In various places, we use the DOM to measure text like at
|
export function measureFontHeight(font: string): number { |
or we explicitly lay out text like at
|
// Single word. Applying text wrap on word by splitting |
|
// it into characters and fitting the maximum number of |
|
// characters possible per line (box width). |
|
if (wordsInColumn.length === 0) { |
|
let curLineTextWidth = gc.measureText(textInCurrentLine).width; |
|
while (curLineTextWidth > boxWidth && textInCurrentLine !== '') { |
|
// Iterating from the end of the string until we find a |
|
// substring (0,i) which has a width less than the box width. |
|
for (let i = textInCurrentLine.length; i > 0; i--) { |
|
const curSubString = textInCurrentLine.substring(0, i); |
|
const curSubStringWidth = gc.measureText(curSubString).width; |
|
if (curSubStringWidth < boxWidth || curSubString.length === 1) { |
|
// Found a substring which has a width less than the current |
|
// box width. Rendering that substring on the current line |
|
// and setting the remainder of the parent string as the next |
|
// string to iterate on for the next line. |
|
const nextLineText = textInCurrentLine.substring( |
|
i, |
|
textInCurrentLine.length |
|
); |
|
textInCurrentLine = nextLineText; |
|
curLineTextWidth = gc.measureText(textInCurrentLine).width; |
|
gc.fillText(curSubString, textX, curY); |
|
curY += textHeight; |
|
// No need to continue iterating after we identified |
|
// an index to break the string on. |
|
break; |
|
} |
|
} |
|
} |
Pretext (HN discussion) is a new library that measures and lays out text efficiently, so it may be faster or more accurate. This issue is to explore using pretext instead.
In various places, we use the DOM to measure text like at
lumino/packages/datagrid/src/textrenderer.ts
Line 944 in d4b1494
or we explicitly lay out text like at
lumino/packages/datagrid/src/textrenderer.ts
Lines 253 to 282 in d4b1494
Pretext (HN discussion) is a new library that measures and lays out text efficiently, so it may be faster or more accurate. This issue is to explore using pretext instead.