unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
From: Stephen Leake <stephen_leake@stephe-leake.org>
To: Juanma Barranquero <lekktu@gmail.com>
Cc: 21816@debbugs.gnu.org
Subject: bug#21816: elisp-mode-tests fails on a case-preserving filesystem
Date: Tue, 03 Nov 2015 03:17:42 -0600	[thread overview]
Message-ID: <86mvuv30i1.fsf@stephe-leake.org> (raw)
In-Reply-To: <CAAeL0SQLncO69=TuhRwWjh0OX5+HJnpcxwOccQLOu-hdO4uX+A@mail.gmail.com> (Juanma Barranquero's message of "Mon, 2 Nov 2015 17:41:58 +0100")

Juanma Barranquero <lekktu@gmail.com> writes:

> Package: emacs
> Version: 25.0.50
> Severity: minor
> X-Debbugs-CC: stephen_leake@stephe-leake.org
>
>
> Test xref-elisp-test-find-defs-feature-el fails on my Windows system
> because of a case-sensitive filename comparison.
>
> Basically
>
> ELISP> (aref (aref (car (elisp--xref-find-definitions 'xref)) 2) 3)
> "c:/devel/emacs/repo/trunk/lisp/progmodes/xref.el"
>
> ELISP> (aref (xref-make-elisp-location
>      'xref 'feature
>      (expand-file-name "../../lisp/progmodes/xref.el"
> emacs-test-dir))
>     3)
> "c:/Devel/emacs/repo/trunk/lisp/progmodes/xref.el"
>
> and the test fails at the d/D in c:/Devel vs. c:/devel (as seen in the log
> below).
>
> The value returned by `expand-file-name' is the real name as stored in the
> filesystem (the directory is called "Devel"). 

This is not caused by expand-file-name, but by the values passed to it.
In this case, `emacs-test-dir' has the "c:/Devel"; it is set by:

(defconst emacs-test-dir (file-name-directory (or load-file-name (buffer-file-name))))

I can reproduce the problem by adding a "downcase" to this; my emacs
test dir is in c:/Projects/..., but my load-path has the correct case.

> The value from
> `elisp--xref-find-definitions' ultimately comes from `locate-file', and
> reflects what's stored in `load-path' (which can be set by the user, so we
> cannot force it to have "correct" casing).

Well, since this is only a problem in a test run by developers (not
general users), we _could_ insist on that, but I agree it would be nice
not to.

> The test code in elisp-mode-tests.el uses `equal' to compare the xref
> structures. I can think a couple of ways to fix the test, but they are all
> quite ugly....

The canonical solution for case-insensitive string comparisons is to use
`downcase' on both values before using `equal'. This works for me:

(defun xref-elisp-test-run (xrefs expected-xrefs)
  (should (= (length xrefs) (length expected-xrefs)))
  (while xrefs
    (let* ((xref (pop xrefs))
           (expected (pop expected-xrefs))
           (expected-xref (or (when (consp expected) (car expected)) expected))
           (expected-source (when (consp expected) (cdr expected))))

      ;; Downcase the filename for case-insensitive file systems.
      (setf (xref-elisp-location-file (oref xref :location))
            (downcase (xref-elisp-location-file (oref xref :location))))

      (setf (xref-elisp-location-file (oref expected-xref :location))
            (downcase (xref-elisp-location-file (oref expected-xref :location))))

      (should (equal xref expected-xref))

      (xref--goto-location (xref-item-location xref))
      (back-to-indentation)
      (should (looking-at (or expected-source
                              (xref-elisp-test-descr-to-target expected)))))
    ))

together with:

;; We add 'downcase' here to deliberately cause a potential problem on
;; case-insensitive file systems. On such systems, `load-file-name'
;; may not have the same case as the real file system, since the user
;; can set `load-path' to have the wrong case. This is handled in
;; `xref-elisp-test-run'.
(defconst emacs-test-dir (downcase (file-name-directory (or load-file-name (buffer-file-name)))))


If no one objects to this, I'll commit it.

-- 
-- Stephe





  parent reply	other threads:[~2015-11-03  9:17 UTC|newest]

Thread overview: 67+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-02 16:41 bug#21816: elisp-mode-tests fails on a case-preserving filesystem Juanma Barranquero
2015-11-02 17:57 ` Eli Zaretskii
2015-11-02 18:20   ` Juanma Barranquero
2015-11-03  1:19     ` Juanma Barranquero
2015-11-03  3:42       ` Eli Zaretskii
2015-11-03  9:17 ` Stephen Leake [this message]
2015-11-03 10:06   ` Juanma Barranquero
2015-11-03 14:51     ` Stephen Leake
2015-11-03 14:58       ` Juanma Barranquero
2015-11-03 19:43         ` Stephen Leake
2015-11-03 22:24           ` Stephen Leake
2015-11-04  8:48           ` Juanma Barranquero
2015-11-09  3:38           ` Dmitry Gutov
2019-10-02  0:57             ` Juanma Barranquero
2015-11-03 15:40       ` Eli Zaretskii
2015-11-03 15:45         ` Juanma Barranquero
2015-11-03 15:52           ` Dmitry Gutov
2015-11-03 16:04             ` Juanma Barranquero
2015-11-03 16:18               ` Dmitry Gutov
2015-11-03 16:33                 ` Juanma Barranquero
2015-11-03 18:12                   ` Dmitry Gutov
2015-11-04  8:40                     ` Juanma Barranquero
2015-11-03 16:34                 ` Eli Zaretskii
2015-11-03 16:42                   ` Juanma Barranquero
2015-11-03 16:53                     ` Eli Zaretskii
2015-11-03 18:13                       ` Dmitry Gutov
2015-11-03 19:54                         ` Random832
2015-11-03 20:03                           ` Dmitry Gutov
2015-11-03 20:25                             ` Random832
2015-11-03 21:25                               ` Dmitry Gutov
2015-11-03 20:51                         ` Eli Zaretskii
2015-11-04  8:19                       ` Juanma Barranquero
2015-11-03 18:18                   ` Dmitry Gutov
2015-11-03 20:56                     ` Eli Zaretskii
2015-11-04  8:53                       ` Juanma Barranquero
2015-11-03 19:50                   ` Random832
2015-11-03 21:02                     ` Eli Zaretskii
2015-11-03 21:17                       ` Random832
2015-11-03 21:22                         ` Eli Zaretskii
2015-11-03 19:54                 ` Stephen Leake
2015-11-03 20:05                   ` Dmitry Gutov
2015-11-03 21:07                     ` Eli Zaretskii
2015-11-03 21:52                     ` Stephen Leake
2015-11-03 22:07                       ` Stephen Leake
2015-11-03 22:18                         ` Dmitry Gutov
2015-11-03 22:20                         ` Stephen Leake
2015-11-04  3:46                         ` Eli Zaretskii
2015-11-04  8:28                           ` Stephen Leake
2015-11-04  5:24                         ` Alexis
2015-11-04  6:33                           ` Random832
2015-11-04  6:52                             ` Alexis
2015-11-04 15:35                               ` Eli Zaretskii
2015-11-04 15:38                             ` Eli Zaretskii
2015-11-04  9:04                           ` Stephen Leake
2015-11-04 10:37                             ` Alexis
2015-11-04 16:08                             ` Eli Zaretskii
2015-11-04 15:34                           ` Eli Zaretskii
2015-11-03 21:05                   ` Eli Zaretskii
2015-11-03 21:58                     ` Stephen Leake
2015-11-04  3:42                       ` Eli Zaretskii
2015-11-04 10:00                         ` Stephen Leake
2015-11-04 19:14                           ` Stephen Leake
2015-11-03 19:24         ` Stephen Leake
2015-11-03 21:00           ` Eli Zaretskii
2015-11-04 15:29     ` Stephen Leake
2015-11-05 12:26       ` Juanma Barranquero
2015-11-05 13:23         ` Stephen Leake

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.gnu.org/software/emacs/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=86mvuv30i1.fsf@stephe-leake.org \
    --to=stephen_leake@stephe-leake.org \
    --cc=21816@debbugs.gnu.org \
    --cc=lekktu@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).