Adding SCP support

.gitignore

@@ -46,7 +46,7 @@ config.py
 *~
 
 # don't show the data on git status
-qiita_db/support_files/test_data
+qiita_db/support_files/test_data/
 
 # ignoring redis files
 dump.rdb
@@ -60,3 +60,6 @@ qiita_pet/static/doc/
 
 # webdis log
 webdis.log
+
+# test keys should generated in travis
+qiita_ware/test/test_data/test_key

.travis.yml

@@ -1,6 +1,6 @@
 dist: precise
 language: python
-sudo: false
+sudo: true
 env:
   global:
     - PYTHON_VERSION=3.5
@@ -50,6 +50,8 @@ install:
   - export REDBIOM_HOST=http://127.0.0.1:7379
   - cp $PWD/qiita_core/support_files/BIOM\ type_2.1.4.conf ~/.qiita_plugins/BIOM\ type_2.1.4\ -\ Qiime2.conf
   - touch ~/.bash_profile
+  - ssh-keygen -t rsa -C "ssh test key" -f $PWD/qiita_ware/test/test_data/test_key -P ""
+  - mv $PWD/qiita_ware/test/test_data/test_key.pub ~/.ssh/
   # Install the biom plugin so we can run the analysis tests
   - wget https://data.qiime2.org/distro/core/qiime2-2017.12-py35-linux-conda.yml
   - travis_retry conda env create -q -n qtp-biom --file qiime2-2017.12-py35-linux-conda.yml

qiita_db/support_files/patches/66.sql

@@ -0,0 +1,3 @@
+-- August 6, 2018
+
+SELECT 42;

qiita_db/support_files/patches/python_patches/66.py

@@ -0,0 +1,166 @@
+# August 6, 2018
+# Create parameters for the ssh/scp remote file upload commands
+
+
+from json import loads, dumps
+
+from qiita_db.sql_connection import TRN
+from qiita_db.software import Software, Command
+from qiita_db.exceptions import (QiitaDBError, QiitaDBDuplicateError)
+from qiita_db.util import convert_to_id
+
+
+# Copied from patch 58.py. Couldn't import due to how patching system works
+def create_command(software, name, description, parameters, outputs=None,
+                   analysis_only=False):
+    r"""Replicates the Command.create code at the time the patch was written"""
+    # Perform some sanity checks in the parameters dictionary
+    if not parameters:
+        raise QiitaDBError(
+            "Error creating command %s. At least one parameter should "
+            "be provided." % name)
+    sql_param_values = []
+    sql_artifact_params = []
+    for pname, vals in parameters.items():
+        if len(vals) != 2:
+            raise QiitaDBError(
+                "Malformed parameters dictionary, the format should be "
+                "{param_name: [parameter_type, default]}. Found: "
+                "%s for parameter name %s" % (vals, pname))
+
+        ptype, dflt = vals
+        # Check that the type is one of the supported types
+        supported_types = ['string', 'integer', 'float', 'reference',
+                           'boolean', 'prep_template', 'analysis']
+        if ptype not in supported_types and not ptype.startswith(
+                ('choice', 'mchoice', 'artifact')):
+            supported_types.extend(['choice', 'mchoice', 'artifact'])
+            raise QiitaDBError(
+                "Unsupported parameters type '%s' for parameter %s. "
+                "Supported types are: %s"
+                % (ptype, pname, ', '.join(supported_types)))
+
+        if ptype.startswith(('choice', 'mchoice')) and dflt is not None:
+            choices = set(loads(ptype.split(':')[1]))
+            dflt_val = dflt
+            if ptype.startswith('choice'):
+                # In the choice case, the dflt value is a single string,
+                # create a list with it the string on it to use the
+                # issuperset call below
+                dflt_val = [dflt_val]
+            else:
+                # jsonize the list to store it in the DB
+                dflt = dumps(dflt)
+            if not choices.issuperset(dflt_val):
+                raise QiitaDBError(
+                    "The default value '%s' for the parameter %s is not "
+                    "listed in the available choices: %s"
+                    % (dflt, pname, ', '.join(choices)))
+
+        if ptype.startswith('artifact'):
+            atypes = loads(ptype.split(':')[1])
+            sql_artifact_params.append(
+                [pname, 'artifact', atypes])
+        else:
+            if dflt is not None:
+                sql_param_values.append([pname, ptype, False, dflt])
+            else:
+                sql_param_values.append([pname, ptype, True, None])
+
+    with TRN:
+        sql = """SELECT EXISTS(SELECT *
+                               FROM qiita.software_command
+                               WHERE software_id = %s AND name = %s)"""
+        TRN.add(sql, [software.id, name])
+        if TRN.execute_fetchlast():
+            raise QiitaDBDuplicateError(
+                "command", "software: %d, name: %s"
+                           % (software.id, name))
+        # Add the command to the DB
+        sql = """INSERT INTO qiita.software_command
+                        (name, software_id, description, is_analysis)
+                 VALUES (%s, %s, %s, %s)
+                 RETURNING command_id"""
+        sql_params = [name, software.id, description, analysis_only]
+        TRN.add(sql, sql_params)
+        c_id = TRN.execute_fetchlast()
+
+        # Add the parameters to the DB
+        sql = """INSERT INTO qiita.command_parameter
+                    (command_id, parameter_name, parameter_type, required,
+                     default_value)
+                 VALUES (%s, %s, %s, %s, %s)
+                 RETURNING command_parameter_id"""
+        sql_params = [[c_id, pname, p_type, reqd, default]
+                      for pname, p_type, reqd, default in sql_param_values]
+        TRN.add(sql, sql_params, many=True)
+        TRN.execute()
+
+        # Add the artifact parameters
+        sql_type = """INSERT INTO qiita.parameter_artifact_type
+                        (command_parameter_id, artifact_type_id)
+                      VALUES (%s, %s)"""
+        supported_types = []
+        for pname, p_type, atypes in sql_artifact_params:
+            sql_params = [c_id, pname, p_type, True, None]
+            TRN.add(sql, sql_params)
+            pid = TRN.execute_fetchlast()
+            sql_params = [[pid, convert_to_id(at, 'artifact_type')]
+                          for at in atypes]
+            TRN.add(sql_type, sql_params, many=True)
+            supported_types.extend([atid for _, atid in sql_params])
+
+        # If the software type is 'artifact definition', there are a couple
+        # of extra steps
+        if software.type == 'artifact definition':
+            # If supported types is not empty, link the software with these
+            # types
+            if supported_types:
+                sql = """INSERT INTO qiita.software_artifact_type
+                                (software_id, artifact_type_id)
+                            VALUES (%s, %s)"""
+                sql_params = [[software.id, atid]
+                              for atid in supported_types]
+                TRN.add(sql, sql_params, many=True)
+            # If this is the validate command, we need to add the
+            # provenance and name parameters. These are used internally,
+            # that's why we are adding them here
+            if name == 'Validate':
+                sql = """INSERT INTO qiita.command_parameter
+                            (command_id, parameter_name, parameter_type,
+                             required, default_value)
+                         VALUES (%s, 'name', 'string', 'False',
+                                 'dflt_name'),
+                                (%s, 'provenance', 'string', 'False', NULL)
+                         """
+                TRN.add(sql, [c_id, c_id])
+
+        # Add the outputs to the command
+        if outputs:
+            sql = """INSERT INTO qiita.command_output
+                        (name, command_id, artifact_type_id)
+                     VALUES (%s, %s, %s)"""
+            sql_args = [[pname, c_id, convert_to_id(at, 'artifact_type')]
+                        for pname, at in outputs.items()]
+            TRN.add(sql, sql_args, many=True)
+            TRN.execute()
+
+    return Command(c_id)
+
+
+with TRN:
+    qiita_plugin = Software.from_name_and_version('Qiita', 'alpha')
+
+    # Create the 'list_remote_files' command
+    parameters = {'url': ['string', None],
+                  'private_key': ['string', None]}
+    create_command(qiita_plugin, "list_remote_files",
+                   "retrieves list of valid study files from remote dir",
+                   parameters)
+
+    # Create the 'download_remote_files' command
+    parameters = {'url': ['string', None],
+                  'destination': ['string', None],
+                  'private_key': ['string', None]}
+    create_command(qiita_plugin, "download_remote_files",
+                   "downloads valid study files from remote dir", parameters)

qiita_pet/templates/upload.html

@@ -59,7 +59,7 @@ <h3><u>Files</u></h3>
           </td>
           <td width="30px">&nbsp;</td>
           <td>
-            <button type="button" class="btn btn-light" onclick="$('input[name*=\'files_to_erase\']').removeAttr('checked');">Select None</button>
+            <button type="button" class="btn btn-light" onclick="$('input[name*=\'files_to_erase\']').removeAttr('checked');">Unselect All</button>
           </td>
         </tr>
       </table>

qiita_ware/commands.py

@@ -6,12 +6,16 @@
 # The full license is in the file LICENSE, distributed with this software.
 # -----------------------------------------------------------------------------
 
-from os.path import join, isdir
+from os.path import basename, isdir, join
 from shutil import rmtree
 from tarfile import open as taropen
 from tempfile import mkdtemp
 from os import environ
 from traceback import format_exc
+from paramiko import AutoAddPolicy, RSAKey, SSHClient
+from scp import SCPClient
+from urlparse import urlparse
+from functools import partial
 
 from qiita_db.artifact import Artifact
 from qiita_db.logger import LogEntry
@@ -21,6 +25,130 @@
 from qiita_ware.exceptions import ComputeError, EBISubmissionError
 
 
+def _ssh_session(p_url, private_key):
+    """Initializes an SSH session
+
+    Parameters
+    ----------
+    URL : urlparse object
+        a parsed url
+    private_key : str
+        Path to the private key used to authenticate connection
+
+    Returns
+    -------
+    paramiko.SSHClient
+        the SSH session
+    """
+    scheme = p_url.scheme
+    hostname = p_url.hostname
+    port = p_url.port
+    username = p_url.username
+
+    if scheme == 'scp' or scheme == 'sftp':
+
+        # if port not specified, use default 22 as port
+        if not port:
+            port = 22
+
+        # step 1: both schemes requires an SSH connection
+        ssh = SSHClient()
+        ssh.set_missing_host_key_policy(AutoAddPolicy)
+
+        # step 2: connect to fileserver
+        key = RSAKey.from_private_key_file(private_key)
+        ssh.connect(hostname, port=port, username=username,
+                    pkey=key, look_for_keys=False)
+        return ssh
+
+
+def _list_valid_files(ssh, directory):
+    """Gets a list of valid study files from ssh session
+
+    Parameters
+    ----------
+    ssh : paramiko.SSHClient
+        An initializeed ssh session
+    dir : the directory to search for files
+
+    Returns
+    -------
+    list of str
+        list of valid study files (basenames)
+    """
+
+    valid_file_extensions = tuple(qiita_config.valid_upload_extension)
+    sftp = ssh.open_sftp()
+    files = sftp.listdir(directory)
+
+    valid_files = [f for f in files if f.endswith(valid_file_extensions)]
+    sftp.close()
+    return valid_files
+
+
+def list_remote(URL, private_key):
+    """Retrieve a valid study files from a remote directory
+
+    Parameters
+    ----------
+    URL : str
+        The url to the remote directory
+    private_key : str
+        Path to the private key used to authenticate connection
+
+    Returns
+    -------
+    list of str
+        list of files that are valid study files
+    """
+    p_url = urlparse(URL)
+    directory = p_url.path
+    ssh = _ssh_session(p_url, private_key)
+    valid_files = _list_valid_files(ssh, directory)
+    ssh.close()
+    return valid_files
+
+
+def download_remote(URL, private_key, destination):
+    """Add study files by specifying a remote directory to download from
+
+    Parameters
+    ----------
+    URL : str
+        The url to the remote directory
+    destination : str
+        The path to the study upload folder
+    private_key : str
+        Path to the private key used to authenticate connection
+    """
+
+    # step 1: initialize connection and list valid files
+    p_url = urlparse(URL)
+    ssh = _ssh_session(p_url, private_key)
+
+    directory = p_url.path
+    valid_files = _list_valid_files(ssh, directory)
+    file_paths = [join(directory, f) for f in valid_files]
+
+    # step 2: download files
+    scheme = p_url.scheme
+    if scheme == 'scp':
+        scp = SCPClient(ssh.get_transport())
+        for f in file_paths:
+            download = partial(scp.get,
+                               local_path=join(destination, basename(f)))
+            download(f)
+    elif scheme == 'sftp':
+        sftp = ssh.open_sftp()
+        for f in file_paths:
+            download = partial(sftp.get,
+                               localpath=join(destination, basename(f)))
+            download(f)
+
+    # step 3: close the connection
+    ssh.close()
+
+
 def submit_EBI(artifact_id, action, send, test=False):
     """Submit an artifact to EBI