Closed kleinesfilmroellchen closed 4 years ago
update: commit e7dc798d01fb1272f30403e691c4ee53ebf4dec7 fixes:
However, the same commit does break the index indication for errors, which is not fatal (error message is way more informative now than before) but needs to be addressed.
update: indication of line and index in line is now fixed. Also, the Tokenizer refactor exposed the Java Matcher auto-reset behavior, which is unwanted and has been circumvented.
update: code block behavior is broken; again because of the Java Matcher
auto-reset behavior, which also deletes regions. There are two routes from here: build a memory-heavy AST parser or circumvent all matcher functionality except when actual regex searches are performed (in essence, reimplement Java's Matcher
with desired behavior). I will go with the second route first, as it will be the easiest.
update: all Matcher
behavior is circumvented by keeping track of state separately. Code block still does not work, also, bug 'linenumber' is introduced.
Closing, future bugs get their own issue.