diff options
| author | Robin Watts <robin.watts@artifex.com> | 2017-09-25 19:22:49 +0100 |
|---|---|---|
| committer | Robin Watts <robin.watts@artifex.com> | 2017-10-24 15:16:37 +0100 |
| commit | 79bc9c182e74bcd58c6de33be8dfeb384a07199b (patch) | |
| tree | 6e5155ac4558a785e6a483804b4e16675ae066a7 /source | |
| parent | c8376f966de15dac704666cc7446366328ea8737 (diff) | |
| download | mupdf-79bc9c182e74bcd58c6de33be8dfeb384a07199b.tar.xz | |
Fix separations object handling when cloning for spots.
The 'producing spots' case is poorly tested, but I believe we
should be keeping the spot information internally, rather than
NULLing it out.
Diffstat (limited to 'source')
| -rw-r--r-- | source/fitz/separation.c | 7 |
1 files changed, 5 insertions, 2 deletions
diff --git a/source/fitz/separation.c b/source/fitz/separation.c index 5f8fee41..f4f5e262 100644 --- a/source/fitz/separation.c +++ b/source/fitz/separation.c @@ -195,6 +195,8 @@ fz_separations *fz_clone_separations_for_overprint(fz_context *ctx, fz_separatio return NULL; n = sep->num_separations; + if (n == 0) + return NULL; c = 0; for (i = 0; i < n; i++) { @@ -203,9 +205,10 @@ fz_separations *fz_clone_separations_for_overprint(fz_context *ctx, fz_separatio c++; } - /* If no composites, then we are fine to render direct. */ + /* If no composites, then we don't need to create a new seps object + * with the composite ones enabled, so just reuse our current object. */ if (c == 0) - return NULL; + return fz_keep_separations(ctx, sep); /* We need to clone us a separation structure, with all * the composite separations marked as enabled. */ |