From: Jehan <jehan@girinstud.io>
Date: Tue, 14 Feb 2023 09:50:20 +0000 (+0100)
Subject: Issue #85: Troubles opening 16-bit floating point grayscale files.
X-Git-Tag: archive/raspbian/1%0.1.106-3+rpi1^2~15^2^2~28
X-Git-Url: https://dgit.raspbian.org/?a=commitdiff_plain;h=722b7851b1df793152f2280cd17b2672d3b731b6;p=babl.git

Issue #85: Troubles opening 16-bit floating point grayscale files.

All current LUT conversion and creation clearly assume that we are
operating on 1 byte per channel (the flag we use right now to determine
if we may have a LUT is even called `is_u8_color_conv`). So let's
enforce this.

Otherwise we might end up with problems such as in issue #85, where we
convert from or to "YA u16", which is still under 4bpp but not 1bpc. As
a consequence, the first byte of the alpha only was copied and the rest
(including the second alpha byte) was used as belonging to the color
components.

Of course, we can update this code later to also support such cases
(taking into account the source and target alpha size), but for now,
let's simply exclude these.
---

diff --git a/babl/babl-fish-path.c b/babl/babl-fish-path.c
index a4b002a..26072d5 100644
--- a/babl/babl-fish-path.c
+++ b/babl/babl-fish-path.c
@@ -958,8 +958,12 @@ _babl_fish_prepare_bpp (Babl *babl)
      }
 
   {
-  int source_bpp = babl->fish_path.source_bpp;
-  int dest_bpp = babl->fish_path.dest_bpp;
+  int         source_bpp  = babl->fish_path.source_bpp;
+  int         dest_bpp    = babl->fish_path.dest_bpp;
+  const Babl *source_type = babl_format_get_type (babl_source,
+                                                  babl_format_get_n_components (babl_source) - 1);
+  const Babl *dest_type   = babl_format_get_type (babl_dest,
+                                                  babl_format_get_n_components (babl_dest) - 1);
   if (//source->format.space != destination->format.space &&
      (
         (source_bpp == 2 && dest_bpp == 16)
@@ -977,12 +981,16 @@ _babl_fish_prepare_bpp (Babl *babl)
      // as long as the highest 8bit of the 32bit of a 4 byte input is ignored
      // (alpha) - and it is not an associated color model. A 24 bit LUT provides
      // exact data. Thus this is valid for instance for "YA half"
-     // Additionally we also avoid associated destination models too.
 
-     if ((babl->conversion.source->format.type[0]->bits < 32 &&
+     if ((babl->conversion.source->format.type[0]->bits < 32                                  &&
           (source_bpp < 4 ||
-           (babl->conversion.source->format.model->flags & BABL_MODEL_FLAG_ASSOCIATED)==0) &&
-         (babl->conversion.destination->format.model->flags & BABL_MODEL_FLAG_ASSOCIATED)==0))
+           (babl->conversion.source->format.model->flags & BABL_MODEL_FLAG_ASSOCIATED)==0)    &&
+         (babl->conversion.destination->format.model->flags & BABL_MODEL_FLAG_ASSOCIATED)==0) &&
+         /* Checking the exact type to be u8, so that we don't generate LUT for
+          * "YA u16" (4 bpp) or similar formats.
+          */
+         source_type == babl_type ("u8")                                                      &&
+         dest_type   == babl_type ("u8"))
      {
        static int measured_timings = 0;
        float scaling = 10.0;