[QUESTION] ref-filter: how to support atom like %an

[QUESTION] ref-filter: how to support atom like %an

[ZheNing Hu]

My current thinking is this:

First add a new member `short_name` in `struct valid_atom`,
Then we can share a valid_atom item for the long name of
an atom such as `%(authoremail)` and the short name of
the atom `%ae` (%aE is not considered at the time).
Something like:

[ATOM_AUTHOREMAIL] = { "ae", "authoremail", SOURCE_OBJ, FIELD_STR,
person_email_atom_parser },

Then we need to do some work on atoms parsing:
In the original design of `parse_ref_filter_atom()`, it relied on
searching for `%(` and `)` to find an atom. Now we have to
change the way of finding atoms: Currently I just replace
`find_next()` to find a single `%`. If there is `(` after `%`,
we can keep the original long name atoms parsing method,
otherwise we are dealing with the short name atom,
we need to pass the interval between the current `%`
and the next single `%` to `parse_ref_filter_atom()`.
Then we are in `parse_ref_filter_atom()`, we need to
consider these situations:

1. %(authoremail:trim)
2. %(authoremail)
3. %ae
4. %ae:trim
5. %aeabc:trim
6. %ae:trimabcd...

As we can see, the situation is starting to get complicated.
The more critical point is as in case 6: We cannot determine
the end of a short name atom because it does not have an
explicit `)` end.
In this way, many of the `*_atom_parser` functions in
`ref-filter.c` will not work, because most of them use `strcmp()`
to check the attribute of an atom. If we want to support the
short name `%ae`, that would require a large-scale rewrite.

Any good ideas, friends?

--
ZheNing Hu

Re: [QUESTION] ref-filter: how to support atom like %an

[Junio C Hamano]

ZheNing Hu <adlternative@gmail.com> writes:

> My current thinking is this:
>
> First add a new member `short_name` in `struct valid_atom`,
> Then we can share a valid_atom item for the long name of
> an atom such as `%(authoremail)` and the short name of
> the atom `%ae` (%aE is not considered at the time).
> Something like:
>
> [ATOM_AUTHOREMAIL] = { "ae", "authoremail", SOURCE_OBJ, FIELD_STR,
> person_email_atom_parser },

My assumption has always been that when enriching "--format" of the
"log" family (e.g. "status", "show") with ingredients from the
"for-each-ref" family (e.g. "branch", "tag"), we would adopt the
notation the latter uses, i.e. "git log --format='%s %(notes)'",
with the full intention that eventually users would choose between
the synonyms to make it "git log --format='%(subject) %(notes)'" for
consistency, without wasting brain cycles to come up with and to
remember two-letter shorthands that will not scale.

Re: [QUESTION] ref-filter: how to support atom like %an

[ZheNing Hu]

Junio C Hamano <gitster@pobox.com> 于2021年5月16日周日 上午11:55写道：
>
> ZheNing Hu <adlternative@gmail.com> writes:
>
> > My current thinking is this:
> >
> > First add a new member `short_name` in `struct valid_atom`,
> > Then we can share a valid_atom item for the long name of
> > an atom such as `%(authoremail)` and the short name of
> > the atom `%ae` (%aE is not considered at the time).
> > Something like:
> >
> > [ATOM_AUTHOREMAIL] = { "ae", "authoremail", SOURCE_OBJ, FIELD_STR,
> > person_email_atom_parser },
>
> My assumption has always been that when enriching "--format" of the
> "log" family (e.g. "status", "show") with ingredients from the
> "for-each-ref" family (e.g. "branch", "tag"), we would adopt the
> notation the latter uses, i.e. "git log --format='%s %(notes)'",
> with the full intention that eventually users would choose between
> the synonyms to make it "git log --format='%(subject) %(notes)'" for
> consistency, without wasting brain cycles to come up with and to
> remember two-letter shorthands that will not scale.
>

Well, it’s truly more useful to provide long name atoms for the "log" family,
But if we need to establish a unified interface for --format, shouldn't
we extend it base on `ref-filter`?

At present, the scalability of the implementation in `ref-filter` is better,
but it does not support two-letter shorthands atoms...

--
ZheNing Hu