daniel543x/artcraft-services
1
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
170 Commits 58 Branches 0 Tags
5fc8f0610c50c830c3ef0d2fce87116a9547b888
Commit Graph
116 Commits
Author SHA1 Message Date
Brandon Thomas 31eb0ea2b7 Move old frontend UI code to new directory 2016-10-20 00:35:15 -04:00
Brandon Thomas c6c3084a07 Cleanup handler setup, static asset serving 2016-10-18 03:10:59 -04:00
Brandon Thomas beb8303884 Static HTML handler 
Maybe slightly wasted effort, but maps routes to static HTML files.
Will be good for doing several demos / concepts and hosting the old UI.
 2016-10-18 01:39:02 -04:00
Brandon Thomas 49547edf42 5-phones 2016-10-15 23:32:16 -04:00
Brandon Thomas 46961eb43d w/ with 2016-10-15 19:11:00 -04:00
Brandon Thomas a70151d1f1 add "o'clock" to times 2016-10-15 17:46:41 -04:00
Brandon Thomas a3ecab05f0 Time parsing 2016-10-15 17:32:56 -04:00
Brandon Thomas 493749a86d time tokenizing 2016-10-14 02:44:19 -04:00
Brandon Thomas 52f986ad0d Twitter username and hashtag parsing 2016-10-13 03:22:08 -04:00
Brandon Thomas 202c290e1f CamelCase parsing to word expansion 2016-10-13 01:23:15 -04:00
Brandon Thomas e00d8abe0e Parse hyphenated words into words if in dictionary 2016-10-13 00:55:16 -04:00
Brandon Thomas e8ab38fda0 Cleanup symbol pattern matching 2016-10-13 00:08:31 -04:00
Brandon Thomas 376b19fc06 Currency symbol tokenizing 2016-10-12 23:50:04 -04:00
Brandon Thomas 514656f296 additional test 2016-10-12 22:44:36 -04:00
Brandon Thomas 0c4a69ba4a CamelCase tokenizing 2016-10-12 22:42:41 -04:00
Brandon Thomas c0a284a1ac Tokenize (but do not parse) hyphenated words 2016-10-12 22:05:04 -04:00
Brandon Thomas fd45d9416a plus symbol 2016-10-12 21:41:09 -04:00
Brandon Thomas dbc776292a easier to read logs; dictionary tweaks 2016-10-12 02:04:44 -04:00
Brandon Thomas 90d91f731f ordinal bugfix 2016-10-12 00:24:49 -04:00
Brandon Thomas bb5d9aca1c Ordinal tokenization and rendering 2016-10-11 23:57:51 -04:00
Brandon Thomas c319b82462 add arpabet words, remove parens, etc. 2016-10-11 03:59:52 -04:00
Brandon Thomas f9fd557a27 log timing 2016-10-10 04:02:38 -04:00
Brandon Thomas 03a3e24d8b Handle percentages 2016-10-10 03:44:23 -04:00
Brandon Thomas 3f29bed4ad get rid of junk (square brackets, asterisk, etc.) 2016-10-10 03:31:37 -04:00
Brandon Thomas 28aee28c12 Number parsing 
Enables audio synthesis of numbers.
 2016-10-10 03:04:01 -04:00
Brandon Thomas 535829e791 Add TODO file to keep track of goals 2016-10-09 18:51:41 -04:00
Brandon Thomas 99f84354ad Add an abbreviations system 
- Abbreviations file maps abbreviations to a word list.

- This will function diffrently than initialisms, which are pronounced
  letter by letter.
 2016-10-09 18:40:07 -04:00
Brandon Thomas 9b8263815b Improved code and quality of tokenization 
- Distinct rounds of tokenization
- Split into modules
- Cleanup
 2016-10-08 14:28:11 -04:00
Brandon Thomas 1850dfe418 dashes and emoji tokenization, added tests 
- emoji tokenization is WIP
 2016-10-06 03:26:40 -04:00
Brandon Thomas 6524be41e1 Configurable logging contexts; refactored logging 
Logging from each component (parsing, synthesis, HTTP server) now logs
as its own target. Configuration options enable turning on or off each
component so it's easy to isolate a component for debugging.
 2016-10-06 02:46:29 -04:00
Brandon Thomas c37c07efc0 Update configs based on the system in tweetspeak 
configs have a default set and a hostname-specified set that act as
overrides.
 2016-10-06 01:27:09 -04:00
Brandon Thomas 343c7f7092 Actually implement a tokenizing pipeline 
Split "tokenizing" and "parsing". It isn't quite CS theory, but it makes
things easier to test and maintain.
 2016-10-06 00:59:00 -04:00
Brandon Thomas 5fc4592369 fix smart quote 2016-10-03 22:18:30 -04:00
Brandon Thomas ea0fb8efcc fix emdashes, smart quotes, periods w/ quotes 2016-10-02 17:29:42 -04:00
Brandon Thomas c7224d4838 Cleanup, fix compile warnings 2016-10-02 16:32:56 -04:00
Brandon Thomas f86f6aa3ea Foundations for sentence tokenization (WIP) 
- Sentence parsing and correction moved into a single module
- Typesafe "Speaker" parameter
- Regex + lazy_static!
 2016-10-02 15:28:48 -04:00
Brandon Thomas 7e302e7665 Word audio spacing params + handler cleanup 2016-09-20 05:19:16 -04:00
Brandon Thomas 4ce9909b4e Add work-related dictionary for fun 2016-09-20 04:27:06 -04:00
Brandon Thomas ec72a3c832 Rename "_words" directory to "words" 2016-09-18 06:22:22 -04:00
Brandon Thomas 6cb4dac63b Support - in arpabet. 2016-09-18 01:43:51 -04:00
Brandon Thomas 196744b976 Fix bug 2016-09-18 00:49:20 -04:00
Brandon Thomas 91d0ab9e67 Improved monophones wrt. ends 2016-09-16 02:17:40 -04:00
Brandon Thomas a5dcdca16e Custom arpabet dictionary 2016-09-16 02:08:47 -04:00
Brandon Thomas c80cf6715a Add start/end concept to the n-phone-based synth 2016-09-16 02:06:34 -04:00
Brandon Thomas 6e7afc677a Forgot to check this in awhile ago 2016-09-10 17:56:30 -04:00
Brandon Thomas 990faf3092 Add a WaveForm UI component (WaveSufer) 
- Drop buzz.js for wavesurfer.js
- Wrapper in TypeScript / Webpack instead of Javascript / Gulp
 2016-09-10 17:55:19 -04:00
Brandon Thomas 9004b204ed Fix n-phone bug 2016-09-10 12:40:57 -04:00
Brandon Thomas 23bd68c761 Allow apostrophe - fix in arpabet regex 2016-09-10 03:21:22 -04:00
Brandon Thomas c7b34146af cleanup, words 2016-09-08 05:32:30 -04:00
Brandon Thomas 93f3a2d372 Synthesize using n-phones! 
n-phones comprise monophones, diphones, 3-phones, and more.
This algorithm allows for smoother sounding synthesis.

Also updates the UI for a new n-phones option and adds/updates
scripts related to the arpabet with respect to n-phones.
 2016-09-08 04:36:49 -04:00