how to make Festival czech voice machac to run in archLinux

[Mgr. Janusz Chmiel]

Because I run Arch Linux compatible with Termux, Proot and Android, there
are no packages like for Arch Linux for X86 or X64 CPUS. As A result, I must
build myself from sources available on Github.
Festival voice for English, which is available on Aur repo for aarch64 work
even with speech-dispatcher.
I simply run
festival --server &
and then I can use speech-dispatcher with Festival speech synthesizer.

But Czech language is specific.
There are festival-freebsoft-utils
festival/czech
and brailcom/voice-czech-ph: Czech diphone database 
But how to configure everything to work properly?
FOrtunately, Festivla work perfectly by using Pulseaudio. No crashes no bad
sound output.
So I want to configure Czech voice because it will be much more easier than
compiling Epos for aarch64 platform.
Because before compiling, some C experts would had to change many lines of
source code to work with newest GCC. Because simply compiling Epos is not
enough. There are some segmentation faults errors when simply compiling it
for newest libraryes and GCC.