ansible/network/basics/uri.py

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

# (c) 2013, Romeo Theriault <romeot () hawaii.edu>
#
# This file is part of Ansible
#
# Ansible is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Ansible is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Ansible.  If not, see <http://www.gnu.org/licenses/>.
#
# see examples/playbooks/uri.yml

import cgi
import shutil
import tempfile
import datetime

try:
    import json
except ImportError:
    import simplejson as json

import ansible.module_utils.six as six


DOCUMENTATION = '''
---
module: uri
short_description: Interacts with webservices
description:
  - Interacts with HTTP and HTTPS web services and supports Digest, Basic and WSSE
    HTTP authentication mechanisms.
version_added: "1.1"
options:
  url:
    description:
      - HTTP or HTTPS URL in the form (http|https)://host.domain[:port]/path
    required: true
    default: null
  dest:
    description:
      - path of where to download the file to (if desired). If I(dest) is a
        directory, the basename of the file on the remote server will be used.
    required: false
    default: null
  user:
    description:
      - username for the module to use for Digest, Basic or WSSE authentication.
    required: false
    default: null
  password:
    description:
      - password for the module to use for Digest, Basic or WSSE authentication.
    required: false
    default: null
  body:
    description:
      - The body of the http request/response to the web service. If C(body_format) is set
        to 'json' it will take an already formated JSON string or convert a data structure
        into JSON.
    required: false
    default: null
  body_format:
    description:
      - The serialization format of the body. When set to json, encodes the
        body argument, if needed, and automatically sets the Content-Type header accordingly.
    required: false
    choices: [ "raw", "json" ]
    default: raw
    version_added: "2.0"
  method:
    description:
      - The HTTP method of the request or response. It MUST be uppercase.
    required: false
    choices: [ "GET", "POST", "PUT", "HEAD", "DELETE", "OPTIONS", "PATCH", "TRACE", "CONNECT", "REFRESH" ]
    default: "GET"
  return_content:
    description:
      - Whether or not to return the body of the request as a "content" key in
        the dictionary result. If the reported Content-type is
        "application/json", then the JSON is additionally loaded into a key
        called C(json) in the dictionary results.
    required: false
    choices: [ "yes", "no" ]
    default: "no"
  force_basic_auth:
    description:
      - The library used by the uri module only sends authentication information when a webservice
        responds to an initial request with a 401 status. Since some basic auth services do not properly
        send a 401, logins will fail. This option forces the sending of the Basic authentication header
        upon initial request.
    required: false
    choices: [ "yes", "no" ]
    default: "no"
  follow_redirects:
    description:
      - Whether or not the URI module should follow redirects. C(all) will follow all redirects.
        C(safe) will follow only "safe" redirects, where "safe" means that the client is only
        doing a GET or HEAD on the URI to which it is being redirected. C(none) will not follow
        any redirects. Note that C(yes) and C(no) choices are accepted for backwards compatibility,
        where C(yes) is the equivalent of C(all) and C(no) is the equivalent of C(safe). C(yes) and C(no)
        are deprecated and will be removed in some future version of Ansible.
    required: false
    choices: [ "all", "safe", "none" ]
    default: "safe"
  creates:
    description:
      - a filename, when it already exists, this step will not be run.
    required: false
  removes:
    description:
      - a filename, when it does not exist, this step will not be run.
    required: false
  status_code:
    description:
      - A valid, numeric, HTTP status code that signifies success of the
        request. Can also be comma separated list of status codes.
    required: false
    default: 200
  timeout:
    description:
      - The socket level timeout in seconds
    required: false
    default: 30
  HEADER_:
    description:
      - Any parameter starting with "HEADER_" is a sent with your request as a header.
        For example, HEADER_Content-Type="application/json" would send the header
        "Content-Type" along with your request with a value of "application/json".
        This option is deprecated as of C(2.1) and may be removed in a future
        release. Use I(headers) instead.
    required: false
    default: null
  headers:
    description:
        - Add custom HTTP headers to a request in the format of a YAML hash
    required: false
    default: null
    version_added: '2.1'
  others:
    description:
      - all arguments accepted by the M(file) module also work here
    required: false
  validate_certs:
    description:
      - If C(no), SSL certificates will not be validated.  This should only
        set to C(no) used on personally controlled sites using self-signed
        certificates.  Prior to 1.9.2 the code defaulted to C(no).
    required: false
    default: 'yes'
    choices: ['yes', 'no']
    version_added: '1.9.2'
notes:
  - The dependency on httplib2 was removed in Ansible 2.1
author: "Romeo Theriault (@romeotheriault)"
'''

EXAMPLES = '''
# Check that you can connect (GET) to a page and it returns a status 200
- uri: url=http://www.example.com

# Check that a page returns a status 200 and fail if the word AWESOME is not
# in the page contents.
- action: uri url=http://www.example.com return_content=yes
  register: webpage

- action: fail
  when: "'AWESOME' not in webpage.content"


# Create a JIRA issue
- uri:
    url: https://your.jira.example.com/rest/api/2/issue/
    method: POST
    user: your_username
    password: your_pass
    body: "{{ lookup('file','issue.json') }}"
    force_basic_auth: yes
    status_code: 201
    body_format: json

# Login to a form based webpage, then use the returned cookie to
# access the app in later tasks

- uri:
    url: https://your.form.based.auth.example.com/index.php
    method: POST
    body: "name=your_username&password=your_password&enter=Sign%20in"
    status_code: 302
    HEADER_Content-Type: "application/x-www-form-urlencoded"
  register: login

- uri:
    url: https://your.form.based.auth.example.com/dashboard.php
    method: GET
    return_content: yes
    HEADER_Cookie: "{{login.set_cookie}}"

# Queue build of a project in Jenkins:
- uri:
    url: "http://{{ jenkins.host }}/job/{{ jenkins.job }}/build?token={{ jenkins.token }}"
    method: GET
    user: "{{ jenkins.user }}"
    password: "{{ jenkins.password }}"
    force_basic_auth: yes
    status_code: 201

'''


def write_file(module, url, dest, content):
    # create a tempfile with some test content
    fd, tmpsrc = tempfile.mkstemp()
    f = open(tmpsrc, 'wb')
    try:
        f.write(content)
    except Exception:
        err = get_exception()
        os.remove(tmpsrc)
        module.fail_json(msg="failed to create temporary content file: %s" % str(err))
    f.close()

    checksum_src   = None
    checksum_dest  = None

    # raise an error if there is no tmpsrc file
    if not os.path.exists(tmpsrc):
        os.remove(tmpsrc)
        module.fail_json(msg="Source %s does not exist" % (tmpsrc))
    if not os.access(tmpsrc, os.R_OK):
        os.remove(tmpsrc)
        module.fail_json( msg="Source %s not readable" % (tmpsrc))
    checksum_src = module.sha1(tmpsrc)

    # check if there is no dest file
    if os.path.exists(dest):
        # raise an error if copy has no permission on dest
        if not os.access(dest, os.W_OK):
            os.remove(tmpsrc)
            module.fail_json(msg="Destination %s not writable" % (dest))
        if not os.access(dest, os.R_OK):
            os.remove(tmpsrc)
            module.fail_json(msg="Destination %s not readable" % (dest))
        checksum_dest = module.sha1(dest)
    else:
        if not os.access(os.path.dirname(dest), os.W_OK):
            os.remove(tmpsrc)
            module.fail_json(msg="Destination dir %s not writable" % (os.path.dirname(dest)))

    if checksum_src != checksum_dest:
        try:
            shutil.copyfile(tmpsrc, dest)
        except Exception:
            err = get_exception()
            os.remove(tmpsrc)
            module.fail_json(msg="failed to copy %s to %s: %s" % (tmpsrc, dest, str(err)))

    os.remove(tmpsrc)


def url_filename(url):
    fn = os.path.basename(six.moves.urllib.parse.urlsplit(url)[2])
    if fn == '':
        return 'index.html'
    return fn


def absolute_location(url, location):
    """Attempts to create an absolute URL based on initial URL, and
    next URL, specifically in the case of a ``Location`` header.
    """

    if '://' in location:
        return location

    elif location.startswith('/'):
        parts = six.moves.urllib.parse.urlsplit(url)
        base = url.replace(parts[2], '')
        return '%s%s' % (base, location)

    elif not location.startswith('/'):
        base = os.path.dirname(url)
        return '%s/%s' % (base, location)

    else:
        return location


def uri(module, url, dest, body, body_format, method, headers, socket_timeout):
    # is dest is set and is a directory, let's check if we get redirected and
    # set the filename from that url
    redirected = False
    redir_info = {}
    r = {}
    if dest is not None:
        # Stash follow_redirects, in this block we don't want to follow
        # we'll reset back to the supplied value soon
        follow_redirects = module.params['follow_redirects']
        module.params['follow_redirects'] = False
        dest = os.path.expanduser(dest)
        if os.path.isdir(dest):
            # first check if we are redirected to a file download
            _, redir_info = fetch_url(module, url, data=body,
                                      headers=headers,
                                      method=method,
                                      timeout=socket_timeout)
            # if we are redirected, update the url with the location header,
            # and update dest with the new url filename
            if redir_info['status'] in (301, 302, 303, 307):
                url = redir_info['location']
                redirected = True
            dest = os.path.join(dest, url_filename(url))
        # if destination file already exist, only download if file newer
        if os.path.exists(dest):
            t = datetime.datetime.utcfromtimestamp(os.path.getmtime(dest))
            tstamp = t.strftime('%a, %d %b %Y %H:%M:%S +0000')
            headers['If-Modified-Since'] = tstamp

        # Reset follow_redirects back to the stashed value
        module.params['follow_redirects'] = follow_redirects

    resp, info = fetch_url(module, url, data=body, headers=headers,
                           method=method, timeout=socket_timeout)

    try:
        content = resp.read()
    except AttributeError:
        # there was no content, but the error read()
        # may have been stored in the info as 'body'
        content = info.pop('body', '')

    r['redirected'] = redirected or info['url'] != url
    r.update(redir_info)
    r.update(info)

    return r, content, dest


def main():
    argument_spec = url_argument_spec()
    argument_spec.update(dict(
        dest = dict(required=False, default=None, type='path'),
        url_username = dict(required=False, default=None, aliases=['user']),
        url_password = dict(required=False, default=None, aliases=['password']),
        body = dict(required=False, default=None, type='raw'),
        body_format = dict(required=False, default='raw', choices=['raw', 'json']),
        method = dict(required=False, default='GET', choices=['GET', 'POST', 'PUT', 'HEAD', 'DELETE', 'OPTIONS', 'PATCH', 'TRACE', 'CONNECT', 'REFRESH']),
        return_content = dict(required=False, default='no', type='bool'),
        follow_redirects = dict(required=False, default='safe', choices=['all', 'safe', 'none', 'yes', 'no']),
        creates = dict(required=False, default=None, type='path'),
        removes = dict(required=False, default=None, type='path'),
        status_code = dict(required=False, default=[200], type='list'),
        timeout = dict(required=False, default=30, type='int'),
        headers = dict(required=False, type='dict', default={})
    ))

    module = AnsibleModule(
        argument_spec=argument_spec,
        check_invalid_arguments=False,
        add_file_common_args=True
    )

    url  = module.params['url']
    body = module.params['body']
    body_format = module.params['body_format'].lower()
    method = module.params['method']
    dest = module.params['dest']
    return_content = module.params['return_content']
    creates = module.params['creates']
    removes = module.params['removes']
    status_code = [int(x) for x in list(module.params['status_code'])]
    socket_timeout = module.params['timeout']

    dict_headers = module.params['headers']

    if body_format == 'json':
        # Encode the body unless its a string, then assume it is preformatted JSON
        if not isinstance(body, basestring):
            body = json.dumps(body)
        dict_headers['Content-Type'] = 'application/json'

    # Grab all the http headers. Need this hack since passing multi-values is
    # currently a bit ugly. (e.g. headers='{"Content-Type":"application/json"}')
    for key, value in module.params.iteritems():
        if key.startswith("HEADER_"):
            skey = key.replace("HEADER_", "")
            dict_headers[skey] = value

    if creates is not None:
        # do not run the command if the line contains creates=filename
        # and the filename already exists.  This allows idempotence
        # of uri executions.
        if os.path.exists(creates):
            module.exit_json(stdout="skipped, since %s exists" % creates,
                             changed=False, stderr=False, rc=0)

    if removes is not None:
        # do not run the command if the line contains removes=filename
        # and the filename do not exists.  This allows idempotence
        # of uri executions.
        if not os.path.exists(removes):
            module.exit_json(stdout="skipped, since %s does not exist" % removes, changed=False, stderr=False, rc=0)

    # Make the request
    resp, content, dest = uri(module, url, dest, body, body_format, method,
                              dict_headers, socket_timeout)
    resp['status'] = int(resp['status'])

    # Write the file out if requested
    if dest is not None:
        if resp['status'] == 304:
            changed = False
        else:
            write_file(module, url, dest, content)
            # allow file attribute changes
            changed = True
            module.params['path'] = dest
            file_args = module.load_file_common_arguments(module.params)
            file_args['path'] = dest
            changed = module.set_fs_attributes_if_different(file_args, changed)
        resp['path'] = dest
    else:
        changed = False

    # Transmogrify the headers, replacing '-' with '_', since variables dont
    # work with dashes.
    uresp = {}
    for key, value in resp.iteritems():
        ukey = key.replace("-", "_")
        uresp[ukey] = value

    try:
        uresp['location'] = absolute_location(url, uresp['location'])
    except KeyError:
        pass

    # Default content_encoding to try
    content_encoding = 'utf-8'
    if 'content_type' in uresp:
        content_type, params = cgi.parse_header(uresp['content_type'])
        if 'charset' in params:
            content_encoding = params['charset']
        u_content = unicode(content, content_encoding, errors='replace')
        if 'application/json' in content_type or 'text/json' in content_type:
            try:
                js = json.loads(u_content)
                uresp['json'] = js
            except:
                pass
    else:
        u_content = unicode(content, content_encoding, errors='replace')

    if resp['status'] not in status_code:
        uresp['msg'] = 'Status code was not %s: %s' % (status_code, uresp.get('msg', ''))
        module.fail_json(content=u_content, **uresp)
    elif return_content:
        module.exit_json(changed=changed, content=u_content, **uresp)
    else:
        module.exit_json(changed=changed, **uresp)


# import module snippets
from ansible.module_utils.basic import *
from ansible.module_utils.urls import *

if __name__ == '__main__':
    main()