Simple python script to generate a maven pom.xml file from a list of jars in a directory (with proxy support but, remember urllib2 proxy does not work for https)

gav.py

import os
import json
import urllib2
import hashlib
import argparse
import urlparse
import traceback


# configurable parameters
def maven_central(sha1_hash):
  """
  Use the hash search of maven central
  """
  url_template = 'http://search.maven.org/solrsearch/select?q=1:%22{sha}%22&rows=20&wt=json'
  content = get_url(url_template.format(sha=sha1_hash))
  if ( content ):
    try:
      js = json.loads(content)
      if ( js['response']['numFound'] > 0 and len(js['response']['docs']) > 0 ):
        dep = js['response']['docs'][0]
        return dep_template.format(groupId=dep['g'], artifactId=dep['a'], version=dep['v'])
    except Exception as e:
      print('[ERROR] ' + str(e))
  return None

def jboss_repo(sha1_hash):
  """
  User the hash search of jboss nexus repo
  """
  url_template = 'https://repository.jboss.org/nexus/service/local/lucene/search?sha1={sha}'
  content = get_url(url_template.format(sha=sha1_hash), None, {'Accept': 'application/json'})
  if ( content ):
    try:
      js = json.loads(content)
      if ( js['totalCount'] > 0 and len(js['data']) > 0 ):
        dep = js['data'][0]
        return dep_template.format(groupId=dep['groupId'], artifactId=dep['artifactId'], version=dep['version'])
    except Exception as e:
      print('[ERROR] ' + str(e))
  return None  


def debug(msg):
  if ( DEBUG ):
    print(msg)
  
def get_dependency(sha1_hash):
  providers = {
    'central': maven_central,
    'jboss'  : jboss_repo
  }
  dep = None
  for name, provider in providers.items():
    debug("Trying %s" % name)
    dep = provider(sha1_hash)
    if ( dep != None ):
      break
  return dep

def list_jars(dr):
  jars = []
  for fn in os.listdir(dr):
    jar = os.path.join(os.path.realpath(dr), fn)
    if ( os.path.isfile(jar) and jar.endswith('.jar') ):
      jars.append(jar)
  return jars

def list_jar_files(dirs=['.']):
  files = []
  for dr in dirs:
    files.extend(list_jars(dr))
  return files

def calculate_hashes(jar_files):
  default_sha1 = hashlib.sha1().hexdigest()
  sha_map = {}
  for jar in jar_files:
    sha1_hash = sha1(jar)
    if ( default_sha1 != sha1_hash ):
      sha_map[os.path.basename(jar)] = sha1_hash
  return sha_map

def set_proxy(proxy_cfg={}):
  debug('setting proxy : {0}'.format(proxy_cfg))
  opener = urllib2.build_opener(urllib2.ProxyHandler(proxy_cfg))
  urllib2.install_opener(opener)

def make_request(url, data=None, headers={}):
  req = urllib2.Request(url, data) if data else urllib2.Request(url)
  for k,v in default_headers.items():
    req.add_header(k, v)
  for k,v in headers.items():
    req.add_header(k, v)
  return req

def get_url(url, data=None, headers={}):
  try:
    debug('GET {0}'.format(url))
    req = make_request(url, data, headers)
    f = urllib2.urlopen(req)
    b = f.read()
    f.close()
    return b
  except Exception as e:
    traceback.print_exc()
    print('[ERROR] ' + str(e))
    print('[ERROR] fetching url : %s' % url)

def sha1(fp):
  blocksize = 64*1024
  sha = hashlib.sha1()
  with open(fp, 'rb') as fp:
    while True:
      data = fp.read(blocksize)
      if not data:
        break
      sha.update(data)
  return sha.hexdigest()


# constants
pom_template = '''<project xmlns="http://maven.apache.org/POM/4.0.0"
         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
  <modelVersion>4.0.0</modelVersion>

  <groupId>group-id</groupId>
  <artifactId>artifact-id</artifactId>
  <version>1.0</version>
  <packaging>jar</packaging>

  <properties>
    <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
  </properties>

  <dependencies>
    {dependencies}
  </dependencies>
</project>'''

dep_template = '''
    <dependency>
      <groupId>{groupId}</groupId>
      <artifactId>{artifactId}</artifactId>
      <version>{version}</version>
    </dependency>'''
default_headers = {
  'User-Agent': 'Mozilla/5.0 (Windows NT 6.1; rv:21.0) Gecko/20100101 Firefox/21.0'
}



# configure argparse
parser = argparse.ArgumentParser(description='Generate sample pom.xml from list of jars')
parser.add_argument('jars_dir', metavar='dir', nargs='*', help='directory containing the jars')
parser.add_argument('-o', '--output', metavar='file', required=False, help='output file name')
parser.add_argument('-p', '--proxy',  metavar='proxy', required=False, help='use specified proxy')
parser.add_argument('-d', '--debug',  required=False, action='store_true', help='print debug info')
args = parser.parse_args()

DEBUG = args.debug
debug(args)
dirs = args.jars_dir

output_file = args.output if args.output else 'pom.xml'
if ( args.proxy ):
  proxy_cfg = {}
  up = urlparse.urlparse(args.proxy)
  proxy_cfg[up.scheme] = args.proxy
  set_proxy(proxy_cfg)
else:
  set_proxy()
  
sha1_hashmap = calculate_hashes(list_jar_files(dirs))
dependencies = []
for jar,sha1 in sha1_hashmap.items():
  debug('checking for {0}...'.format(jar))
  dep = get_dependency(sha1)
  if ( dep ):
    dependencies.append(dep)
    debug('success')
  else:
    print('failed to find dependency for %s' % jar)

f = open(output_file, 'w')
f.write(pom_template.format(dependencies='\n'.join(dependencies)))
f.close()