• A few edge cases where some rare orthographic sequences were being rejected
• Some nuclei not being properly parsed when onset =
• Bug where syllables with both on and offglides were being phonetized without onglide

• -8 option to output Chinese-style tone classes
• "o" dialect option which produces a spelling pronunication (best used in combination with -8 -n -g -p flags)

• Bug where -n flag was not respected for some items

• Major re-write of script logic - now generates internal G2P representation as basis for all output types
• As a result, G2P 'rules' mapping greatly simplified
• Changed default tone encodings; options to force strict lowercasing and suppression of glottals
• Added continuous integration testing
• Both surface and underlying outputs now follow Pham (2006)

syllables.txt test file

• Transcription of <ch> (and, for Northern, <tr>) to /tɕ/
• Transcription of <b> and <d> to reflect canonical preglottalized/implosive pronunciation
• Transcriptions of Central tones
• Internally, <oo> and <ôô> are now represented as long vowels, but they are then shortened before output

• Errors in mappings involving <e> and <ê>

• Python 3 support. To use vPhon with Python 2, see the v1.0.0 release.

• Since optparse is deprecated, upgraded to work with argparse.
• the -d, --dialect flag is no longer obligatory: vPhon will default to Northen dialect.
• Changed behavior of -o, --ortho flag which outputs Viet orthography along with IPA transcription. Like the -m, --delimit flag, this option now requires a delimiter.
• the -8, --cao option now returns sắc and nặng in checked syllables as 7 and 8, respectively, instead of 5b and 6b.

• Final Python 2 version -- legacy release

• Added -o, --ortho flag to output Viet orthography along with IPA transcription.
• Added -m, --delimit flag to have output separated by a user-specified delimiter.

• Output now uses Python 3-style print function
• Supplying a filename as a command-line argument no longer works: vPhon now reads only from STDIN and writes only to STDOUT. Any text processing should be done in usual Unix fashion (i.e. using shell redirection, cat, etc.).

• Fixed bug that was treating gì, gĩ... and qùy, qũy... as non-Viet characters.
• Fixed bug that was treating checked syllables with nặng as 21g instead of 21 in Northern and Southern dialects.
• Fixed bug that was treating sonorant-final syllables with sắc as 45 instead of 24 in Northern and Central dialects.

• Added -t, --tokenize flag to preserve underscores or hyphens in tokenized inputs, so that e.g. anh_ta is output as anh1_ta1. This flag has the effect of not automatically treating inputs with hyphens as non-Viet words.