autocomplete over project for Sublime Text 2

ctags_autocomplete.py

# CTags based autocompletion plugin for Sublime Text 2
# You can add the file to the User Package in ~/Library/Application Support/Sublime Text 2/Packages and restart Sublime Text 2.
# generate the .tags file in your project root with "ctags -R -f .tags"

import sublime, sublime_plugin, os

class AutocompleteAll(sublime_plugin.EventListener):

    def on_query_completions(self, view, prefix, locations):
        tags_path = view.window().folders()[0]+"/.tags"
        results=[]
        if (not view.window().folders() or not os.path.exists(tags_path)): #check if a project is open and the .tags file exists
            return results
        f=os.popen("grep -i '^"+prefix+"' '"+tags_path+"' | awk '{ print $1 }'") # grep tags from project directory .tags file
        for i in f.readlines():
            results.append([i.strip()])
        results = [(item,item) for sublist in results for item in sublist] #flatten
        results = list(set(results)) # make unique
        results.sort() # sort
        return results

You can use this improved version (for windows) to look in all your project folders. Also you will need to download Grep and Gawk for windows (http://gnuwin32.sourceforge.net/packages.html) and add the installed folder to your system PATH and restart windows before try it.

# CTags based autocompletion plugin for Sublime Text 2
# You can add the file to the User Package in ~/Library/Application Support/Sublime Text 2/Packages and restart Sublime Text 2.
# generate the .tags file in your project root with "ctags -R -f .tags"

import sublime, sublime_plugin, os, subprocess

class AutocompleteAll(sublime_plugin.EventListener):

    def on_query_completions(self, view, prefix, locations):
        results=[]
        tags_paths = [folder + '/.tags' for folder in view.window().folders()]
        for tags_path in tags_paths:
            if (not view.window().folders() or not os.path.exists(tags_path)): #check if a project is open and the .tags file exists
                return results
        tags_paths = ' '.join([str(x) for x in tags_paths])    
        f=os.popen('grep -ih "^'+prefix+'" ' + tags_paths + ' | gawk "{ print $1 }"') # egrep tags from project directory .tags file
        for i in f.readlines():
            results.append([i.strip()])
        results = [(item,item) for sublist in results for item in sublist] #flatten
        results = list(set(results)) # make unique
        results.sort() # sort
        return results

sdf

It works, but it's a bit slow for big tags files. I wonder if there's something from the actual ctags plugin you might be able to use to speed it up?

removing -i improves performance considerably, see my fork if needed.

I am the maintainer of the CTags plugin. It would be by far best to integrate this into that plugin, rather than use a separate one. It keeps a dict loaded in RAM, and also sorts the tags file for easier lookup.

It already has a command to list all tags in the quick panel. I think this could tack onto that pretty easily. It just needs to load this list into the completions. I also had filed ticket to add support for ./-> completions (SublimeText/CTags#75). That will be a bit more involved, but along the same vein, and definitely desirable.

I tried this and got ctags: illegal option -- R in Terminal.

@barneywilliams gettings this in CTags would be great.

Cool! It's great !

@Grawl this is because you are using the Mac OS X shipped version of ctags. Install ctags from homebrew and use the one located at /usr/local/bin/ctags.