eessppeelllloo/processing4
1
0
Fork 0
mirror of https://github.com/processing/processing4.git synced 2026-02-03 21:59:20 +01:00
Code Issues Packages Projects Releases Wiki Activity

calculate offsets using a single char (end of #194, #342, #226?)

Browse Source
This commit is contained in:
Ben Fry
2022-01-29 11:59:31 -05:00
parent 4bbf290dca
commit c6d9ba0c7f
3 changed files with 108 additions and 72 deletions
Download Patch File Download Diff File Expand all files Collapse all files

159
app/src/processing/app/syntax/JEditTextArea.java
View File

@@ -29,6 +29,7 @@ import java.awt.im.InputMethodRequests;
import processing.app.syntax.im.InputMethodSupport;
import processing.core.PApplet;
/**
* The text area component from the JEdit Syntax (syntax.jedit.org) project.
* This is a very early version of what later was completely rewritten and
@@ -653,7 +654,6 @@ public class JEditTextArea extends JComponent
public int _offsetToX(int line, int offset) {
TokenMarkerState tokenMarker = getTokenMarker();
// Use painter's cached info for speed
FontMetrics fm = painter.getFontMetrics();
getLineText(line, lineSegment);
@@ -1359,80 +1359,105 @@ public class JEditTextArea extends JComponent
return CharacterKinds.Other;
}
/**
* Get the width in pixels of a segment of text within the IDE.
*
* <p>
* Fractional-font aware implementation of Utilities.getTabbedTextWidth that determines if there
* are fractional character widths present in a font in order to return a more accurate pixel
* width for an input segment.
* </p>
*
* @param s The segment of text for which a pixel width should be returned.
* @param metrics The metrics for the font in which the given segment will be drawn.
* @param x The x origin.
* @param expander The strategy for converting tabs into characters.
* @param startOffset The offset to apply before the text will be drawn.
* @return The width of the input segment in pixels with fractional character widths considered.
*/
private int getTabbedTextWidth(Segment s, FontMetrics metrics, float x,
TabExpander expander, int startOffset) {
float additionalOffset =
getPartialPixelWidth(metrics, x, expander, startOffset) * s.length();
if (Platform.isWindows()) {
// When the OS is using 125% or 250% or other fractional scaling,
// the layout gets hosed. This is fixed (in 4.0 beta 4) by
// subtracting this additionalOffset:
// https://github.com/processing/processing4/issues/226
// However, on macOS, changing the font size makes this show up
// as well, only in the opposite direction:
// https://github.com/processing/processing4/issues/194
// Bottom line, it's related to fractional metrics,
// and is especially bad with Source Code Pro.
// https://github.com/sampottinger/processing/issues/103
additionalOffset = -additionalOffset;
}
return Math.round(
Utilities.getTabbedTextWidth(s, metrics, x, expander, startOffset) + additionalOffset
);
/*
static float getFontCharWidth(char c, FontMetrics fm) {
return getFontCharsWidth(new char[] { c }, 0, 1, fm);
}
/**
* Get any partial widths applied within a font.
*
* <p>
* Get any partial widths applied within a font, caching results for the latest requested font
* (as identified via a FontMetrics object). Note that this is calculated for a sample character
* and is only valid for extrapolation in a monospaced font (that one might want to use in an
* IDE).
* </p>
*
* @param candidateMetrics The FontMetrics for which partial character pixel widths should be
* returned.
* @param x The x origin.
* @param expander The strategy for converting tabs into characters.
* @param startOffset The offset to apply before the text will be drawn.
* @return The partial width of a sample character within a font.
*/
private float getPartialPixelWidth(FontMetrics candidateMetrics, float x, TabExpander expander,
int startOffset) {
// See https://github.com/sampottinger/processing/issues/103
// Requires reference not object equality check
if (candidateMetrics != cachedPartialPixelWidthFont) {
float withFractional =
Utilities.getTabbedTextWidth(TEST_SEGMENT, candidateMetrics,
x, expander, startOffset);
int withoutFractional = (int) withFractional;
static final char[] spaceChar = new char[] { ' ' };
partialPixelWidth = withFractional - withoutFractional;
cachedPartialPixelWidthFont = candidateMetrics;
static float getFontCharsWidth(char[] data, int offset, int len,
FontMetrics fm) {
if (len == 0) {
return 0;
}
// doesn't seem to do anything fractional
float wi = fm.charsWidth(data, offset, len);
if (wi != ((int) wi)) {
System.out.println("extra: " + wi);
}
return partialPixelWidth;
int spaceWidth = fm.charsWidth(spaceChar, 0, 1);
//return fm.charsWidth(data, offset, len);
return len * spaceWidth;
}
*/
/**
* Hacked up version of the function with the same name from
* javax.swing.text.Utilities.
*
* In spite of being a fixed width font, Source Code Pro (the default
* font starting in Processing 3) returns slightly different widths
* depending on the number of characters shown. Using the getBounds()
* method on text won't even give us these metrics for individual
* characters, which returns a float but never with any fractional.
*
* This function forces the width of each character to stay the same,
* just as we're doing by drawing individual characters in the
* TextAreaPainter class.
*
* <a href="https://github.com/processing/processing4/issues/226">#226</a>,
* <a href="https://github.com/processing/processing4/issues/194">#194</a>,
* and <a href="https://github.com/sampottinger/processing/issues/103">Sam's 103</a>
*/
static int getTabbedTextWidth(Segment s,
FontMetrics metrics, int x,
TabExpander e, int startOffset) {
int nextX = x;
char[] txt = s.array;
int txtOffset = s.offset;
int n = s.offset + s.count;
int charCount = 0;
// int spaceAddon = 0;
int spaceWidth = metrics.charWidth(' ');
for (int i = txtOffset; i < n; i++) {
if (txt[i] == '\t') {
//nextX += metrics.charsWidth(txt, i-charCount, charCount);
nextX += charCount * spaceWidth;
charCount = 0;
if (txt[i] == '\t') {
if (e != null) {
nextX = (int) e.nextTabStop(nextX, startOffset + i - txtOffset);
} else {
// if no tab expander, just return the size of a space
//nextX += getFontCharWidth(' ', metrics);
nextX += spaceWidth;
}
} else if (txt[i] == ' ') {
//float spaceWidth = getFontCharWidth(' ', metrics);
//nextX += spaceWidth + spaceAddon;
nextX += spaceWidth;
}
} else if (txt[i] == '\n') {
// Ignore newlines, they take up space, and shouldn't be counted.
//nextX += getFontCharsWidth(txt, i - charCount, charCount, metrics);
nextX += charCount * spaceWidth;
// But this doesn't make any sense: why are we adding horizontally,
// shouldn't nextX be *reset* here? Guessing that segments never
// include a new line, so we never run into this. [fry 220129]
charCount = 0;
} else {
charCount++;
}
}
//nextX += getFontCharsWidth(txt, n - charCount, charCount, metrics);
nextX += charCount * spaceWidth;
// int amt = (int) (nextX - x);
// float spc = getFontCharWidth(' ', metrics);
// System.out.println(amt + " % " + spc + " = " + (amt % spc));
// return (int) (nextX - x); // nextX was a float, this was returning a float [fry 220128]
return nextX - x;
}
protected void setNewSelectionWord( int line, int offset )
{

8
app/src/processing/app/syntax/PdeTextAreaDefaults.java
View File

@@ -35,7 +35,7 @@ import processing.app.ui.Theme;
*/
public class PdeTextAreaDefaults extends TextAreaDefaults {
public PdeTextAreaDefaults(Mode ignoredMode) {
public PdeTextAreaDefaults() {
document = new SyntaxDocument();
// Set to 0 for revision 0215 because it causes strange jumps
@@ -55,6 +55,12 @@ public class PdeTextAreaDefaults extends TextAreaDefaults {
}
@Deprecated
public PdeTextAreaDefaults(Mode ignoredMode) {
this();
}
protected void updateTheme() {
fgcolor = Theme.getColor("editor.fgcolor");
bgcolor = Theme.getColor("editor.bgcolor");

13
todo.txt
View File

@@ -9,10 +9,11 @@ X but also requires checkout of processing-docs, so an unnecessary headache
X update themes with new token colors
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
scaling, text, again
_ Editor cursor position is offset to the right when display scaling >100%
_ https://github.com/processing/processing4/issues/226
_ Remove the need for “Disable HiDPI Scaling” on Windows
_ https://github.com/processing/processing4/issues/342
_ IDE cursor position is wrong if font size is changed in preferences on macOS
_ though at least one report that restarting the PDE doesn't fix the problem
_ probably related to second displays, need to hook one up and test
@@ -21,6 +22,9 @@ _ seems like the Windows workaround may be making this worse?
_ users confirms the correctly working display swapped between beta 3 and 4
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
_ Add support for localizing contributions
_ https://github.com/processing/processing4/pull/237 (updated by Andres)
X https://github.com/processing/processing/pull/2833
@@ -209,6 +213,7 @@ _ just skip the welcome screen on Windows hidpi dipslays?
_ better to launch a web browser
_ also use id/ask for login to help understand community?
_ tell users about JavaFX, Movie Maker install, changing themes
_ Remove “Disable HiDPI Scaling” preference on Windows?
would like to fix (after 4.0 final)