CHICKEN Scheme

Metadata

Id: 0480e82d53a24cadb0956a9bf379a7bc551e094d
Trac id: 1135
Type: defect
Reporter: mario
Owner:
Cc:
Status: closed
Component: core libraries
Estimated difficulty:
Resolution: fixed
Priority: major
Milestone: 4.10.0
Version: 4.9.x
Changetime: 2014-06-30 21:32:33 UTC
Created: 2014-06-27 16:51:02 UTC
Keywords: string-copy!, srfi-13

Attachments

0480e82d53a24cadb0956a9bf379a7bc551e094d/attachments/0001-C_substring_copy-use-C_memmove-instead-of-C_memcpy.patch
0480e82d53a24cadb0956a9bf379a7bc551e094d/attachments/0001-C_substring_copy-use-C_memmove-instead-of-C_memcpy-w.patch

Description

While using the gap-buffer egg (implementation ported from Guile), I noticed a weird behavior in certain cases (see http://lists.gnu.org/archive/html/guile-devel/2014-06/msg00060.html)

Looks like the culprit is string-copy! (probably C_substring_copy), as it was in Guile's case (as pointed by Mark H Weaver).

A test case from Guile's test suite illustrates the problem:

 (let ((str (string-copy "abcde")))
   (string-copy! str 1 str 0 3) str)
 => "aaaae"
 
 Expected: "aabce"

Changes and comments

[2014-06-27 20:54:33 UTC] mario attached 0001-C_substring_copy-use-C_memmove-instead-of-C_memcpy-w.patch (description=#f)

[2014-06-27 20:55:39 UTC] mario wrote:

The attached patch Works For Me, although I don't know if it is the optimal/correct solution.

[2014-06-27 23:03:56 UTC] mario attached 0001-C_substring_copy-use-C_memmove-instead-of-C_memcpy.patch (description=#f)

[2014-06-27 23:04:12 UTC] mario wrote:

Seth Alves pointed out on #chicken that checking for string equality is pointless, since memmove will do that anyway. So, attached is a new patch which just replaces C_memcpy by C_memmove in C_substring_copy.

[2014-06-30 21:32:33 UTC] sjamaan changed status from new to closed

[2014-06-30 21:32:33 UTC] sjamaan set resolution to fixed

[2014-06-30 21:32:33 UTC] sjamaan wrote:

Fixed by 6b36695d94e0bd977e0d85d48438f621128e1101

Summary