Tag generator for Elm files

elmtags.py

#! /usr/bin/env python
# -*- coding: utf-8 -*-

from __future__ import print_function

help_text = """
Extracts tags from Elm files. Useful for the Tagbar plugin.

Usage:
Install Tagbar (http://majutsushi.github.io/tagbar/). Then, put this file
anywhere and add the following to your .vimrc:

let g:tagbar_type_elm = {
          \   'ctagstype':'elm'
          \ , 'kinds':['h:header', 'i:import', 't:type', 'f:function']
          \ , 'sro':'&&&'
          \ , 'kind2scope':{'h':'header', 'i:import', 't:type', 'f:function'}
          \ , 'sort':0
          \ , 'ctagsbin':'/path/to/elmtags.py'
          \ }
"""

import sys
import re

if len(sys.argv) < 2:
    print(help_text)
    exit()

filename = sys.argv[1]

re_header = re.compile(r"^-- (.*)$")
re_import = re.compile(r"^import ([^ \n]+)( as ([^ \n]+))?( exposing (\([^)]+\)))?")
re_type = re.compile(r"^type( alias)? ([^ \n]+)( =)?$")
re_function = re.compile(r"^([^ ]+) : (.*)$")


file_content = []
try:
    with open(filename, "r") as vim_buffer:
        file_content = vim_buffer.readlines()
except:
    exit()

cur_head = ""
for lnum, line in enumerate(file_content):

    match_header = re_header.match(line)
    match_import = re_import.match(line)
    match_type = re_type.match(line)
    match_function = re_function.match(line)
    cur_searchterm = ""
    cur_kind = ""
    args = ""
    lines = ""

    if match_header:
        cur_head = match_header.group(1)
        cur_tag = cur_head
        cur_searchterm = "^-- " + cur_tag + "$"
        cur_kind = "h"
    elif match_import:
        cur_tag = match_import.group(1)
        cur_searchterm = "^import " + cur_tag
        cur_kind = "i"
        args = "\theader:" + cur_head
        if match_import.group(3):
            args = args + "\tsignature:(" + cur_tag + ")"
            cur_tag = match_import.group(3)
        if match_import.group(5):
            exposing = match_import.group(5).strip("()").replace(" ", "").split(",")
            for exposed in exposing:
                lines = lines + '\n{0}\t{1}\t/{2}/;"\t{3}\tline:{4}{5}'.format(exposed, filename, cur_searchterm, "e", str(lnum+1), "\taccess:public\timport:" + cur_head + "&&&" + cur_tag)
    elif match_type:
        cur_tag = match_type.group(2)
        if match_type.group(1) == " alias":
            args = "\taccess:protected"
        cur_searchterm = "^type " + cur_tag + "$"
        cur_kind = "t"
        args = args + "\theader:" + cur_head
    elif match_function:
        cur_tag = match_function.group(1)
        cur_searchterm = "^" + cur_tag + " :"
        cur_kind = "f"
        args = "\theader:" + cur_head + "\tsignature:(" + match_function.group(2) + ")"
    else:
        continue

    print('{0}\t{1}\t/{2}/;"\t{3}\tline:{4}{5}'.format(
        cur_tag, filename, cur_searchterm, cur_kind, str(lnum+1), args) + lines)

Hello Maxine.

On this line -> https://gist.github.com/MaximeWack/388c393b7db290dd732f0a2d403118c5#file-elmtags-py-L17

The key-value pairs need some separation between the key and the value, here's an example.

'kind2scope':{'h':'header', 'i':'import', 't':'type', 'f':'function'}

😬

Oh, and by the way thanks for the post. :D