Re: [NEW] audio/ncspot: ncurses Spotify client written in Rust

On Tue, Nov 26, 2019 at 10:27:58AM +0000, Stuart Henderson wrote:
> On 2019/11/26 09:08, Sebastien Marie wrote:
> > On Mon, Nov 25, 2019 at 10:06:22PM +0000, Stuart Henderson wrote:
> > > +cc semarie@, wondering if you have any suggestions?
> >
> > CONFIGURE_STYLE = cargo
> >
> > is missing. so cargo isn't configured properly.
>
> Thank you.
>
> Would it be OK to add one or both parts of this?
>
> - mention CONFIGURE_STYLE in a comment so it can be found more easily,
> - allow modcargo-gen-crates-licenses to continue with a missing license
>
> Index: cargo.port.mk
> ===================================================================
> RCS file: /cvs/ports/devel/cargo/cargo.port.mk,v
> retrieving revision 1.9
> diff -u -p -r1.9 cargo.port.mk
> --- cargo.port.mk 14 Sep 2019 17:34:46 -0000 1.9
> +++ cargo.port.mk 26 Nov 2019 10:25:22 -0000
> @@ -68,7 +68,7 @@ MODCARGO_post-patch += \
> .endfor
>
> # configure hook. Place a config file for overriding crates-io index by
> -# local source directory.
> +# local source directory. Enabled by use of "CONFIGURE_STYLE=cargo".
> MODCARGO_configure = \
> mkdir -p ${WRKDIR}/.cargo; \
> \

ok for this one.

> @@ -219,5 +219,5 @@ modcargo-gen-crates: extract
> modcargo-gen-crates-licenses: configure
> .for _cratename _cratever in ${MODCARGO_CRATES}
> @echo -n "MODCARGO_CRATES += ${_cratename} ${_cratever} # "
> - @sed -ne 's/^license.*= *"\([^"]*\)".*/\1/p' "${MODCARGO_VENDOR_DIR}/${_cratename}-${_cratever}/Cargo.toml"
> + @sed -ne 's/^license.*= *"\([^"]*\)".*/\1/p' "${MODCARGO_VENDOR_DIR}/${_cratename}-${_cratever}/Cargo.toml" 2>/dev/null || echo '(XXX missing)'

sed will not fail if license marker is absent, but if the Cargo.toml file is absent/unreadable.

currently, libpulse-sys crate has no license marker (but has a LICENSE file): so
sed output is nothing, and it results the MODCARGO_CRATES line to collapse with
the next one.

> .endfor
>
>
> > I will take look to cargo.devel.mk. cargo has, since few releases, a new
> > --offline flag, which should avoid the need to use pf(4) to catch network
> > problem (cargo will just abort without trying network).
>
> That would be helpful.

yes. I just need to check it doesn't introduce regression with current ports
using devel/cargo module.


--
Sebastien Marie