[report] 7/21

0 views
Skip to first unread message

Jim Shargo

unread,
Jul 21, 2010, 4:01:21 PM7/21/10
to Edwin
Today we:
* Fixed a nasty bug in search.scm
* Tested the newly-working line-start, line-end functions in motion

Tomorrow:
* Write some more tests for motion.scm
* ...?

Where should we go next? Comtabs, keystrokes, or "Edwin-Scheme"?

Thanks!
0001-Fixed-argument-order-bug-in-search.scm.patch
0002-Added-line-start-end-tests-to-scratch-test-groups.sc.patch

Duncan Mak

unread,
Jul 21, 2010, 4:12:48 PM7/21/10
to edwin-...@googlegroups.com
On Wed, Jul 21, 2010 at 4:01 PM, Jim Shargo <mnu...@gmail.com> wrote:
Today we:
* Fixed a nasty bug in search.scm
  1. The bulk of 0001 has to do with indentation - which is fine, but could you spin out a separate patch so that I can review the argument order changes separate from the indentation changes?
  2. The change seems to be the order of FIND-NEXT, which changed from (start end char) to (char start end) -- why did that happen? Where is FIND-NEXT defined? Is it an incompatibility between MIT Scheme string library and SRFI-13?
  3. A note - in TEST-LINE-START+END, you used #t as the last case in the COND expression. I think that's preferred in (Common) LISP, but in Scheme, the 'ELSE' keyword is available to you. 
-- 
Duncan.

Jim Shargo

unread,
Jul 21, 2010, 7:49:13 PM7/21/10
to edwin-...@googlegroups.com
On 7/21/10 4:12 PM, Duncan Mak wrote:
  1. The bulk of 0001 has to do with indentation - which is fine, but could you spin out a separate patch so that I can review the argument order changes separate from the indentation changes?
Any quick-and-dirty git magic to doing this? Or should I git revert HEAD^^ search.scm  and not untabify the file?

  1. The change seems to be the order of FIND-NEXT, which changed from (start end char) to (char start end) -- why did that happen? Where is FIND-NEXT defined? Is it an incompatibility between MIT Scheme string library and SRFI-13?
FIND-NEXT is a function passed to make-find-(previous|next) in search.scm. The only functions passed to it are string-index, string-index-ci and their right-moving counter-parts. As seen here-- http://goo.gl/8RKy --the contract for the functions is str char start end. We did a git blame and chalked it up to a typo.

  1. A note - in TEST-LINE-START+END, you used #t as the last case in the COND expression. I think that's preferred in (Common) LISP, but in Scheme, the 'ELSE' keyword is available to you.
Ooops! Will be fixed first thing tomorrow.

-- Jim Shargo

Ryan Schwers

unread,
Jul 22, 2010, 11:04:51 AM7/22/10
to edwin-...@googlegroups.com
Here are the patches re-done. There's a new one that untabifies search.scm.

~Ryan Schwers

0003-Untabified-search.scm.patch
0004-Fixed-an-argument-order-bug-in-search.scm.patch
0005-Added-line-start-end-tests-to-test-groups.scm.patch
Reply all
Reply to author
Forward
0 new messages