	const Settings *settings = buffer_settings(buffer);
-	Font *font = buffer_font(buffer);
-	u8 tab_width = settings->tab_width;
 	if (index > line->len)
 		index = line->len;
-	
-	if (settings->force_monospace) {
-		u32 col = 0;
-		for (u32 i = 0; i < index; ++i) {
-			switch (str[i]) {
-			case '\t': {
-				do
-					++col;
-				while (col % tab_width);
-			} break;
-			default:
-				++col;
-				break;
-			}
-		}
-		return col * (double)text_font_char_width(font, ' ');
-	} else {
-		double x = 0;
-		for (u32 i = 0; i < index; ++i) {
-			switch (str[i]) {
-			case '\t': {
-				x += tab_width * text_font_char_width(font, ' ');
-			} break;
-			default:
-				x += text_font_char_width(font, str[i]);
-				break;
-			}
-		}
-		return x;
+	Font *font = buffer_font(buffer);
+	TextRenderState state = text_render_state_default;
+	state.render = false;
+	for (u32 i = 0; i < index; ++i) {
+		buffer_render_char(buffer, font, &state, str[i]);
 	}
+	return state.x;
 }
 
 // convert line x offset in pixels to character index
@@ -836,52 +835,21 @@ static u32 buffer_xoff_to_index(TextBuffer *buffer, u32 line_number, double xoff
 	}
 	Line *line = &buffer->lines[line_number];
 	char32_t *str = line->str;
-	const Settings *settings = buffer_settings(buffer);
 	Font *font = buffer_font(buffer);
-	u8 tab_width = settings->tab_width;
-	
-	if (settings->force_monospace) {
-		u32 target_col = (u32)(xoff / text_font_char_width(font, ' '));
-		u32 col = 0;
-		for (u32 i = 0; i < line->len; ++i) {
-			switch (str[i]) {
-			case '\t': {
-				do {
-					if (col == target_col)
-						return i;
-					++col;
-				} while (col % tab_width);
-			} break;
-			default:
-				if (col == target_col)
-					return i;
-				++col;
-				break;
-			}
-		}
-		return line->len;
-	} else {
-		double x = 0;
-		for (u32 i = 0; i < line->len; ++i) {
-			double x_prev = x;
-			switch (str[i]) {
-			case '\t': {
-				x += tab_width * text_font_char_width(font, ' ');
-			} break;
-			default:
-				x += text_font_char_width(font, str[i]);
-				break;
-			}
-			if (x > xoff) {
-				if (xoff - x_prev < x - xoff) {
-					return i;
-				} else {
-					return i + 1;
-				}
-			}
+	TextRenderState state = text_render_state_default;
+	state.render = false;
+	for (u32 i = 0; i < line->len; ++i) {
+		double x0 = state.x;
+		buffer_render_char(buffer, font, &state, str[i]);
+		double x1 = state.x;
+		if (x1 > xoff) {
+			if (x1 - xoff > xoff - x0)
+				return i;
+			else
+				return i + 1;
 		}
-		return line->len;
 	}
+	return line->len;
 }
 
 
@@ -2956,7 +2924,9 @@ bool buffer_cursor_move_to_matching_bracket(TextBuffer *buffer) {
 
 // Render the text buffer in the given rectangle
 void buffer_render(TextBuffer *buffer, Rect r) {
+	const Settings *settings = buffer_settings(buffer);
 	
+	buffer->tab_width = settings->tab_width;
 	buffer_lsp(buffer); // this will send didOpen/didClose if the buffer's LSP changed
 	
 	if (r.size.x < 1 || r.size.y < 1) {
@@ -2978,7 +2948,6 @@ void buffer_render(TextBuffer *buffer, Rect r) {
 	float char_height = text_font_char_height(font);
 
 	Ted *ted = buffer->ted;
-	const Settings *settings = buffer_settings(buffer);
 	const u32 *colors = settings->colors;
 	const float padding = settings->padding;
 	const float border_thickness = settings->border_thickness;
@@ -3081,7 +3050,6 @@ void buffer_render(TextBuffer *buffer, Rect r) {
 	
 	// what x coordinate to start rendering the text from
 	double render_start_x = x1 - (double)buffer->scroll_x * text_font_char_width(font, ' ');
-	u32 column = 0;
 
 	if (buffer->selection) { // draw selection
 		BufferPos sel_start = {0}, sel_end = {0};
@@ -3155,7 +3123,10 @@ void buffer_render(TextBuffer *buffer, Rect r) {
 
 
 	TextRenderState text_state = text_render_state_default;
-	text_state.x = render_start_x;
+	text_state.x = 0;
+	// we need to start at x = 0 to be consistent with
+	// buffer_index_to_xoff and the like (because tabs are difficult).
+	text_state.x_render_offset = (float)render_start_x;
 	text_state.y = render_start_y;
 	text_state.min_x = x1;
 	text_state.min_y = y1;
@@ -3182,33 +3153,18 @@ void buffer_render(TextBuffer *buffer, Rect r) {
 				ColorSetting color = syntax_char_type_to_color_setting(type);
 				rgba_u32_to_floats(colors[color], text_state.color);
 			}
-			switch (c) {
-			case '\n': assert(0); break;
-			case '\r': break; // for CRLF line endings
-			case '\t': {
-				uint tab_width = buffer_tab_width(buffer);
-				do {
-					text_char_with_state(font, &text_state, ' ');
-					++column;
-				} while (column % tab_width);
-			} break;
-			default:
-				text_char_with_state(font, &text_state, c);
-				++column;
-				break;
-			}
+			buffer_render_char(buffer, font, &text_state, c);
 		}
 
 		// next line
 		text_state_break_kerning(&text_state);
-		text_state.x = render_start_x;
+		text_state.x = 0;
 		if (text_state.y > text_state.max_y) {
 			buffer->last_line_on_screen = line_idx;
 			// made it to the bottom of the buffer view.
 			break;
 		}
 		text_state.y += text_font_char_height(font);
-		column = 0;
 	}
 	if (buffer->last_line_on_screen == 0) buffer->last_line_on_screen = nlines - 1;
 	
diff --git a/development.md b/development.md
index 7ad5b7d..c633885 100644
--- a/development.md
+++ b/development.md
@@ -5,21 +5,32 @@
 To build the debug version of ted, run `make` (or `make.bat` on Windows).
 By default we use the `ninja` build system (`sudo apt install ninja` or similar).
 
+## Architecture
+
+The big `Ted` struct has basically all of the data for ted.
+
+In general, fixed-size types such as `u32` should be preferred to `unsigned int` & co.
+
+
 ## Files
 
 Most function declarations should go in `ted.h`.
 The exceptions are only for self-contained files, like `text.c`,
 which gets its own header file `text.h`.
 
-As much as possible, OS-dependent functions should be put in `os.h/os-*.c`.
-(But "pure" functions like `qsort_with_context`, which could
-in theory be implemented on any platform with just plain C, should be put
-in `util.c` even if they use OS-specific library functions.)
+Everything includes `base.h`, which has useful type definitions and stuff.
+This includes shorter names for fixed-size types (e.g. `i8` for `int8_t`).
+
+`util.h`/`util.c` has "utility" functions which aren't specific to ted.
+
+`os.h`/`os-*.c` has OS-dependent functions such as file system
+stuff which is (annoyingly) lacking from standard C.
 
 ## Unicode
 
-ted stores text as UTF-32. We assume that code points are characters.
-This is not correct for combining diacritics and will hopefully be fixed at some point.
+ted stores text as UTF-32. Left/right move by code points,
+which isn't correct for e.g. combining diacritics. Hopefully this will
+be fixed at some point.
 
 All paths are stored as UTF-8, and annoyingly have to be converted to UTF-16 for Windows
 functions (why hasn't Windows made UTF-8 versions of all their API functions yet to save
diff --git a/ted.h b/ted.h
index 956ccef..f6c0e78 100644
--- a/ted.h
+++ b/ted.h
@@ -412,6 +412,8 @@ typedef struct {
 	u32 nlines;
 	/// capacity of \ref lines
 	u32 lines_capacity;
+	/// cached tab width, updated every frame
+	u8 tab_width;
 	
 	/// which LSP this document is open in
 	LSPID lsp_opened_in;
diff --git a/text.c b/text.c
index 8b2acc3..de5438c 100644
--- a/text.c
+++ b/text.c
@@ -63,6 +63,7 @@ const TextRenderState text_render_state_default = {
 	.color = {1, 0, 1, 1},
 	.x_largest = -FLT_MAX, .y_largest = -FLT_MAX,
 	.prev_glyph = 0,
+	.x_render_offset = 0,
 };
 
 static char text_err[200];
@@ -411,8 +412,8 @@ top:;
 	
 	float s0 = q.s0, t0 = q.t0;
 	float s1 = q.s1, t1 = q.t1;
-	float x0 = roundf(q.x0), y0 = roundf(q.y0);
-	float x1 = roundf(q.x1), y1 = roundf(q.y1);
+	float x0 = roundf(q.x0 + state->x_render_offset), y0 = roundf(q.y0);
+	float x1 = roundf(q.x1 + state->x_render_offset), y1 = roundf(q.y1);
 	const float min_x = state->min_x, max_x = state->max_x;
 	const float min_y = state->min_y, max_y = state->max_y;
 	
diff --git a/text.h b/text.h
index 8b2f133..3fd603c 100644
--- a/text.h
+++ b/text.h
@@ -44,6 +44,12 @@ typedef struct {
 	/// index of previous glyph rendered, or 0 if this is the first
 	int prev_glyph;
 	
+	/// added to x for rendering
+	/// this exists for complicated reasons
+	/// basically we want a way of consistently getting the size
+	/// of text without error from floating point imprecision
+	float x_render_offset;
+	
 	/// used for forwards-compatibility
 	char _reserved[64];
 } TextRenderState;