Commit Graph
1225 Commits
Author SHA1 Message Date
Reece H. Dunn bb051e6086 ky: don't speak numbers character by character 2017-02-14 22:48:04 +00:00
Reece H. Dunn c4ca7a1202 Add wchar.h and wctype.h compatibility shims using ucd-tools. This is applied to all platforms to provide consistent behaviour. 2017-02-13 19:49:30 +00:00
Reece H. Dunn 1378a81a06 Remove the wchar tables from readclause, now the standard is* and to* APIs are being used. 2017-02-13 19:48:57 +00:00
Reece H. Dunn 5975f07095 Replace the *2 wchar/wctype APIs with their standard versions. These will be fixed using compatibility headers. 2017-02-13 07:43:23 +00:00
Reece H. Dunn d2a919bc7b Merge commit '6346ede604270202b2aad8159cb914b7a2276c34' 2017-02-12 21:32:06 +00:00
Reece H. Dunn 6346ede604 isalnum: use the Alphabetic derived core property to check that the character is in the alpha class. 2017-02-12 21:22:01 +00:00
Reece H. Dunn cbddf092c9 isalpha: use the Alphabetic derived core property to check that the character is in the alpha class. 2017-02-12 21:20:03 +00:00
Reece H. Dunn b38659582f islower: include the Other_Lowercase characters. 2017-02-12 17:28:22 +00:00
Reece H. Dunn 1fc20f39f5 isupper: include the Other_Uppercase characters. 2017-02-12 14:51:24 +00:00
Valdis Vitolins 17168a5669 More proper fix for issue #196 2017-02-12 15:56:27 +02:00
Reece H. Dunn eb08dc422d Update the copyright dates. 2017-02-12 12:24:18 +00:00
Reece H. Dunn c9f2940373 isblank: don't include <noBreak> characters, and add tests for this API. 2017-02-12 11:47:30 +00:00
Reece H. Dunn f109bb918f isspace: don't include <noBreak> characters. 2017-02-12 11:11:52 +00:00
Reece H. Dunn bd71fed013 ctype: return true in isupper/islower if there is a simple case mapping present 2017-02-12 09:22:25 +00:00
Reece H. Dunn fe3e3350e4 Merge remote-tracking branch 'ValdisVitolins/master' 2017-02-11 22:25:15 +00:00
Valdis Vitolins fd4b397ad7 Merge branch 'test' 2017-02-11 21:49:02 +02:00
Valdis Vitolins 9558e60e4b Merge branch 'master' of https://github.com/espeak-ng/espeak-ng 2017-02-11 21:48:04 +02:00
Valdis Vitolins 872433d62c Fix for Jump rule to allow no letters in between current and pre/post (literal letter/letter group) 2017-02-11 21:42:09 +02:00
Valdis Vitolins 41f29f3e87 Fix for issue #196 Word end mark _ doesn't work properly with ~ character group 2017-02-11 20:52:52 +02:00
Reece H. Dunn 01852ee90d isdigit: C99 compatibility -- only check [0-9]. 2017-02-11 18:20:08 +00:00
Reece H. Dunn 34167771fd Add an iswxdigit compatibility API. 2017-02-11 18:18:38 +00:00
Reece H. Dunn 3e321158e7 Update the documentation on the is* APIs. 2017-02-11 17:54:25 +00:00
Reece H. Dunn d585c77804 Add an iswblank compatibility API. 2017-02-11 17:37:42 +00:00
Reece H. Dunn c80829a5fd Remove the NEED_WCHAR_FUNCTIONS code for wchar/wctype compatibility. 2017-02-11 08:39:29 +00:00
Reece H. Dunn 71c010b712 Remove unused wchar.h and wctype.h includes. 2017-02-10 08:03:40 +00:00
Reece H. Dunn 9ecde56231 Build the ucd-tools files with libespeak. 2017-02-10 07:45:43 +00:00
Valdis Vitolins af7fc1b657 Issue #199 Jump statement now supports letter groups 2017-02-05 23:31:24 +02:00
Reece H. Dunn 36a2872b4b Don't allow the gender property to be set in language files (they are only relevant in voice files). 2017-02-05 19:56:07 +00:00
Reece H. Dunn 1c5bbe31a4 Use LookupMnem in ReadVoiceFile when parsing lines. 2017-02-05 19:55:48 +00:00
Reece H. Dunn d30cf6f0d7 Default voices/languages to be male to match the default voice. 2017-02-05 19:08:27 +00:00
Reece H. Dunn f6147085cd Use enumeration constants for the voice gender field values. 2017-02-05 19:06:27 +00:00
Reece H. Dunn 46cd5f48e1 Don't allow 'gender unknown' in voice/language files. 2017-02-05 18:47:31 +00:00
Reece H. Dunn ea365af43c Remove the 'dictdialect' feature. 2017-02-05 17:03:44 +00:00
Valdis Vitolins 0c3ceb2e70 As secondary stress is still spelled differently it is disabled 2017-02-05 15:41:26 +00:00
Valdis Vitolins b2057635c4 PRE jump rule e.g. 'xyJ)' implemented 2017-02-05 15:40:28 +00:00
Valdis Vitolins 29f0b673ee Updated comments for POST jump rule 2017-02-05 15:40:14 +00:00
Valdis Vitolins 6bd656cf4a Limit search for jump POST rule to null byte for last word in sentence 2017-02-05 15:40:00 +00:00
Reece H. Dunn 6eaf1d2ddc Look in espeak-ng-data/lang for voices.
This is to support splitting language specification (dictionaries)
from voice specification (phon*).
2017-02-04 16:50:57 +00:00
Reece H. Dunn bb7630e4c1 Make len_path_voices a parameter to GetVoices. 2017-02-04 16:44:12 +00:00
Reece H. Dunn 022e4e82dd Don't check if the MBROLA voice data is present on Windows.
With the original eSpeak Windows port, the MBROLA voices had to
be present in an mbrola folder of the espeak-data. This will be
different in espeak-ng, allowing the MBROLA voice files to be
located elsewhere on the system.
2017-02-04 16:34:09 +00:00
Reece H. Dunn 3ce7fab7db Return the actual error from GetFileLength instead of 0. 2017-02-04 16:03:56 +00:00
Reece H. Dunn df6a2228b7 Use -EISDIR instead of -2 in GetFileLength for directories. 2017-02-04 16:02:56 +00:00
chrislm d757fd9743 IT: Reading roman numerals as ordinals. 2017-02-04 11:53:43 +01:00
chrislm 59c9408376 IT: Removed automatic secondari stress
IT: Unstressed final syllable is diminished

IT: Preserve unstressed monosyllable words and allow secondary stress in multisillable words indicated as unstressed in it_list.
2017-02-04 11:43:44 +01:00
Reece H. Dunn 1a0594c62f Fix reading stdin buffers larger than 1000.
Patch by https://github.com/ChliHug.
2017-02-02 07:22:59 +00:00
Reece H. Dunn 901bc0d1ba Allow 'apr liquid' to be used together in a phoneme. 2017-01-30 20:27:10 +00:00
Reece H. Dunn 9448834298 Use '_' to indicate no parent phoneme table in the 'phonemetable' command. 2017-01-30 19:14:00 +00:00
Reece H. Dunn 574ecbd634 Inline the InstnPlusPhoneme function. 2017-01-29 10:53:33 +00:00
Reece H. Dunn e7374059e5 Use the stress condition values with i_CHANGE_IF. 2017-01-29 10:43:02 +00:00
Reece H. Dunn f2bb482782 Use OPCODE for the RETURN and CONTINUE opcodes. 2017-01-29 10:29:48 +00:00