these changes have been down for Persian (Farsi) Language voice fa, fa-en-us, fa-latn
the bug fixed, this line has been modified to 
fa_list file, Line 200: _)	paRAntezbaste:
for solving "twice speaking right prentice" issue.

these line has been added to fa_rules, .replace section. (line: 26) these new rules help to replace unknown characters to standard Persian UTF8 characters in process .
Line 47: ٫	،
Line 109: ﺾ	ض
Line 116: ﻆ	ظ
Line 141: ﻙ	ک
Line 184: ۆ	ؤ

this rule has been modified to 
fa_rules file, .group ی section, Line 5639: L09L09) یوم (_Sm3	ijom
for improving pronunciation

Improved spelling of wide/narrow e/ē, o,ō and uo.
Improved intonation of syllables.
More unstressed words are added.

added definition for  phoneme l in ph_italian

updated dictionaries, march 2016

This can occur in cases where the voice is not found, or
espeak_ng_Initialize is called more than once.

This reverts commit 6d9b9c4db6.

This is handled by calls to espeak_ng_Cancel/espeak_Cancel and
espeak_ng_Terminate/espeak_Terminate.

Removing these calls do not impact on the behaviour of the program
and simplify the wave.h API.

These calls are used to pre-emptively exit when writing audio
data. This is also handled within dispatch_audio. The difference
is that this callback could allow espeak to cancel quicker. In
practice it does not make much difference.

This call was used to update the written sample count used by
wave_get_remaining_time. As that function has been removed, this
code is no longer needed.

This logic was implemented to limit the number of audio events sent
to a device, but the logic is not too relevant in practice (these
calls return 0 most of the time), and the logic is broken (e.g. when
calling Synthesize with a very long block of text).