Idan Kamara
Idan Kamara
Throwing another comment here to say I'd like to see this implemented, for a tool I'm working on. It seems pretty basic to have position information in a parser. :)
I meant absolute position. I ended up with [this hack](https://github.com/idank/showthedocs/blob/master/showdocs/parsers/sql.py#L29), for those interested: ``` def calcpositions(root, index=0): positions = {} if root.is_group(): startindex = index for token in root.tokens: nestedpos,...
Yeah I happen to need position of groups too. This whole thing falls on its face if the parser decides to suppress parts of the input one day, say whitespace...
nit: please use this commit message `tokenizer: fix backslash newline and multiple newlines at end (fixes #...)`
I'm a bit hesitant to merge it then. Could a fix alter the new logic significantly?
Got it, that makes sense. I'm a bit cautious to steer from bash but if we think the direction we've taken with these changes is the right one then I'm...
Sounds good to me. It was left empty because I didn't have a need for it at the time. If it's easy to add, go for it.
I didn't notice that pull request, sorry. See the comment I left there.