I agree that regex could probably do all the text processing. However everytime it is raised comments are made as to it not handling typeform. Have you a suggestion for how it can deal with that?
I personally have questions over how typeform is handled and whether that way really is needed. Would it really be that problematic to embed tags such as XML in the translation string and still preserve indexing, I would have thought it is fully possible. May be I need a further explanation as to why embedding tags would cause problems, and examples might be useful.
Michael Whapples On 26/10/2013 19:44, Ken Perry wrote:
This could all be handled by a regex opp code in the current liblouis though. Ken -----Original Message----- From: liblouis-liblouisxml-bounce@xxxxxxxxxxxxx [mailto:liblouis-liblouisxml-bounce@xxxxxxxxxxxxx] On Behalf Of Susan Jolly Sent: Saturday, October 26, 2013 2:15 PM To: liblouis-liblouisxml@xxxxxxxxxxxxx Subject: [liblouis-liblouisxml] Please don't misunderstand the importance of what John Boyer is saying I think we are using the term "scripting language" in two different ways that is leading to misunderstanding and confusion, Certain existing programming languages such as Python are often referred to as scripting languages. You can read more about the use of this term in this Wikipedia article: http://en.wikipedia.org/wiki/Scripting_language I believe John Boyer is using this term as a synonym for what is also called a Domain Specific Language or DSL. He is not advocating developing a language such as Python. The concern is that we definitely need some high-level and uniform way for people familiar with the rules of a particular braille code to express or encode the possibly quite complex rules for the use of indicators in that code. As an example of the complexity of these rules I will contrast some of the rules for the use of the number sign in American English braille and in UEB. In our current English braille (EBAE) a single number sign is used before a braille sequence to indicate integers, real numbers, and also numeric sequences. A numeric sequence consists of digits plus one or more of six special characters: colon, comma, decimal, fraction line, hyphen, and slash. That is, the scope of the number sign is not terminated by any of these six characters but it is terminated by a space, slash, letter, etc. There are also various rules for when a letter sign is needed if a letter or group of letters immediately follows a number. In older versions of EBAE a letter sign was only needed to avoid ambiguity where the same braille cell could represent a letter or digit. The latest rule is that a letter sign is always needed if a letter immediate follows a number or is joined to it by a hyphen with the exception that if a single letter is immediately followed by another number sign, then the letter sign is not required. The rules for the use of the UEB numeric indicator are quite different from the EBAE ones. In the first place, the indicator is actually two cells. The first cell is always the standard dots 3456. The second cell can be a digit, a period, a decimal point, a comma, a space, or, if necessary, a second dots 3456. There are six symbols that don't end the scope of the indicator. Also the dot 5 numerical space doesn't end the scope if it is immediately followed by another digit. The UEB numeric indicator also establishes a special Grade 1 mode unless terminated by a hyphen or dash. This means that in many cases contractions cannot be used immediately after numbers. I hope it is clear from just this one example that this is a very serious issue that is going to require a lot of thought. Susan Jolly For a description of the software, to download it and links to project pages go to http://www.abilitiessoft.com For a description of the software, to download it and links to project pages go to http://www.abilitiessoft.com
For a description of the software, to download it and links to project pages go to http://www.abilitiessoft.com
