Correctly handle \u in Notation 3 files. #426
Conversation
It was using case insensitive regex, so it was mistaking \u and \U.
|
hmm, from what you write and the changes this seems legit, but other tests now fail... @gromgull you wrote that :p |
It was using case insensitive regex, so it was mistaking \u and \U.
… Add correct case for prefix escape in unicode expansion in litterals
|
Hello, I was encountering the same problem and was about to apply the exact same correction when I found this pull request. Ideally, I would like to add a pull request to this pull request, but i do not quite know how to do it. Regards, |
…scape-n3 This should have an extra commit that might fix some tests
|
@ymph I have merged that commit into this branch. The Travis CI tests mostly pass, except for the python 3.4 one, which failed due to a network problem of accessing github |
|
@rory Thank you very much, this is certainly the best way. Regards, |
|
re-running the test... |
correctly handle \u and \U in n3 files
It was using case insensitive regex, so it was mistaking \u and \U.
Sec 6.4 of the W3C spec (http://www.w3.org/TR/turtle/#sec-escapes) says that it's either \uXXXX (4 chars) or \UXXXXXXXX (8 chars). The current code uses a regex, but the regex has the case-insensitive flag set. So if there is a \uXXXX in the turtle file, the 8 character regex for \U will match and it'll try to pull in 8 character (rather than 4).
I've included tests that demostrate this.