Here is a complex bug I stumbled upon: when you recreate an indent by yourself (adding a new line, erasing inserted indent, adding your desired indent) and then insert a snippet with a newline at the end, then if there is no tabstops, then the indent is added ok, but when there are tabstops before newline, then the auto_indent is not created. And fun thing: that happens only when `auto_indent` option is enabled, if you'd disable it, there would be an automatic indent for snippet, but the bug won't be there.

Well, that's complex, but here is an example code that reproduces the problem and what you could run through ST's console:

edit = view.begin_edit("lol");view.run_command('insert', {"characters": "\n"});view.erase(edit, sublime.Region(view.line(view.sel()[0]).a, view.sel()[0].a));view.run_command('insert', {"characters": "    "});view.run_command("insert_snippet", {"contents" : "snippet: $1foo\n"});view.end_edit(edit)

Or if expanded:

edit = view.begin_edit("lol")
view.run_command('insert', {"characters": "\n"})
view.erase(edit, sublime.Region(view.line(view.sel()[0]).a, view.sel()[0].a))
view.run_command('insert', {"characters": "    "})
view.run_command("insert_snippet", {"contents" : "snippet: $1foo\n"})
view.end_edit(edit)

If you'd run it somewhere with an indent, the indent would be created at the end of the snippet when you'd `tab` to the ending tabstop. But if you'd remove the `$1` or set `"auto_indent": false`, then it would magically work.

it would be nice to have dark menus (or customizable) in OSX, the sublime window is pretty, why not make the menus pretty too? i made a little example 

http://cl.ly/2y2t3H3L0T2M 

and here is a screenshot  

http://cl.ly/image/0w0w3C1C1o3v

Hi,

I'm am using ST2 on mac with REPL and R. What I would like to do is open a quartz display (i.e. generate a figure) without having ST2 switch to a different screen. Any help you could provide would be greatly appreciated. Thanks!

~Alex

Isn't there a good possibility to replace the regular expression engine with a faster one? I guess you use "re" from python, but there are faster ones.

Find/replace on large files is rather slow and could crash. For example:

a file of 1.000.000 rows with empty lines, replace "^\n" with "" (without the quotes), to remove the empty lines. This will take a while or crash.

This is because the "re" uses backtracking, which could be rather slow. But there are also regular expression engines which use finite state machine (FSA). You can see this page for more information: http://swtch.com/~rsc/regexp/regexp1.html

Possible options could be (but not limited to):

• RE2 - http://code.google.com/p/re2/ 
  (with python bindings: https://github.com/axiak/pyre2)
• PIRE - https://github.com/dprokoptsev/pire
• Iregexp (regular expression engine of V8) - http://blog.quenta.org/2009/02/irregexp.html (but no python yet, as i know of)
• http://blog.errstr.com/2013/01/22/implementing-a-more-powerful-regex/

I would like to be able to edit the attributes of the "NON-HIGHLIGHTING THIN RED ZIGZAG with TRANSPARENT-BACKGROUND" that appears when you misspell a word.   In general  it is not clearly visible with certain themes, and not scaled appropriately with small and large text, so I think manually adjustments are in order.   Maybe the style (underline, zigzag, dotted), pixel width, (whether or not it highlights the text, or changes the font color of the misspelled words/characters, or the background of the underline element or text.