notice in std/socket.e the Socket Option enums
public enum
...
ESO_OOBINLINE,
--****
-- ==== Windows Socket Option enums
ESO_USELOOPBACK,
ESO_DONTLINGER,
ESO_OKLINGER
this renders as
2.3.7.19 ESO_USELOOPBACK
2.3.7.20 Windows Socket Option enums
2.3.7.21 ESO_DONTLINGER
...
previously eudoc would stop collecting constants all together when comments or markup other than {--** was between a list of constants.
now it seems like it wants to work to accept a new header inline, except it should have put the header after the current constant. then continue on collecting the public or export item.
one work around would require restarting the public constant list. this was used later on in socket.e
in this case of a list of public enum using the previous_enum+1 would become error prone and read more complex than it should. and in the current euphoria, this is not even possible
enum ESO_USELOOPBACK = ESO_OOBINLINE+1,
next=last+1
will error
<0117>:: Not expecting to see '+' here
ESO_USELOOPBACK = ESO_OOBINLINE+1,
workaround is to place the header one enum less than it should actually be.
public enum
ESO_PREVIOUS,
--****
-- !! actually ESO_OOBINLINE+1
-- ==== Option enums
ESO_OOBINLINE,
ESO_USELOOPBACK,
ESO_DONTLINGER,
ESO_OKLINGER
...
and BTW, do these enum even need to be public? won't someone try to use the enum directly instead of the actual socket option constant with anyones guess what kind of a result?
Deferred until 4.1.
changeset: 66:8fca3aab9b8c tag: tip user: Shawn Pringle <shawn.pringle@gmail.com> date: Tue Feb 03 13:31:10 2015 -0300 files: tests/res/s_enum.e tests/t_enum.e description:
changeset: 67:14cf81f7c528 parent: 64:e12d01792ad1 user: Shawn Pringle <shawn.pringle@gmail.com> date: Wed Feb 04 10:33:35 2015 -0300 files: euparser.e description:
changeset: 68:df22d5284e3a tag: tip user: Shawn Pringle <shawn.pringle@gmail.com> date: Tue Feb 03 13:31:10 2015 -0300 files: tests/res/s_enum.e tests/t_enum.e description: