Using clustering model for the text width estimation #594
Conversation
|
|
||
| private fun getFaceCoefficient(font: Font): Double { | ||
| return when { | ||
| font.isBold && font.isItalic -> 0.9172120995070999 |
There was a problem hiding this comment.
I don't understand how bold-italic style can have coefficient 0.9 and just italic - only 0.18?
Why "normal" coefficient is near 0.0?
There was a problem hiding this comment.
It is an additive coefficient.
So, coeff(italic) + coeff(bold) is approximately equal to coeff(bold-italic).
And coeff(normal) should be zero, but it is just almost zero because of calculations (face coefficient calculated after the family coefficient).
There was a problem hiding this comment.
Do you think a correction by e-17 makes any difference?
| private const val SIZE_COEFFICIENT = 0.9843304096547842 | ||
|
|
||
| private fun getCharWidth(char: Char): Double { | ||
| val clusterId = CLUSTERS.indexOfFirst { char in it } |
There was a problem hiding this comment.
Let use a hash map for the symbol lookup.
Init the hash map using elements of the CLUSTERS list.
| private const val DEFAULT_FAMILY = "Lucida Grande" | ||
| private const val DEFAULT_FONT_SIZE = 14 | ||
|
|
||
| private val CLUSTERS = listOf( |
There was a problem hiding this comment.
Let's add a reference on GitHub: https://github.com/ASmirnov-HORIS/text-width-estimation
There was a problem hiding this comment.
Now I added it, but a bit higher.
| @@ -12,6 +12,7 @@ class FontFamily private constructor(private val myName: String) { | |||
| } | |||
|
|
|||
| companion object { | |||
| val DEFAULT_FONT_FAMILY = forName("\"Lucida Grande\", sans-serif") | |||
| val MONOSPACED = forName("monospace") | |||
There was a problem hiding this comment.
Curious, does "monospace" family work. Would be nice if it does.
There was a problem hiding this comment.
It work.
But this particular variable is never used.
No description provided.