[gegl] operations/workshop/shadows-highlights-correction: Simplify code



commit 315a19696940cd9ace9f7914ec642eada208d8bc
Author: Debarshi Ray <debarshir gnome org>
Date:   Mon Dec 11 19:06:28 2017 +0100

    operations/workshop/shadows-highlights-correction: Simplify code
    
    It is hard to follow a maze of nested fminf and fmaxf calls. CLAMP is a
    more readable way to achieve the same. However, the values are already
    bounded by the limits imposed by the operation's properties. So,
    calling a function to re-impose the limit needlessly complicates
    things - wasteful at run-time and harder to read.
    
    It is better to use G_LIKELY annotated assertions instead.

 .../workshop/shadows-highlights-correction.c       |   35 +++++++++++++++----
 1 files changed, 27 insertions(+), 8 deletions(-)
---
diff --git a/operations/workshop/shadows-highlights-correction.c 
b/operations/workshop/shadows-highlights-correction.c
index ffac603..9bcf96a 100644
--- a/operations/workshop/shadows-highlights-correction.c
+++ b/operations/workshop/shadows-highlights-correction.c
@@ -82,16 +82,18 @@ process (GeglOperation       *operation,
   gfloat *dst  = out_buf;
   gfloat *aux  = aux_buf;
 
-  gfloat shadows    = 2.f * fminf (fmaxf (-1.0, ((gfloat) o->shadows / 100.f)), 1.f);
-  gfloat highlights = 2.f * fminf (fmaxf (-1.0, ((gfloat) o->highlights / 100.f)), 1.f);
-  gfloat whitepoint = fmaxf (1.f - (gfloat) o->whitepoint / 100.f, 0.01f);
-  gfloat compress   = fminf (fmaxf (0, ((gfloat) o->compress / 100.f)), 0.99f);
+  gfloat shadows;
+  gfloat shadows_100 = (gfloat) o->shadows / 100.0f;
+  gfloat highlights;
+  gfloat highlights_100 = (gfloat) o->highlights / 100.0f;
+  gfloat whitepoint = 1.0f - (gfloat) o->whitepoint / 100.0f;
+  gfloat compress;
 
-  gfloat shadows_ccorrect = (fminf (fmaxf (0.0f, ((gfloat) o->shadows_ccorrect / 100.f)), 1.f) - 0.5f)
-                             * SIGN(shadows) + 0.5f;
+  gfloat shadows_ccorrect;
+  gfloat shadows_ccorrect_100 = (gfloat) o->shadows_ccorrect / 100.0f;
 
-  gfloat highlights_ccorrect = (fminf (fmaxf (0.0f, ((gfloat) o->highlights_ccorrect / 100.f)), 1.f) - 0.5f)
-                                * SIGN(-highlights) + 0.5f;
+  gfloat highlights_ccorrect;
+  gfloat highlights_ccorrect_100 = (gfloat) o->highlights_ccorrect / 100.0f;
 
   gfloat max[4] = { 1.0f, 1.0f, 1.0f, 1.0f };
   gfloat min[4] = { 0.0f, -1.0f, -1.0f, 0.0f };
@@ -100,6 +102,23 @@ process (GeglOperation       *operation,
   gfloat doublemax = lmax * 2.0f;
   gfloat low_approximation = 0.01f;
 
+  compress = fminf ((gfloat) o->compress / 100.0f, 0.99f);
+  g_return_val_if_fail (compress >= 0.0f, FALSE);
+
+  g_return_val_if_fail (-1.0f <= highlights_100 && highlights_100 <= 1.0f, FALSE);
+  highlights = 2.0f * highlights_100;
+
+  g_return_val_if_fail (0.0f <= highlights_ccorrect_100 && highlights_ccorrect_100 <= 1.0f, FALSE);
+  highlights_ccorrect = (highlights_ccorrect_100 - 0.5f) * SIGN (-highlights) + 0.5f;
+
+  g_return_val_if_fail (-1.0f <= shadows_100 && shadows_100 <= 1.0f, FALSE);
+  shadows = 2.0f * shadows_100;
+
+  g_return_val_if_fail (0.0f <= shadows_ccorrect_100 && shadows_ccorrect_100 <= 1.0f, FALSE);
+  shadows_ccorrect = (shadows_ccorrect_100 - 0.5f) * SIGN (shadows) + 0.5f;
+
+  g_return_val_if_fail (whitepoint >= 0.01f, FALSE);
+
   if (!aux)
     {
       memcpy (out_buf, in_buf, sizeof (gfloat) * 4 * n_pixels);


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]