Re: [PATCH 3/8] merge-ort: record the reason that we want a rename for a directory
- From
Elijah Newren <newren@gmail.com>
- Date
- Mar 15, 2021, 15:27 UTC
- Message-ID
- <CABPp-BHLL_u_E-Ss=AYWUZM+_9hbC6D6WzRjxEUwmexTEgQ2zw@mail.gmail.com>
- In-Reply-To
- <baca2700-0959-c4ca-6ba9-ceb5188f96f4@gmail.com>
On Mon, Mar 15, 2021 at 7:31 AM Derrick Stolee <stolee@gmail.com> wrote:
Show 42 quoted lines
>
> On 3/13/2021 5:22 PM, Elijah Newren via GitGitGadget wrote:
> > From: Elijah Newren <newren@gmail.com>
> >
> > When one side of history renames a directory, and the other side of
> > history added files to the old directory, directory rename detection is
> > used to warn about the location of the added files so the user can
> > move them to the old directory or keep them with the new one.
> >
> > This sets up three different types of directories:
> > * directories that had new files added to them
> > * directories underneath a directory that had new files added to them
> > * directories where no new files were added to it or any leading path
> >
> > Save this information in dirs_removed; the next several commits will
> > make use of this information.
> ...
> > +/* dir_rename_relevance: the reason we want rename information for a dir */
> > +enum dir_rename_relevance {
> > + NOT_RELEVANT = 0,
> > + RELEVANT_FOR_ANCESTOR = 1,
> > + RELEVANT_FOR_SELF = 2
> > +};
>
> Is this potentially a flag list? It's hard to tell because we don't
> have another item (3 or 4?).
>
> > unsigned sides = (0x07 - dirmask)/2;
> > + unsigned relevance = (renames->dir_rename_mask == 0x07) ?
> > + RELEVANT_FOR_ANCESTOR : NOT_RELEVANT;
> > + /*
> > + * Record relevance of this directory. However, note that
> > + * when collect_merge_info_callback() recurses into this
> > + * directory and calls collect_rename_info() on paths
> > + * within that directory, if we find a path that was added
> > + * to this directory on the other side of history, we will
> > + * upgrade this value to RELEVANT_FOR_SELF; see below.
> > + */
>
> This comment seems to imply that RELEVANT_FOR_SELF is "more important"
> than RELEVANT_FOR_ANCESTOR, so the value will just be changed (not a
> flag).Yes.
Show 20 quoted lines
> > + /*
> > + * Here's the block that potentially upgrades to RELEVANT_FOR_SELF.
> > + * When we run across a file added to a directory. In such a case,
> > + * find the directory of the file and upgrade its relevance.
> > + */
> > + if (renames->dir_rename_mask == 0x07 &&
> > + (filemask == 2 || filemask == 4)) {
> > + /*
> > + * Need directory rename for parent directory on other side
> > + * of history from added file. Thus
> > + * side = (~filemask & 0x06) >> 1
> > + * or
> > + * side = 3 - (filemask/2).
> > + */
> > + unsigned side = 3 - (filemask >> 1);
> > + strintmap_set(&renames->dirs_removed[side], dirname,
> > + RELEVANT_FOR_SELF);
>
> Yes, using "RELEVANT_FOR_SELF" here, not "relevance | RELEVANT_FOR_SELF".
> OK. This should make the later consumers simpler.Yep, indeed. Would it make it clearer earlier if I were to stop assigning the explicit values in the enum? Would adding a comment when I introduce the enum be easier? Or was it just "thinking out loud"?