forked from Nick3C/pinyin-toolkit
-
Notifications
You must be signed in to change notification settings - Fork 14
Changelog
batterseapower edited this page Sep 13, 2010
·
2 revisions
Max Bolingbroke | Website | |
Nick Cook | Website |
- Big performance and memory-usage improvements
- This was obtained by moving to a sqlite based backend using cjklib
- Automatic reformatting of text in the “Reading” field to match your color and tone-indication preferences
- Automatic introduction of audio corresponding to Pinyin entered in the “Audio” field
- Generation of decent readings for Western numbers, decimals. dates, percentages and fractions
- Generation of decent meanings for Chinese numbers, decimals. dates, percentages and fractions
- Automatic filling of ‘Traditional’ and ‘Simplified’ fields on facts with the appropriate characters
- Audio generation for a noun’s measure words (using a random quantity or 几) in a field called something like “MW Audio”
- Meaning emphasis: you can make all but the first meaning from the dictionary small automatically
- Can now guess field purpose from a wider variety of field names
- In the meaning field, mask the pinyin from CEDICT that appears when a character is a common surname
- Lots of bug fixes, notably to:
- Ubuntu support
- Tone sandhi
- The configuration dialog
- … and lots more small issues
Max Bolingbroke | Website | |
Nick Cook | Website |
- Large-scale re-write and optimisation of the code by Max Bolingbroke (many thanks!)
- ported KanjiGraph to Hanzigraph
- Automatic translation of non-dictionary words & phrase (can be used for almost any language)
- Add limited support for new CFDICT (French)
- All distributions will now include all dictionaries for simplicity. Unwanted dictionaries can be deleted.
- MW with pinyin added automatically to your deck if in dictionary (if enabled) only applies to English version
- Dictionary definition and measure word have their simplified/traditional variant selected according
to user preferences - Don’t generate audio tags if sounds are missing
- Use a fifth tone audio sample if one is provided, otherwise switch to 4th tone
- Try several file extensions when we need a bit of audio, using the order: .mp3, .ogg, .wav
- Optional entry-number indicators in translations such as “(1)” “(2)” “(3)”
- Internal refactoring of code to remove the incidence of unreliable string manipulations, leading to bugfixes:
- Remove space at the end of colored pinyin
- Remove spaces between punctuation in pinyin
- Remove the space between erhua “r” suffix and main word in pinyin
- Prevent loss of punctuation when colorizing hanzi
- Erhua and third tone sandhi now handled properly
- Line-index support and separator support added for dictionary lookup
- Squash bug that means character colorisation to fail if audio generation off
- Pinyin is recognised and colored anywhere in the text
- Support for numbers has been greatly improved
- Added code testsuite
- Preferences window now used for config
- Many smaller modifications to improve usability
- Improved documentation & launch of website
Nick Cook | Website |
- Two versions are now being distributed: English (using CC-CEDICT) and German (using HanDeDict)
- Thanks to Rainer Menes for suggesting use of HanDeDict
- New character colorization feature
- If enabled, colored hanzi will be generated and placed in a field called “Color”
- to get the benefit of this you should alter models in a way to make use of colored Hanzi
- Fix to give better handle the formatting of dictionary entries
- Customisable dictionary meaning separator (defaults to line break)
- More reliable sound tag generation
- Automatic blanking feature if hanzi field is emptied and de-focused (makes input faster)
- Audio field is ignored if longer than 40 characters (i.e. if Anki has improved or recorded audio) [a better way to do this is in the works]
- Strip html from expression field before doing anything (prevents formating bugs)
- Several minor fixes and improvements to general usage
- Assigned version numbers retroactively to v0.01 to v0.03
Nick Cook | Website |
- Simplified and traditional generation now happens each lookup (no need to chose one or the other)
- Automatic English generation from dictionary
- English will only be generated for exact matches (words) not phrases
- Automatic colorization of pinyin by tone
- Text-to-speech feature (automatic auto generation)
- Support for turning off new features
- Tidy code (remove unused & rationalise messy bits)
- Update dictionary to latest CC-CEDICT
- Fix bug from 谁 [actually bug is no longer relevant under new directory structure]
- Fix problems with pinyin generation of “v”/“u:”
- Rearrange to more logical filenames (pinyinfetcher.py etc)
- Change dictionary structure for better management, easier updates, and to allow new features
- Split dictionary into 3 parts (CC-CEDICT, supplementary, and user) [entries prioritised from latter to former]
- Supplementary dictionary contains previously hard-coded entries and new data such as numbers
Nick Cook | Website |
- Dictionary updated to use adapted version of CC-CEDICT
- Entries increasing from 44,783 to 82,941
Damien Elmes | Website |
- Ported to Anki 0.9.9.6 and tidied up
- Brian Vaughan no longer maintaing plugin
Brian Vaughan | Website |
- Original release