C++ header include graph

find_headers.py

#!/usr/bin/env python3

import sys
import os
import re
from os import path
from glob import glob

from graph_tool.all import *

ignore_dirs = '''
    _build
    .svn
'''.split()

include_pattern = re.compile('^\s*#\s*include\s*(<(.*)>|"(.*)")')

def included_headers(filename):
    ret = []
    with open(filename,'r') as fin:
        for m in include_pattern.findall(fin.read().decode(),re.MULTILINE):
            ret.append(m.group(2) or m.group(3))
    return ret

g = Graph()
header_names = g.new_vertex_property('string')
header_types = g.new_vertex_property('string')
vertices = {}

for root,dirs,files in os.walk(sys.argv[1]):
    for d in ignore_dirs:
        if d in dirs:
            dirs.remove(d)
    for f in files:
        if path.splitext(f)[-1] == '.h':
            print(f)
            v = g.add_vertex()
            vertices[f] = v
            header_names[v] = f
            header_types[v] = 'internal'

for root,dirs,files in os.walk(sys.argv[1]):
    for d in ignore_dirs:
        if d in dirs:
            dirs.remove(d)
    for f in files:
        if path.splitext(f)[-1] in ['.h','.cpp']:
            if f in header_names:
                headers = included_headers(path.join(root,f))
                for h in headers:
                    if h in header_names:
                        print(f,'-->',h)
                        g.add_edge(vertices[f],vertices[h])
                    else:
                        print(h)
                        v = g.add_vertex()
                        vertices[h] = v
                        header_names[v] = h
                        header_types[v] = 'external'

g.vertex_properties['header_name'] = header_name
g.vertex_properties['header_type'] = header_type
g.save('header_graph.xml.gz')