-
Version 3.1.0
- Uses udpated external g2p library for conversions (v0.4+)
- You must upgrade your installed version of the g2p library to >= 0.4.x
-
Version 3.0.0
- Uses udpated external g2p library for conversions
- Breaking change: both an input_language and and output_language now have to be specified instead of just table as in the previous API. In the command line, this is done with the
-il
and-ol
arguments respectively. - Breaking change: The argument language
-l
has now been deprecated and if using a custom mapping file (csv, tsv, psv, json, or xlsx) you must write your rules in the style used byg2p
and provide the path with the mapping-m
argument.
-
Version 2.5.0
- Uses external g2p library for transductions
- Breaking change: from is now labelled in, to is now labelled out, after is now labelled context_after and before is now labelled context_before
- Breaking change: custom tables can still be passed as a path with the
-l
flag, but all default languages are now from the g2p library and must be called with the language-l
flag and an ISO code and the table-t
flag and the name of the lookup table. Typeconvertextract -h
for a list of available languages.
-
Version 2.0.1
- Added ability to take csv or plain list as cors
- Fixed some problems with regex. Before columns are processed as positive lookbehinds, after columns are processed as positive lookaheads
-
Version 2.0
- Migrated to Python 3
- Added processText method for processing of plain text
- Added --no-write argument
-
Version 1.6
- Added support for feeding edge case https://medium.com/digital-linguistics/transliteration-in-javascript-99d306996752
- Note: this solution will cause problems for documents that contain characters in the geometric shapes code block: Hex 25A0-25FF.
-
Version 1.5.1
- Fixed typo from Duolos to Doulos
-
Version 1.5
- Added Regex support for .txt, .docx, .pptx and .xlsx conversions
- Breaking change: correspondence charts must now be ordered with the character to convert from in column A and the character to convert to in column B
-
Version 1.3
- Fixed an issue with pptx not preserving underlines.
- Updated URL in warning messages
- Fixed package includes for default correspondences
-
Version 1.2.2
- Fixed a typo for xlsx parser
-
Version 1.2.1
- Fixed a typo in PyPi submitted version 1.2
-
Version 1.2
- Fixed a silly issue causing pptx conversions to take far too long, they are faster now.
- Revised docx conversions to allow for pictures to be preserved
-
Version 1.1
- Fixed txt_parser error with new line breaks. First release of stable version.
-
Version 1.0.8 & 1.0.9
- Fixed correspondence path for server deployment
-
Version 1.0.7
- Fixed script kwargs issue
-
Version 1.0.6
- Added -p argument for .xlsx files
-
Version 1.0.5
- Fixed shebang issue
-
Version 1.0.4
- Supports .docx, .pptx, .txt, .xlsx
- Supports Heiltsuk Duolos, Heiltsuk Times, Ts'ilhqot'in Duolos