Attachment 47761 Details for Bug 38079 – Bugfix (workaround?)

[patch] Bugfix (workaround?)

0001-path-bounder-Fix-empty-extents-computation.patch (text/plain), 3.17 KB, created by Andrea Canciani on 2011-06-09 01:54:06 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Andrea Canciani

Created: 2011-06-09 01:54:06 UTC

Size: 3.17 KB

patch

obsolete

>From f2c0f544d5d8b0e63e22e657dbccd881c181b25b Mon Sep 17 00:00:00 2001
>From: Andrea Canciani <ranma42@nirvana-fedora.(none)>
>Date: Thu, 9 Jun 2011 10:39:58 +0200
>Subject: [PATCH] path-bounder: Fix empty extents computation
>
>Fix long-standing bug in computation of path extents.
>
>Fixes get-path-extents, (test-from-38079).
>
>See https://bugs.freedesktop.org/show_bug.cgi?id=38079
>---
> KNOWN_ISSUES            |    6 ------
> src/cairo-path-bounds.c |   12 ++++++++++++
> test/get-path-extents.c |    6 ++----
> 3 files changed, 14 insertions(+), 10 deletions(-)
>
>diff --git a/KNOWN_ISSUES b/KNOWN_ISSUES
>index 5fda683..65fbf12 100644
>--- a/KNOWN_ISSUES
>+++ b/KNOWN_ISSUES
>@@ -2,9 +2,3 @@ There are a few known bugs in 1.10 that have been fixed in master, but
> appear to be non-trivial to backport without fear of causing other
> regressions. The impact of these bugs is considered to be less than of a
> risk than rewriting the code.
>-
>-Zero Path Extents
>------------------
>-A closed degenerate path is reported as having extents (0, 0) x (0, 0),
>-whereas the expected value is (x, y) x (0, 0). This regression has existed
>-since at least 1.2.
>diff --git a/src/cairo-path-bounds.c b/src/cairo-path-bounds.c
>index 2e7b6bd..61d2f69 100644
>--- a/src/cairo-path-bounds.c
>+++ b/src/cairo-path-bounds.c
>@@ -114,6 +114,11 @@ _cairo_path_bounder_curve_to (void *closure,
> {
>     cairo_path_bounder_t *bounder = closure;
> 
>+    if (bounder->has_initial_point) {
>+	_cairo_path_bounder_add_point (bounder, &bounder->current_point);
>+	bounder->has_initial_point = FALSE;
>+    }
>+
>     /* If the bbox of the control points is entirely inside, then we
>      * do not need to further evaluate the spline.
>      */
>@@ -140,6 +145,13 @@ _cairo_path_bounder_curve_to (void *closure,
> static cairo_status_t
> _cairo_path_bounder_close_path (void *closure)
> {
>+    cairo_path_bounder_t *bounder = closure;
>+
>+    if (bounder->has_initial_point) {
>+	_cairo_path_bounder_add_point (bounder, &bounder->current_point);
>+	bounder->has_initial_point = FALSE;
>+    }
>+
>     return CAIRO_STATUS_SUCCESS;
> }
> 
>diff --git a/test/get-path-extents.c b/test/get-path-extents.c
>index 3a05eff..5bd3eb7 100644
>--- a/test/get-path-extents.c
>+++ b/test/get-path-extents.c
>@@ -148,16 +148,14 @@ draw (cairo_t *cr, int width, int height)
>     phase = "Degenerate arc (R=0)";
>     errors += !check_extents (ctx, phase, cr2, FILL, EQUALS, 0, 0, 0, 0);
>     errors += !check_extents (ctx, phase, cr2, STROKE, EQUALS, 0, 0, 0, 0);
>-    /*XXX: I'd expect these extents to be oriented at 200, 400 */
>-    errors += !check_extents (ctx, phase, cr2, PATH, EQUALS, 0, 0, 0, 0);
>+    errors += !check_extents (ctx, phase, cr2, PATH, EQUALS, 200, 400, 0, 0);
> 
>     cairo_new_path (cr2);
>     cairo_arc (cr2, 200, 400, 10., 0, 0);
>     phase = "Degenerate arc (Î=0)";
>     errors += !check_extents (ctx, phase, cr2, FILL, EQUALS, 0, 0, 0, 0);
>     errors += !check_extents (ctx, phase, cr2, STROKE, EQUALS, 0, 0, 0, 0);
>-    /*XXX: I'd expect these extents to be oriented at 200, 400 */
>-    errors += !check_extents (ctx, phase, cr2, PATH, EQUALS, 0, 0, 0, 0);
>+    errors += !check_extents (ctx, phase, cr2, PATH, EQUALS, 210, 400, 0, 0);
> 
>     cairo_new_path (cr2);
>     cairo_restore (cr2);
>-- 
>1.7.5.2
>

Actions: View | Splinter Review

Attachments on bug 38079: 47720 | 47761

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.