[PATCH] [GSoC 2014]diff: Imported dir.h and renamed read_directory()

[PATCH] [GSoC 2014]diff: Imported dir.h and renamed read_directory()

[Brian Bourn]

this was done in order to implement the GSoC
Micro project for diff-no-index.c this is the
first patch importing dir.h, for the use of
is_dot_or_dotdot(), and renaming
read_directory() to read_directory_contents()
in order to deal with the conflicting function
in dir.h

Signed-off-by: Brian Bourn <bab2177@columbia.edu>
---

I plan on applying to GSoC 2014

 diff-no-index.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/diff-no-index.c b/diff-no-index.c
index 8e10bff..ba915af 100644
--- a/diff-no-index.c
+++ b/diff-no-index.c
@@ -10,13 +10,14 @@
 #include "blob.h"
 #include "tag.h"
 #include "diff.h"
+#include "dir.h"
 #include "diffcore.h"
 #include "revision.h"
 #include "log-tree.h"
 #include "builtin.h"
 #include "string-list.h"

-static int read_directory(const char *path, struct string_list *list)
+static int read_directory_contents(const char *path, struct string_list *list)
 {
  DIR *dir;
  struct dirent *e;
@@ -107,9 +108,9 @@ static int queue_diff(struct diff_options *o,
  int i1, i2, ret = 0;
  size_t len1 = 0, len2 = 0;

- if (name1 && read_directory(name1, &p1))
+ if (name1 && read_directory_contents(name1, &p1))
  return -1;
- if (name2 && read_directory(name2, &p2)) {
+ if (name2 && read_directory_contents(name2, &p2)) {
  string_list_clear(&p1, 0);
  return -1;
  }
-- 
1.9.0

Re: [PATCH] [GSoC 2014]diff: Imported dir.h and renamed read_directory()

[Eric Sunshine]

Thanks for the submission. Comments below to give you a taste of the
Git review process...

On Tue, Mar 18, 2014 at 8:35 PM, Brian Bourn <ba.bourn@gmail.com> wrote:
> Subject: diff: Imported dir.h and renamed read_directory()

Use imperative voice: "import dir.h and rename read_directory()"

> this was done in order to implement the GSoC
> Micro project for diff-no-index.c

This commentary won't be relevant to anyone reading the commit message
months or years from now. Place it below the "---" line following your
sign-off.

> this is the
> first patch importing dir.h, for the use of
> is_dot_or_dotdot(), and renaming
> read_directory() to read_directory_contents()
> in order to deal with the conflicting function
> in dir.h

Since this preparatory change and the one which will follow are so
closely related, they should be submitted together as a two-patch
series.

The commit message itself it somewhat rambling. It would be clearer if
you could explain the problem precisely, then the solution. Perhaps
something like this:

    It would be desirable to replace manual checking of "." or ".."
    in diff-no-index.c with is_dot_or_dotdot(), which is defined in
    dir.h, however, dir.h declares a read_directory() which conflicts
    with a (different) static read_directory() defined in
    diff-no-index.c. As a preparatory step, rename the local
    read_directory() to avoid the collision.

> Signed-off-by: Brian Bourn <bab2177@columbia.edu>
> ---
>
> I plan on applying to GSoC 2014
>
>  diff-no-index.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/diff-no-index.c b/diff-no-index.c
> index 8e10bff..ba915af 100644
> --- a/diff-no-index.c
> +++ b/diff-no-index.c
> @@ -10,13 +10,14 @@
>  #include "blob.h"
>  #include "tag.h"
>  #include "diff.h"
> +#include "dir.h"
>  #include "diffcore.h"

dir.h is not needed for this patch, so including it here only confuses
matters. Include it instead when it is actually required by the
follow-up patch which uses is_dot_or_dotdot().

(Was it intentional to place the new #include between diff.h and
diffcore.h? Just being curious; it's not particularly important.)

Other than this, the patch looks sane.

>  #include "revision.h"
>  #include "log-tree.h"
>  #include "builtin.h"
>  #include "string-list.h"
>
> -static int read_directory(const char *path, struct string_list *list)
> +static int read_directory_contents(const char *path, struct string_list *list)
>  {
>   DIR *dir;
>   struct dirent *e;
> @@ -107,9 +108,9 @@ static int queue_diff(struct diff_options *o,
>   int i1, i2, ret = 0;
>   size_t len1 = 0, len2 = 0;
>
> - if (name1 && read_directory(name1, &p1))
> + if (name1 && read_directory_contents(name1, &p1))
>   return -1;
> - if (name2 && read_directory(name2, &p2)) {
> + if (name2 && read_directory_contents(name2, &p2)) {
>   string_list_clear(&p1, 0);
>   return -1;
>   }
> --
> 1.9.0

[PATCH 2/2][GSoC 2014] diff: used is_dot_or_dotdot() in code

[babourn]

in accordance with the GSoC Microproject implemented
the call is_dot_or_dotdot() in the code in order to
further universalize the call to the function and
increase code continuity.

Signed-off-by: Brian Bourn <bab2177@columbia.edu>
---
 diff-no-index.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/diff-no-index.c b/diff-no-index.c
index ba915af..44cce25 100644
--- a/diff-no-index.c
+++ b/diff-no-index.c
@@ -26,7 +26,7 @@ static int read_directory_contents(const char *path,
struct string_list *list)
  return error("Could not open directory %s", path);

  while ((e = readdir(dir)))
- if (strcmp(".", e->d_name) && strcmp("..", e->d_name))
+ if (!is_dot_or_dotdot(e->d_name))
  string_list_insert(list, e->d_name);

  closedir(dir);
--
1.9.0



--
View this message in context: http://git.661346.n2.nabble.com/PATCH-GSoC-2014-diff-Imported-dir-h-and-renamed-read-directory-tp7605950p7605956.html
Sent from the git mailing list archive at Nabble.com.

Re: [PATCH 2/2][GSoC 2014] diff: used is_dot_or_dotdot() in code

[Eric Sunshine]

On Tue, Mar 18, 2014 at 9:30 PM, babourn <ba.bourn@gmail.com> wrote:
> Subject: diff: used is_dot_or_dotdot() in code

Use imperative voice: "use" rather than "used"

> in accordance with the GSoC Microproject implemented

This commentary will not have much meaning to someone reading the
commit log months or years from now. Place it below the "---" line
following your sign-off.

> the call is_dot_or_dotdot() in the code in order to
> further universalize the call to the function and
> increase code continuity.

It should be sufficient to explain the patch by just saying:

    Subject: replace manual "."/".." check with is_dot_or_dotdot()

The rest of the explanatory text can be dropped since it doesn't add
anything (meaningful) beyond what the subject says.

> Signed-off-by: Brian Bourn <bab2177@columbia.edu>
> ---
>  diff-no-index.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/diff-no-index.c b/diff-no-index.c
> index ba915af..44cce25 100644
> --- a/diff-no-index.c
> +++ b/diff-no-index.c
> @@ -26,7 +26,7 @@ static int read_directory_contents(const char *path,
> struct string_list *list)
>   return error("Could not open directory %s", path);
>
>   while ((e = readdir(dir)))
> - if (strcmp(".", e->d_name) && strcmp("..", e->d_name))
> + if (!is_dot_or_dotdot(e->d_name))

The patch is severely whitespace-damaged. (Did you post it through Nabble?)

>   string_list_insert(list, e->d_name);
>
>   closedir(dir);
> --
> 1.9.0
>
>
>
> --
> View this message in context: http://git.661346.n2.nabble.com/PATCH-GSoC-2014-diff-Imported-dir-h-and-renamed-read-directory-tp7605950p7605956.html
> Sent from the git mailing list archive at Nabble.com.
> --

Re: [PATCH 2/2][GSoC 2014] diff: used is_dot_or_dotdot() in code

[Eric Sunshine]

On Tue, Mar 18, 2014 at 11:58 PM, Brian Bourn <ba.bourn@gmail.com> wrote:
> On Tue, Mar 18, 2014 at 11:45 PM, Eric Sunshine <sunshine@sunshineco.com>
> wrote:
>> On Tue, Mar 18, 2014 at 9:30 PM, babourn <ba.bourn@gmail.com> wrote:
>> > Subject: diff: used is_dot_or_dotdot() in code
>> > Signed-off-by: Brian Bourn <bab2177@columbia.edu>
>> > ---
>> >  diff-no-index.c | 2 +-
>> >  1 file changed, 1 insertion(+), 1 deletion(-)
>> >
>> > diff --git a/diff-no-index.c b/diff-no-index.c
>> > index ba915af..44cce25 100644
>> > --- a/diff-no-index.c
>> > +++ b/diff-no-index.c
>> > @@ -26,7 +26,7 @@ static int read_directory_contents(const char *path,
>> > struct string_list *list)
>> >   return error("Could not open directory %s", path);
>> >
>> >   while ((e = readdir(dir)))
>> > - if (strcmp(".", e->d_name) && strcmp("..", e->d_name))
>> > + if (!is_dot_or_dotdot(e->d_name))
>>
>> The patch is severely whitespace-damaged. (Did you post it through
>> Nabble?)
>
>    I did post through Nabble, My email with the patch didn't seem to be
> going through.
>    should I keep trying to resend it through email to undo the whitspace
> damage?

It's probably not necessary to try resending this version of the patch
since you'll (hopefully) be sending a newer version which takes
reviewer comments into consideration.

What method are you using to send the patches? git send-email?
Something other? This particular mailing list rejects HTML-formatted
messages, so that could be the culprit if you pasted the patch into
your email client. It's a good idea to try sending patches to yourself
via "git send-email". If you can get that to work successfully, then
they should be accepted by the mailing list when sent via the same
mechanism.