Fetch and convert PDB files from Microsoft's symbol server to Breakpad .sym format

fetch_microsoft_symbols.py

#!/usr/bin/env python

import distutils.spawn
import os
import requests
import shutil
import subprocess
import sys
import tempfile
import urlparse

URL = 'http://msdl.microsoft.com/download/symbols/'
USER_AGENT = 'Microsoft-Symbol-Server/6.3.0.0'

dump_syms_command = ['wine64', 'dump_syms.exe']

def fetch_symbol(debug_id, debug_file):
    '''
    Attempt to fetch a PDB file from Microsoft's symbol server.
    '''
    url = urlparse.urljoin(URL, os.path.join(debug_file,
                                             debug_id,
                                             debug_file[:-1] + '_'))
    r = requests.get(url,
                     headers={'User-Agent': USER_AGENT})
    if r.status_code == 200:
        return r.content
    return None

def fetch_and_dump_symbols(tmpdir, debug_id, debug_file):
    pdb_bytes = fetch_symbol(debug_id, debug_file)
    if not pdb_bytes or not pdb_bytes.startswith(b'MSCF'):
        return False
    pdb_path = os.path.join(tmpdir, debug_file[:-1] + '_')
    with open(pdb_path, 'wb') as f:
        f.write(pdb_bytes)
    try:
        # Decompress it
        subprocess.check_call(['cabextract', '-d', tmpdir, pdb_path])
        pdb_path = os.path.join(tmpdir, debug_file)
        # Dump it
        return subprocess.check_output(dump_syms_command + [pdb_path])
    except subprocess.CalledProcessError:
        return None

def main():
    cabextract = distutils.spawn.find_executable('cabextract')
    if not cabextract:
        # error
        pass
    files = [
        ('69DDFBCBBC14421D8CB974F8EDC414102', 'wntdll.pdb'),
        ('0FCE9CC301ED4567A819705B2718E1D62', 'wuser32.pdb')
    ]
    tmpdir = tempfile.mkdtemp(prefix='symbols')
    try:
        for debug_id, debug_file in files:
            symbols = fetch_and_dump_symbols(tmpdir, debug_id, debug_file)
            if not symbols:
                continue
            dirs = os.path.join(tmpdir, 'symbols', debug_file, debug_id)
            os.makedirs(dirs)
            with open(os.path.join(dirs, debug_file[:-4] + '.sym'), 'w') as f:
                f.write(symbols)
    finally:
        pass
        #shutil.rmtree(tmpdir)

if __name__ == '__main__':
    main()