When using serverstart("ip.ad.d.r:") to listen on a random port, we need
to abide by getaddrinfo()'s API and pass in a NULL service, rather than
an empty string.
When given an empty string, getaddrinfo() is free to search for a
service by the given name (since the string isn't a number) which will
fail. At least FreeBSD does perform this lookup.
This change implicitly adds IPv6 support.
If the address contains ":", we try to use a TCP socket instead of a Unix domain
socket. Everything in front of the last occurrence of ":" is the hostname and
everything after it the port.
If the hostname lookup fails, we fall back to using a Unix domain socket.
If the port is empty ("localhost:"), a random port will be assigned.
Examples:
NVIM_LISTEN_ADDRESS=localhost:12345 -> TCP (IPv4 or IPv6), port: 12345
NVIM_LISTEN_ADDRESS=localhost: -> TCP (IPv4 or IPv6), port: random (> 1024)
NVIM_LISTEN_ADDRESS=localhost:0 -> TCP (IPv4 or IPv6), port: random (> 1024)
NVIM_LISTEN_ADDRESS=localhost -> Unix domain socket "localhost" in current dir
`blocksize` was checked against UINT_MAX after it was checked against
MAX_SWAP_PAGE_SIZE which makes it always pass the check. Better use
STATIC_ASSERT instead.
AFAIK there is no way NULL can be there, including from the line it points to.
Dunno what analyser was thinking, but dereferencing of `argv[0]` happened just
before `get_number_arg()` call: in `ascii_isdigit()` two lines above. And `idx`
cannot possibly be NULL ever, it comes from `&varname`, this could not ever give
anything, but a valid pointer.
Could not be uninitialized because `func_or_func_caller_profiling` is true only
if `do_profiling` is `YES`, and if `do_profiling` is `YES` then
`script_prof_save()` was called to initialize the variable.
Based on comments it appears that some non-printable characters intended to be
shown as `|x` (0xA0..0xFE) and some as `~x` (0x80..0x9F, 0xFF, excluding
previous). But this never happens because this is being catched by condition `c
>= 0x80` above which makes them be represented as `<A0>`. Since I find this
variant more useful and it additionally is backwards compatible (Vim does the
same thing) I just dropped dead branches.
Compile `nvim` executable against Lua if PREFER_LUA=ON.
As the testing library `nvim-test` requires LuaJIT, it is
still compiled against LuaJIT. If LuaJIT is not available,
`nvim-test` is not built.
When the end character in a range matches a different standard range
(e.g., [0-z]), the range would be incorrectly detected as the class of
the end character (CLASS_az).
Instead of using a fallthrough, immediately FAIL when the end character
doesn't match the expected range.
