Module xwaves_mark
source code
Read in .in files produced by ESPS xmark. Returns (header, data),
where data is (time, word_or_phoneme, type), where type is 0 for words, 1
for phonemes. Times are guarenteed to be increasing inside the set of all
words, and also inside the set of all phonemes. Word marks preceed the
corresponding phoneme marks.
|
|
|
|
|
mark_to_lab(data,
ty)
This function lets you take a mixed list of word and phone labels,
such as provided by xwaves_mark.read(), and will select out one or
the other type. |
source code
|
|
|
|
|
DecrTol = 0.004999
|
|
PHONE = 1
|
|
WORD = 0
|
|
__doc__ = read.__doc__
|
|
__package__ = ' gmisclib '
|
Imports:
os,
sys,
die,
types,
DataError,
NoSuchFileError,
DataOutOfOrderError,
BadFileFormatError,
Error
Read in .in files produced by ESPS xmark. Returns (header, data),
where data is (time, word_or_phoneme, type), where type is 0 for words, 1
for phonemes. Times are guarenteed to be increasing inside the set of all
words, and also inside the set of all phonemes. Word marks preceed the
corresponding phoneme marks.
|
This function lets you take a mixed list of word and phone labels,
such as provided by xwaves_mark.read(), and will select out one or the
other type. Type is PHONE or WORD.
|
combine_2labs(whdr,
wd,
phdr,
pd,
utterance,
TOL=0.001)
| source code
|
Combine two XLAB files into one XMARK file. It forces (within rounding
errors) the words to enclose the phones.
|