On Tue, Jan 30, 2018 at 3:25 PM, Elijah Newren <new...@gmail.com> wrote:
> This populates a list of directory renames for us. The list of
> directory renames is not yet used, but will be in subsequent commits.
>
> Signed-off-by: Elijah Newren <new...@gmail.com>
> ---
> merge-recursive.c | 155
> ++++++++++++++++++++++++++++++++++++++++++++++++++++--
> 1 file changed, 152 insertions(+), 3 deletions(-)
>
> diff --git a/merge-recursive.c b/merge-recursive.c
> index 40ed8e1f39..c75d3a5139 100644
> --- a/merge-recursive.c
> +++ b/merge-recursive.c
> @@ -1393,6 +1393,138 @@ static struct diff_queue_struct *get_diffpairs(struct
> merge_options *o,
> return ret;
> }
>
> +static void get_renamed_dir_portion(const char *old_path, const char
> *new_path,
> + char **old_dir, char **new_dir)
> +{
> + char *end_of_old, *end_of_new;
> + int old_len, new_len;
> +
> + *old_dir = NULL;
> + *new_dir = NULL;
> +
> + /* For
comment style.
/*
* we prefer to keep the beginning
* and ending line of a comment free.
*/
/* unless you write single line comments */
> + * "a/b/c/d/foo.c" -> "a/b/something-else/d/foo.c"
> + * the "d/foo.c" part is the same, we just want to know that
> + * "a/b/c" was renamed to "a/b/something-else"
> + * so, for this example, this function returns "a/b/c" in
> + * *old_dir and "a/b/something-else" in *new_dir.
So we would not see multi-directory renames?
"a/b/c/d/foo.c" -> "a/b/something-else/e/foo.c"
could be detected as
"a/b/{c/d/ => something-else/e}/foo.c"
I assume this patch series is not bringing that to the table.
(which is fine, I am just wondering)
> + *
> + * Also, if the basename of the file changed, we don't care. We
> + * want to know which portion of the directory, if any, changed.
> + */
> + end_of_old = strrchr(old_path, '/');
> + end_of_new = strrchr(new_path, '/');
> +
> + if (end_of_old == NULL || end_of_new == NULL)
> + return;
return early as the files are in the top level, and apparently we do
not support top level renaming?
What about a commit like 81b50f3ce4 (Move 'builtin-*' into
a 'builtin/' subdirectory, 2010-02-22) ?
Well that specific commit left many files outside the new builtin dir,
but conceptually we could see a directory rename of
/* => /src/*
> + while (*--end_of_new == *--end_of_old &&
> + end_of_old != old_path &&
> + end_of_new != new_path)
> + ; /* Do nothing; all in the while loop */
We have to compare manually as we'd want to find
the first non-equal and there doesn't seem to be a good
library function for that.
Assuming many repos are UTF8 (including in their paths),
how does this work with display characters longer than one char?
It should be fine as we cut at the slash?
> + /*
> + * We've found the first non-matching character in the directory
> + * paths. That means the current directory we were comparing
> + * represents the rename. Move end_of_old and end_of_new back
> + * to the full directory name.
> + */
> + if (*end_of_old == '/')
> + end_of_old++;
> + if (*end_of_old != '/')
> + end_of_new++;
> + end_of_old = strchr(end_of_old, '/');
> + end_of_new = strchr(end_of_new, '/');
> +
> + /*
> + * It may have been the case that old_path and new_path were the same
> + * directory all along. Don't claim a rename if they're the same.
> + */
> + old_len = end_of_old - old_path;
> + new_len = end_of_new - new_path;
> +
> + if (old_len != new_len || strncmp(old_path, new_path, old_len)) {
How often are we going to see this string-is-equal case?
Would it make sense to do that first in the function?
> + *old_dir = xstrndup(old_path, old_len);
> + *new_dir = xstrndup(new_path, new_len);
> + }
> +}
> +
> +static struct hashmap *get_directory_renames(struct diff_queue_struct *pairs,
> + struct tree *tree)
> +{
> + struct hashmap *dir_renames;
> + struct hashmap_iter iter;
> + struct dir_rename_entry *entry;
> + int i;
> +
> + dir_renames = malloc(sizeof(struct hashmap));
xmalloc
> + dir_rename_init(dir_renames);
> + for (i = 0; i < pairs->nr; ++i) {
> + struct string_list_item *item;
> + int *count;
> + struct diff_filepair *pair = pairs->queue[i];
> + char *old_dir, *new_dir;
> +
> + /* File not part of directory rename if it wasn't renamed */
> + if (pair->status != 'R')
> + continue;
> +
> + get_renamed_dir_portion(pair->one->path, pair->two->path,
> + &old_dir, &new_dir);
> + if (!old_dir)
> + /* Directory didn't change at all; ignore this one. */
> + continue;
So the first loop is about counting the number of files in each directory
that are renamed and the later while loop is about mapping them?
> + /* Strings were xstrndup'ed before inserting into string-list,
> + * so ask string_list to remove the entries for us.
> + */
comment style.
> + entry->possible_new_dirs.strdup_strings = 1;
Why do we need to set strdup_strings here (so late in the
game, we are about to clear it?) Could we set it earlier?
Or rather have the string list duplicate the strings instead of
get_renamed_dir_portion ?
