#!/usr/bin/env python
# encoding: utf-8
"""
@author: delagarza
"""
import sys
import os
import traceback
import ntpath
import string
from argparse import ArgumentParser
from argparse import RawDescriptionHelpFormatter
from collections import OrderedDict
from string import strip
from lxml import etree
from lxml.etree import SubElement, Element, ElementTree, ParseError, parse
from CTDopts.CTDopts import CTDModel, _InFile, _OutFile, ParameterGroup, _Choices, _NumericRange, _FileFormat, \
ModelError, _Null
__all__ = []
__version__ = 1.0
__date__ = '2014-09-17'
__updated__ = '2016-05-09'
MESSAGE_INDENTATION_INCREMENT = 2
TYPE_TO_GALAXY_TYPE = {int: 'integer', float: 'float', str: 'text', bool: 'boolean', _InFile: 'data',
_OutFile: 'data', _Choices: 'select'}
STDIO_MACRO_NAME = "stdio"
REQUIREMENTS_MACRO_NAME = "requirements"
ADVANCED_OPTIONS_MACRO_NAME = "advanced_options"
REQUIRED_MACROS = [STDIO_MACRO_NAME, REQUIREMENTS_MACRO_NAME, ADVANCED_OPTIONS_MACRO_NAME]
class CLIError(Exception):
# Generic exception to raise and log different fatal errors.
def __init__(self, msg):
super(CLIError).__init__(type(self))
self.msg = "E: %s" % msg
def __str__(self):
return self.msg
def __unicode__(self):
return self.msg
class InvalidModelException(ModelError):
def __init__(self, message):
super(InvalidModelException, self).__init__()
self.message = message
def __str__(self):
return self.message
def __repr__(self):
return self.message
class ApplicationException(Exception):
def __init__(self, msg):
super(ApplicationException).__init__(type(self))
self.msg = msg
def __str__(self):
return self.msg
def __unicode__(self):
return self.msg
class ExitCode:
def __init__(self, code_range="", level="", description=None):
self.range = code_range
self.level = level
self.description = description
class DataType:
def __init__(self, extension, galaxy_extension=None, galaxy_type=None, mimetype=None):
self.extension = extension
self.galaxy_extension = galaxy_extension
self.galaxy_type = galaxy_type
self.mimetype = mimetype
class ParameterHardcoder:
def __init__(self):
# map whose keys are the composite names of tools and parameters in the following pattern:
# [ToolName][separator][ParameterName] -> HardcodedValue
# if the parameter applies to all tools, then the following pattern is used:
# [ParameterName] -> HardcodedValue
# examples (assuming separator is '#'):
# threads -> 24
# XtandemAdapter#adapter -> xtandem.exe
# adapter -> adapter.exe
self.separator = "!"
self.parameter_map = {}
# the most specific value will be returned in case of overlap
def get_hardcoded_value(self, parameter_name, tool_name):
# look for the value that would apply for all tools
generic_value = self.parameter_map.get(parameter_name, None)
specific_value = self.parameter_map.get(self.build_key(parameter_name, tool_name), None)
if specific_value is not None:
return specific_value
return generic_value
def register_parameter(self, parameter_name, parameter_value, tool_name=None):
self.parameter_map[self.build_key(parameter_name, tool_name)] = parameter_value
def build_key(self, parameter_name, tool_name):
if tool_name is None:
return parameter_name
return "%s%s%s" % (parameter_name, self.separator, tool_name)
def main(argv=None): # IGNORE:C0111
# Command line options.
if argv is None:
argv = sys.argv
else:
sys.argv.extend(argv)
program_version = "v%s" % __version__
program_build_date = str(__updated__)
program_version_message = '%%(prog)s %s (%s)' % (program_version, program_build_date)
program_short_description = "CTD2Galaxy - A project from the GenericWorkflowNodes family " \
"(https://github.com/orgs/genericworkflownodes)"
program_usage = '''
USAGE:
I - Parsing a single CTD file and generate a Galaxy wrapper:
$ python generator.py -i input.ctd -o output.xml
II - Parsing all found CTD files (files with .ctd and .xml extension) in a given folder and
output converted Galaxy wrappers in a given folder:
$ python generator.py -i /home/user/*.ctd -o /home/user/galaxywrappers
III - Providing file formats, mimetypes
Galaxy supports the concept of file format in order to connect compatible ports, that is, input ports of a certain
data format will be able to receive data from a port from the same format. This converter allows you to provide
a personalized file in which you can relate the CTD data formats with supported Galaxy data formats. The layout of
this file consists of lines, each of either one or four columns separated by any amount of whitespace. The content
of each column is as follows:
* 1st column: file extension
* 2nd column: data type, as listed in Galaxy
* 3rd column: full-named Galaxy data type, as it will appear on datatypes_conf.xml
* 4th column: mimetype (optional)
The following is an example of a valid "file formats" file:
########################################## FILE FORMATS example ##########################################
# Every line starting with a # will be handled as a comment and will not be parsed.
# The first column is the file format as given in the CTD and second column is the Galaxy data format.
# The second, third, fourth and fifth column can be left empty if the data type has already been registered
# in Galaxy, otherwise, all but the mimetype must be provided.
# CTD type # Galaxy type # Long Galaxy data type # Mimetype
csv tabular galaxy.datatypes.data:Text
fasta
ini txt galaxy.datatypes.data:Text
txt
idxml txt galaxy.datatypes.xml:GenericXml application/xml
options txt galaxy.datatypes.data:Text
grid grid galaxy.datatypes.data:Grid
##########################################################################################################
Note that each line consists precisely of either one, three or four columns. In the case of data types already
registered in Galaxy (such as fasta and txt in the above example), only the first column is needed. In the case of
data types that haven't been yet registered in Galaxy, the first three columns are needed (mimetype is optional).
For information about Galaxy data types and subclasses, see the following page:
https://wiki.galaxyproject.org/Admin/Datatypes/Adding%20Datatypes
IV - Hardcoding parameters
It is possible to hardcode parameters. This makes sense if you want to set a tool in Galaxy in 'quiet' mode or if
your tools support multi-threading and accept the number of threads via a parameter, without giving the end user the
chance to change the values for these parameters.
In order to generate hardcoded parameters, you need to provide a simple file. Each line of this file contains two
or three columns separated by whitespace. Any line starting with a '#' will be ignored. The first column contains
the name of the parameter, the second column contains the value that will always be set for this parameter. The
first two columns are mandatory.
If the parameter is to be hardcoded only for a set of tools, then a third column can be added. This column includes
a comma-separated list of tool names for which the parameter will be hardcoded. If a third column is not included,
then all processed tools containing the given parameter will get a hardcoded value for it.
The following is an example of a valid file:
##################################### HARDCODED PARAMETERS example #####################################
# Every line starting with a # will be handled as a comment and will not be parsed.
# The first column is the name of the parameter and the second column is the value that will be used.
# Parameter name # Value # Tool(s)
threads \${GALAXY_SLOTS:-24}
mode quiet
xtandem_executable xtandem XTandemAdapter
verbosity high Foo, Bar
#########################################################################################################
Using the above file will produce a <command> similar to:
[tool_name] ... -threads \${GALAXY_SLOTS:-24} -mode quiet ...
For all tools. For XTandemAdapter, the <command> will be similar to:
XtandemAdapter ... -threads \${GALAXY_SLOTS:-24} -mode quiet -xtandem_executable xtandem ...
And for tools Foo and Bar, the <command> will be similar to:
Foo ... ... -threads \${GALAXY_SLOTS:-24} -mode quiet -verbosity high ...
V - Control which tools will be converted
Sometimes only a subset of CTDs needs to be converted. It is possible to either explicitly specify which tools will
be converted or which tools will not be converted.
The value of the -s/--skip-tools parameter is a file in which each line will be interpreted as the name of a tool
that will not be converted. Conversely, the value of the -r/--required-tools is a file in which each line will be
interpreted as a tool that is required. Only one of these parameters can be specified at a given time.
The format of both files is exactly the same. As stated before, each line will be interpreted as the name of a tool;
any line starting with a '#' will be ignored.
'''
program_license = '''%(short_description)s
Copyright 2015, Luis de la Garza
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
%(usage)s
''' % {'short_description': program_short_description, 'usage': program_usage}
try:
# Setup argument parser
parser = ArgumentParser(prog="CTD2Galaxy", description=program_license,
formatter_class=RawDescriptionHelpFormatter, add_help=True)
parser.add_argument("-i", "--input", dest="input_files", default=[], required=True, nargs="+", action="append",
help="List of CTD files to convert.")
parser.add_argument("-o", "--output-destination", dest="output_destination", required=True,
help="If multiple input files are given, then a folder in which all generated "
"XMLs will be generated is expected;"
"if a single input file is given, then a destination file is expected.")
parser.add_argument("-f", "--formats-file", dest="formats_file",
help="File containing the supported file formats. Run with '-h' or '--help' to see a "
"brief example on the layout of this file.", default=None, required=False)
parser.add_argument("-a", "--add-to-command-line", dest="add_to_command_line",
help="Adds content to the command line", default="", required=False)
parser.add_argument("-d", "--datatypes-destination", dest="data_types_destination",
help="Specify the location of a datatypes_conf.xml to modify and add the registered "
"data types. If the provided destination does not exist, a new file will be created.",
default=None, required=False)
parser.add_argument("-x", "--default-executable-path", dest="default_executable_path",
help="Use this executable path when <executablePath> is not present in the CTD",
default=None, required=False)
parser.add_argument("-b", "--blacklist-parameters", dest="blacklisted_parameters", default=[], nargs="+", action="append",
help="List of parameters that will be ignored and won't appear on the galaxy stub",
required=False)
parser.add_argument("-c", "--default-category", dest="default_category", default="DEFAULT", required=False,
help="Default category to use for tools lacking a category when generating tool_conf.xml")
parser.add_argument("-t", "--tool-conf-destination", dest="tool_conf_destination", default=None, required=False,
help="Specify the location of an existing tool_conf.xml that will be modified to include "
"the converted tools. If the provided destination does not exist, a new file will"
"be created.")
parser.add_argument("-g", "--galaxy-tool-path", dest="galaxy_tool_path", default=None, required=False,
help="The path that will be prepended to the file names when generating tool_conf.xml")
parser.add_argument("-r", "--required-tools", dest="required_tools_file", default=None, required=False,
help="Each line of the file will be interpreted as a tool name that needs translation. "
"Run with '-h' or '--help' to see a brief example on the format of this file.")
parser.add_argument("-s", "--skip-tools", dest="skip_tools_file", default=None, required=False,
help="File containing a list of tools for which a Galaxy stub will not be generated. "
"Run with '-h' or '--help' to see a brief example on the format of this file.")
parser.add_argument("-m", "--macros", dest="macros_files", default=[], nargs="*",
action="append", required=None, help="Import the additional given file(s) as macros. "
"The macros stdio, requirements and advanced_options are required. Please see "
"macros.xml for an example of a valid macros file. Al defined macros will be imported.")
parser.add_argument("-p", "--hardcoded-parameters", dest="hardcoded_parameters", default=None, required=False,
help="File containing hardcoded values for the given parameters. Run with '-h' or '--help' "
"to see a brief example on the format of this file.")
parser.add_argument("-v", "--validation-schema", dest="xsd_location", default=None, required=False,
help="Location of the schema to use to validate CTDs.")
# TODO: add verbosity, maybe?
parser.add_argument("-V", "--version", action='version', version=program_version_message)
# Process arguments
args = parser.parse_args()
# validate and prepare the passed arguments
validate_and_prepare_args(args)
# extract the names of the macros and check that we have found the ones we need
macros_to_expand = parse_macros_files(args.macros_files)
# parse the given supported file-formats file
supported_file_formats = parse_file_formats(args.formats_file)
# parse the hardcoded parameters file¬
parameter_hardcoder = parse_hardcoded_parameters(args.hardcoded_parameters)
# parse the skip/required tools files
skip_tools = parse_tools_list_file(args.skip_tools_file)
required_tools = parse_tools_list_file(args.required_tools_file)
#if verbose > 0:
# print("Verbose mode on")
parsed_models = convert(args.input_files,
args.output_destination,
supported_file_formats=supported_file_formats,
default_executable_path=args.default_executable_path,
add_to_command_line=args.add_to_command_line,
blacklisted_parameters=args.blacklisted_parameters,
required_tools=required_tools,
skip_tools=skip_tools,
macros_file_names=args.macros_files,
macros_to_expand=macros_to_expand,
parameter_hardcoder=parameter_hardcoder,
xsd_location=args.xsd_location)
#TODO: add some sort of warning if a macro that doesn't exist is to be expanded
# it is not needed to copy the macros files, since the user has provided them
# generation of galaxy stubs is ready... now, let's see if we need to generate a tool_conf.xml
if args.tool_conf_destination is not None:
generate_tool_conf(parsed_models, args.tool_conf_destination,
args.galaxy_tool_path, args.default_category)
# now datatypes_conf.xml
if args.data_types_destination is not None:
generate_data_type_conf(supported_file_formats, args.data_types_destination)
return 0
except KeyboardInterrupt:
# handle keyboard interrupt
return 0
except ApplicationException, e:
error("CTD2Galaxy could not complete the requested operation.", 0)
error("Reason: " + e.msg, 0)
return 1
except ModelError, e:
error("There seems to be a problem with one of your input CTDs.", 0)
error("Reason: " + e.msg, 0)
return 1
except Exception, e:
traceback.print_exc()
return 2
def parse_tools_list_file(tools_list_file):
tools_list = None
if tools_list_file is not None:
tools_list = []
with open(tools_list_file) as f:
for line in f:
if line is None or not line.strip() or line.strip().startswith("#"):
continue
else:
tools_list.append(line.strip())
return tools_list
def parse_macros_files(macros_file_names):
macros_to_expand = set()
for macros_file_name in macros_file_names:
try:
macros_file = open(macros_file_name)
info("Loading macros from %s" % macros_file_name, 0)
root = parse(macros_file).getroot()
for xml_element in root.findall("xml"):
name = xml_element.attrib["name"]
if name in macros_to_expand:
warning("Macro %s has already been found. Duplicate found in file %s." %
(name, macros_file_name), 0)
else:
info("Macro %s found" % name, 1)
macros_to_expand.add(name)
except ParseError, e:
raise ApplicationException("The macros file " + macros_file_name + " could not be parsed. Cause: " +
str(e))
except IOError, e:
raise ApplicationException("The macros file " + macros_file_name + " could not be opened. Cause: " +
str(e))
# we depend on "stdio", "requirements" and "advanced_options" to exist on all the given macros files
missing_needed_macros = []
for required_macro in REQUIRED_MACROS:
if required_macro not in macros_to_expand:
missing_needed_macros.append(required_macro)
if missing_needed_macros:
raise ApplicationException(
"The following required macro(s) were not found in any of the given macros files: %s, "
"see sample_files/macros.xml for an example of a valid macros file."
% ", ".join(missing_needed_macros))
# we do not need to "expand" the advanced_options macro
macros_to_expand.remove(ADVANCED_OPTIONS_MACRO_NAME)
return macros_to_expand
def parse_hardcoded_parameters(hardcoded_parameters_file):
parameter_hardcoder = ParameterHardcoder()
if hardcoded_parameters_file is not None:
line_number = 0
with open(hardcoded_parameters_file) as f:
for line in f:
line_number += 1
if line is None or not line.strip() or line.strip().startswith("#"):
pass
else:
# the third column must not be obtained as a whole, and not split
parsed_hardcoded_parameter = line.strip().split(None, 2)
# valid lines contain two or three columns
if len(parsed_hardcoded_parameter) != 2 and len(parsed_hardcoded_parameter) != 3:
warning("Invalid line at line number %d of the given hardcoded parameters file. Line will be"
"ignored:\n%s" % (line_number, line), 0)
continue
parameter_name = parsed_hardcoded_parameter[0]
hardcoded_value = parsed_hardcoded_parameter[1]
tool_names = None
if len(parsed_hardcoded_parameter) == 3:
tool_names = parsed_hardcoded_parameter[2].split(',')
if tool_names:
for tool_name in tool_names:
parameter_hardcoder.register_parameter(parameter_name, hardcoded_value, tool_name.strip())
else:
parameter_hardcoder.register_parameter(parameter_name, hardcoded_value)
return parameter_hardcoder
def parse_file_formats(formats_file):
supported_formats = {}
if formats_file is not None:
line_number = 0
with open(formats_file) as f:
for line in f:
line_number += 1
if line is None or not line.strip() or line.strip().startswith("#"):
# ignore (it'd be weird to have something like:
# if line is not None and not (not line.strip()) ...
pass
else:
# not an empty line, no comment
# strip the line and split by whitespace
parsed_formats = line.strip().split()
# valid lines contain either one or four columns
if not (len(parsed_formats) == 1 or len(parsed_formats) == 3 or len(parsed_formats) == 4):
warning("Invalid line at line number %d of the given formats file. Line will be ignored:\n%s" %
(line_number, line), 0)
# ignore the line
continue
elif len(parsed_formats) == 1:
supported_formats[parsed_formats[0]] = DataType(parsed_formats[0], parsed_formats[0])
else:
mimetype = None
# check if mimetype was provided
if len(parsed_formats) == 4:
mimetype = parsed_formats[3]
supported_formats[parsed_formats[0]] = DataType(parsed_formats[0], parsed_formats[1],
parsed_formats[2], mimetype)
return supported_formats
def validate_and_prepare_args(args):
# check that only one of skip_tools_file and required_tools_file has been provided
if args.skip_tools_file is not None and args.required_tools_file is not None:
raise ApplicationException(
"You have provided both a file with tools to ignore and a file with required tools.\n"
"Only one of -s/--skip-tools, -r/--required-tools can be provided.")
# first, we convert all list of lists in args to flat lists
lists_to_flatten = ["input_files", "blacklisted_parameters", "macros_files"]
for list_to_flatten in lists_to_flatten:
setattr(args, list_to_flatten, [item for sub_list in getattr(args, list_to_flatten) for item in sub_list])
# if input is a single file, we expect output to be a file (and not a dir that already exists)
if len(args.input_files) == 1:
if os.path.isdir(args.output_destination):
raise ApplicationException("If a single input file is provided, output (%s) is expected to be a file "
"and not a folder.\n" % args.output_destination)
# if input is a list of files, we expect output to be a folder
if len(args.input_files) > 1:
if not os.path.isdir(args.output_destination):
raise ApplicationException("If several input files are provided, output (%s) is expected to be an "
"existing directory.\n" % args.output_destination)
# check that the provided input files, if provided, contain a valid file path
input_variables_to_check = ["skip_tools_file", "required_tools_file", "macros_files", "xsd_location",
"input_files", "formats_file", "hardcoded_parameters"]
for variable_name in input_variables_to_check:
paths_to_check = []
# check if we are handling a single file or a list of files
member_value = getattr(args, variable_name)
if member_value is not None:
if isinstance(member_value, list):
for file_name in member_value:
paths_to_check.append(strip(str(file_name)))
else:
paths_to_check.append(strip(str(member_value)))
for path_to_check in paths_to_check:
if not os.path.isfile(path_to_check) or not os.path.exists(path_to_check):
raise ApplicationException(
"The provided input file (%s) does not exist or is not a valid file path."
% path_to_check)
# check that the provided output files, if provided, contain a valid file path (i.e., not a folder)
output_variables_to_check = ["data_types_destination", "tool_conf_destination"]
for variable_name in output_variables_to_check:
file_name = getattr(args, variable_name)
if file_name is not None and os.path.isdir(file_name):
raise ApplicationException("The provided output file name (%s) points to a directory." % file_name)
if not args.macros_files:
# list is empty, provide the default value
warning("Using default macros from macros.xml", 0)
args.macros_files = ["macros.xml"]
def convert(input_files, output_destination, **kwargs):
# first, generate a model
is_converting_multiple_ctds = len(input_files) > 1
parsed_models = []
schema = None
if kwargs["xsd_location"] is not None:
try:
info("Loading validation schema from %s" % kwargs["xsd_location"], 0)
schema = etree.XMLSchema(etree.parse(kwargs["xsd_location"]))
except Exception, e:
error("Could not load validation schema %s. Reason: %s" % (kwargs["xsd_location"], str(e)), 0)
else:
info("Validation against a schema has not been enabled.", 0)
for input_file in input_files:
try:
if schema is not None:
validate_against_schema(input_file, schema)
model = CTDModel(from_file=input_file)
except Exception, e:
error(str(e), 1)
continue
if kwargs["skip_tools"] is not None and model.name in kwargs["skip_tools"]:
info("Skipping tool %s" % model.name, 0)
continue
elif kwargs["required_tools"] is not None and model.name not in kwargs["required_tools"]:
info("Tool %s is not required, skipping it" % model.name, 0)
continue
else:
info("Converting from %s " % input_file, 0)
tool = create_tool(model)
write_header(tool, model)
create_description(tool, model)
expand_macros(tool, model, **kwargs)
create_command(tool, model, **kwargs)
create_inputs(tool, model, **kwargs)
create_outputs(tool, model, **kwargs)
create_help(tool, model)
# finally, serialize the tool
output_file = output_destination
# if multiple inputs are being converted,
# then we need to generate a different output_file for each input
if is_converting_multiple_ctds:
output_file = os.path.join(output_file, get_filename_without_suffix(input_file) + ".xml")
# wrap our tool element into a tree to be able to serialize it
tree = ElementTree(tool)
tree.write(open(output_file, 'w'), encoding="UTF-8", xml_declaration=True, pretty_print=True)
# let's use model to hold the name of the output file
parsed_models.append([model, get_filename(output_file)])
return parsed_models
# validates a ctd file against the schema
def validate_against_schema(ctd_file, schema):
try:
parser = etree.XMLParser(schema=schema)
etree.parse(ctd_file, parser=parser)
except etree.XMLSyntaxError, e:
raise ApplicationException("Input ctd file %s is not valid. Reason: %s" % (ctd_file, str(e)))
def write_header(tool, model):
tool.addprevious(etree.Comment(
"This is a configuration file for the integration of a tools into Galaxy (https://galaxyproject.org/). "
"This file was automatically generated using CTD2Galaxy."))
tool.addprevious(etree.Comment('Proposed Tool Section: [%s]' % model.opt_attribs.get("category", "")))
def generate_tool_conf(parsed_models, tool_conf_destination, galaxy_tool_path, default_category):
# for each category, we keep a list of models corresponding to it
categories_to_tools = dict()
for model in parsed_models:
category = strip(model[0].opt_attribs.get("category", ""))
if not category.strip():
category = default_category
if category not in categories_to_tools:
categories_to_tools[category] = []
categories_to_tools[category].append(model[1])
# at this point, we should have a map for all categories->tools
toolbox_node = Element("toolbox")
if galaxy_tool_path is not None and not galaxy_tool_path.strip().endswith("/"):
galaxy_tool_path = galaxy_tool_path.strip() + "/"
if galaxy_tool_path is None:
galaxy_tool_path = ""
for category, file_names in categories_to_tools.iteritems():
section_node = add_child_node(toolbox_node, "section")
section_node.attrib["id"] = "section-id-" + "".join(category.split())
section_node.attrib["name"] = category
for filename in file_names:
tool_node = add_child_node(section_node, "tool")
tool_node.attrib["file"] = galaxy_tool_path + filename
toolconf_tree = ElementTree(toolbox_node)
toolconf_tree.write(open(tool_conf_destination,'w'), encoding="UTF-8", xml_declaration=True, pretty_print=True)
info("Generated Galaxy tool_conf.xml in %s" % tool_conf_destination, 0)
def generate_data_type_conf(supported_file_formats, data_types_destination):
data_types_node = Element("datatypes")
registration_node = add_child_node(data_types_node, "registration")
registration_node.attrib["converters_path"] = "lib/galaxy/datatypes/converters"
registration_node.attrib["display_path"] = "display_applications"
for format_name in supported_file_formats:
data_type = supported_file_formats[format_name]
# add only if it's a data type that does not exist in Galaxy
if data_type.galaxy_type is not None:
data_type_node = add_child_node(registration_node, "datatype")
# we know galaxy_extension is not None
data_type_node.attrib["extension"] = data_type.galaxy_extension
data_type_node.attrib["type"] = data_type.galaxy_type
if data_type.mimetype is not None:
data_type_node.attrib["mimetype"] = data_type.mimetype
data_types_tree = ElementTree(data_types_node)
data_types_tree.write(open(data_types_destination,'w'), encoding="UTF-8", xml_declaration=True, pretty_print=True)
info("Generated Galaxy datatypes_conf.xml in %s" % data_types_destination, 0)
# taken from
# http://stackoverflow.com/questions/8384737/python-extract-file-name-from-path-no-matter-what-the-os-path-format
def get_filename(path):
head, tail = ntpath.split(path)
return tail or ntpath.basename(head)
def get_filename_without_suffix(path):
root, ext = os.path.splitext(os.path.basename(path))
return root
def create_tool(model):
return Element("tool", OrderedDict([("id", model.name), ("name", model.name), ("version", model.version)]))
def create_description(tool, model):
if "description" in model.opt_attribs.keys() and model.opt_attribs["description"] is not None:
description = SubElement(tool,"description")
description.text = model.opt_attribs["description"]
def get_param_cli_name(param, model):
# we generate parameters with colons for subgroups, but not for the two topmost parents (OpenMS legacy)
if type(param.parent) == ParameterGroup:
if not hasattr(param.parent.parent, 'parent'):
return resolve_param_mapping(param, model)
elif not hasattr(param.parent.parent.parent, 'parent'):
return resolve_param_mapping(param, model)
else:
if model.cli:
warning("Using nested parameter sections (NODE elements) is not compatible with <cli>", py1)
return get_param_name(param.parent) + ":" + resolve_param_mapping(param, model)
else:
return resolve_param_mapping(param, model)
def get_param_name(param):
# we generate parameters with colons for subgroups, but not for the two topmost parents (OpenMS legacy)
if type(param.parent) == ParameterGroup:
if not hasattr(param.parent.parent, 'parent'):
return param.name
elif not hasattr(param.parent.parent.parent, 'parent'):
return param.name
else:
return get_param_name(param.parent) + ":" + param.name
else:
return param.name
# some parameters are mapped to command line options, this method helps resolve those mappings, if any
def resolve_param_mapping(param, model):
# go through all mappings and find if the given param appears as a reference name in a mapping element
param_mapping = None
for cli_element in model.cli:
for mapping_element in cli_element.mappings:
if mapping_element.reference_name == param.name:
if param_mapping is not None:
warning("The parameter %s has more than one mapping in the <cli> section. "
"The first found mapping, %s, will be used." % (param.name, param_mapping), 1)
else:
param_mapping = cli_element.option_identifier
return param_mapping if param_mapping is not None else param.name
def create_command(tool, model, **kwargs):
final_command = get_tool_executable_path(model, kwargs["default_executable_path"]) + '\n'
final_command += kwargs["add_to_command_line"] + '\n'
advanced_command_start = "#if $adv_opts.adv_opts_selector=='advanced':\n"
advanced_command_end = '#end if'
advanced_command = ''
parameter_hardcoder = kwargs["parameter_hardcoder"]
found_output_parameter = False
for param in extract_parameters(model):
if param.type is _OutFile:
found_output_parameter = True
command = ''
param_name = get_param_name(param)
param_cli_name = get_param_cli_name(param, model)
if param_name == param_cli_name:
# there was no mapping, so for the cli name we will use a '-' in the prefix
param_cli_name = '-' + param_name
if param.name in kwargs["blacklisted_parameters"]:
continue
hardcoded_value = parameter_hardcoder.get_hardcoded_value(param_name, model.name)
if hardcoded_value:
command += '%s %s\n' % (param_cli_name, hardcoded_value)
else:
# parameter is neither blacklisted nor hardcoded...
galaxy_parameter_name = get_galaxy_parameter_name(param)
repeat_galaxy_parameter_name = get_repeat_galaxy_parameter_name(param)
# logic for ITEMLISTs
if param.is_list:
if param.type is _InFile:
command += param_cli_name + "\n"
command += " #for token in $" + galaxy_parameter_name + ":\n"
command += " $token\n"
command += " #end for\n"
else:
command += "\n#if $" + repeat_galaxy_parameter_name + ":\n"
command += param_cli_name + "\n"
command += " #for token in $" + repeat_galaxy_parameter_name + ":\n"
command += " #if \" \" in str(token):\n"
command += " \"$token." + galaxy_parameter_name + "\"\n"
command += " #else\n"
command += " $token." + galaxy_parameter_name + "\n"
command += " #end if\n"
command += " #end for\n"
command += "#end if\n"
# logic for other ITEMs
else:
if param.advanced and param.type is not _OutFile:
actual_parameter = "$adv_opts.%s" % galaxy_parameter_name
else:
actual_parameter = "$%s" % galaxy_parameter_name
## if whitespace_validation has been set, we need to generate, for each parameter:
## #if str( $t ).split() != '':
## -t "$t"
## #end if
## TODO only useful for text fields, integers or floats
## not useful for choices, input fields ...
if not is_boolean_parameter(param) and type(param.restrictions) is _Choices :
command += "#if " + actual_parameter + ":\n"
command += ' %s\n' % param_cli_name
command += " #if \" \" in str(" + actual_parameter + "):\n"
command += " \"" + actual_parameter + "\"\n"
command += " #else\n"
command += " " + actual_parameter + "\n"
command += " #end if\n"
command += "#end if\n"
elif is_boolean_parameter(param):
command += "#if " + actual_parameter + ":\n"
command += ' %s\n' % param_cli_name
command += "#end if\n"
elif TYPE_TO_GALAXY_TYPE[param.type] is 'text':
command += "#if " + actual_parameter + ":\n"
command += " %s " % param_cli_name
command += " \"" + actual_parameter + "\"\n"
command += "#end if\n"
else:
command += "#if " + actual_parameter + ":\n"
command += ' %s ' % param_cli_name
command += actual_parameter + "\n"
command += "#end if\n"
if param.advanced and param.type is not _OutFile:
advanced_command += " %s" % command
else:
final_command += command
if advanced_command:
final_command += "%s%s%s\n" % (advanced_command_start, advanced_command, advanced_command_end)
if not found_output_parameter:
final_command += "> $param_stdout\n"
command_node = add_child_node(tool, "command")
command_node.text = final_command
# creates the xml elements needed to import the needed macros files
# and to "expand" the macros
def expand_macros(tool, model, **kwargs):
macros_node = add_child_node(tool, "macros")
token_node = add_child_node(macros_node, "token")
token_node.attrib["name"] = "@EXECUTABLE@"
token_node.text = get_tool_executable_path(model, kwargs["default_executable_path"])
# add <import> nodes
for macro_file_name in kwargs["macros_file_names"]:
macro_file = open(macro_file_name)
import_node = add_child_node(macros_node, "import")
# do not add the path of the file, rather, just its basename
import_node.text = os.path.basename(macro_file.name)
# add <expand> nodes
for expand_macro in kwargs["macros_to_expand"]:
expand_node = add_child_node(tool, "expand")
expand_node.attrib["macro"] = expand_macro
def get_tool_executable_path(model, default_executable_path):
# rules to build the galaxy executable path:
# if executablePath is null, then use default_executable_path and store it in executablePath
# if executablePath is null and executableName is null, then the name of the tool will be used
# if executablePath is null and executableName is not null, then executableName will be used
# if executablePath is not null and executableName is null,
# then executablePath and the name of the tool will be used
# if executablePath is not null and executableName is not null, then both will be used
# first, check if the model has executablePath / executableName defined
executable_path = model.opt_attribs.get("executablePath", None)
executable_name = model.opt_attribs.get("executableName", None)
# check if we need to use the default_executable_path
if executable_path is None:
executable_path = default_executable_path
# fix the executablePath to make sure that there is a '/' in the end
if executable_path is not None:
executable_path = executable_path.strip()
if not executable_path.endswith('/'):
executable_path += '/'
# assume that we have all information present
command = str(executable_path) + str(executable_name)
if executable_path is None:
if executable_name is None:
command = model.name
else:
command = executable_name
else:
if executable_name is None:
command = executable_path + model.name
return command
def get_galaxy_parameter_name(param):
return "param_%s" % get_param_name(param).replace(':', '_').replace('-', '_')
def get_input_with_same_restrictions(out_param, model, supported_file_formats):
for param in extract_parameters(model):
if param.type is _InFile:
if param.restrictions is not None:
in_param_formats = get_supported_file_types(param.restrictions.formats, supported_file_formats)
out_param_formats = get_supported_file_types(out_param.restrictions.formats, supported_file_formats)
if in_param_formats == out_param_formats:
return param
def create_inputs(tool, model, **kwargs):
inputs_node = SubElement(tool, "inputs")
# some suites (such as OpenMS) need some advanced options when handling inputs
expand_advanced_node = add_child_node(tool, "expand", OrderedDict([("macro", ADVANCED_OPTIONS_MACRO_NAME)]))
parameter_hardcoder = kwargs["parameter_hardcoder"]
# treat all non output-file parameters as inputs
for param in extract_parameters(model):
# no need to show hardcoded parameters
hardcoded_value = parameter_hardcoder.get_hardcoded_value(param.name, model.name)
if param.name in kwargs["blacklisted_parameters"] or hardcoded_value:
# let's not use an extra level of indentation and use NOP
continue
if param.type is not _OutFile:
if param.advanced:
if expand_advanced_node is not None:
parent_node = expand_advanced_node
else:
# something went wrong... we are handling an advanced parameter and the
# advanced input macro was not set... inform the user about it
info("The parameter %s has been set as advanced, but advanced_input_macro has "
"not been set." % param.name, 1)
# there is not much we can do, other than use the inputs_node as a parent node!
parent_node = inputs_node
else:
parent_node = inputs_node
# for lists we need a repeat tag
if param.is_list and param.type is not _InFile:
rep_node = add_child_node(parent_node, "repeat")
create_repeat_attribute_list(rep_node, param)
parent_node = rep_node
param_node = add_child_node(parent_node, "param")
create_param_attribute_list(param_node, param, kwargs["supported_file_formats"])
# advanced parameter selection should be at the end
# and only available if an advanced parameter exists
if expand_advanced_node is not None and len(expand_advanced_node) > 0:
inputs_node.append(expand_advanced_node)
def get_repeat_galaxy_parameter_name(param):
return "rep_" + get_galaxy_parameter_name(param)
def create_repeat_attribute_list(rep_node, param):
rep_node.attrib["name"] = get_repeat_galaxy_parameter_name(param)
if param.required:
rep_node.attrib["min"] = "1"
else:
rep_node.attrib["min"] = "0"
# for the ITEMLISTs which have LISTITEM children we only
# need one parameter as it is given as a string
if param.default is not None:
rep_node.attrib["max"] = "1"
rep_node.attrib["title"] = get_galaxy_parameter_name(param)
def create_param_attribute_list(param_node, param, supported_file_formats):
param_node.attrib["name"] = get_galaxy_parameter_name(param)
param_type = TYPE_TO_GALAXY_TYPE[param.type]
if param_type is None:
raise ModelError("Unrecognized parameter type %(type)s for parameter %(name)s"
% {"type": param.type, "name": param.name})
if param.is_list:
param_type = "text"
if is_selection_parameter(param):
param_type = "select"
if len(param.restrictions.choices) < 5:
param_node.attrib["display"] = "radio"
if is_boolean_parameter(param):
param_type = "boolean"
if param.type is _InFile:
# assume it's just text unless restrictions are provided
param_format = "text"
if param.restrictions is not None:
# join all formats of the file, take mapping from supported_file if available for an entry
if type(param.restrictions) is _FileFormat:
param_format = ','.join([get_supported_file_types(i, supported_file_formats) if
get_supported_file_types(i, supported_file_formats)
else i for i in param.restrictions.formats])
else:
raise InvalidModelException("Expected 'file type' restrictions for input file [%(name)s], "
"but instead got [%(type)s]"
% {"name": param.name, "type": type(param.restrictions)})
param_node.attrib["type"] = "data"
param_node.attrib["format"] = param_format
# in the case of multiple input set multiple flag
if param.is_list:
param_node.attrib["multiple"] = "true"
else:
param_node.attrib["type"] = param_type
# check for parameters with restricted values (which will correspond to a "select" in galaxy)
if param.restrictions is not None:
# it could be either _Choices or _NumericRange, with special case for boolean types
if param_type == "boolean":
create_boolean_parameter(param_node, param)
elif type(param.restrictions) is _Choices:
# create as many <option> elements as restriction values
for choice in param.restrictions.choices:
option_node = add_child_node(param_node, "option", OrderedDict([("value", str(choice))]))
option_node.text = str(choice)
# preselect the default value
if param.default == choice:
option_node.attrib["selected"] = "true"
elif type(param.restrictions) is _NumericRange:
if param.type is not int and param.type is not float:
raise InvalidModelException("Expected either 'int' or 'float' in the numeric range restriction for "
"parameter [%(name)s], but instead got [%(type)s]" %
{"name": param.name, "type": type(param.restrictions)})
# extract the min and max values and add them as attributes
# validate the provided min and max values
if param.restrictions.n_min is not None:
param_node.attrib["min"] = str(param.restrictions.n_min)
if param.restrictions.n_max is not None:
param_node.attrib["max"] = str(param.restrictions.n_max)
elif type(param.restrictions) is _FileFormat:
param_node.attrib["format"] = ",".join(
get_supported_file_types(param.restrictions.formats, supported_file_formats))
else:
raise InvalidModelException("Unrecognized restriction type [%(type)s] for parameter [%(name)s]"
% {"type": type(param.restrictions), "name": param.name})
if param_type == "select" and param.default in param.restrictions.choices:
param_node.attrib["optional"] = "False"
else:
param_node.attrib["optional"] = str(not param.required)
if param_type == "text":
# add size attribute... this is the length of a textbox field in Galaxy (it could also be 15x2, for instance)
param_node.attrib["size"] = "30"
# add sanitizer nodes, this is needed for special character like "["
# which are used for example by FeatureFinderMultiplex
sanitizer_node = SubElement(param_node, "sanitizer")
valid_node = SubElement(sanitizer_node, "valid", OrderedDict([("initial", "string.printable")]))
add_child_node(valid_node, "remove", OrderedDict([("value", '\'')]))
add_child_node(valid_node, "remove", OrderedDict([("value", '"')]))
# check for default value
if param.default is not None and param.default is not _Null:
if type(param.default) is list:
# we ASSUME that a list of parameters looks like:
# $ tool -ignore He Ar Xe
# meaning, that, for example, Helium, Argon and Xenon will be ignored
param_node.attrib["value"] = ' '.join(map(str, param.default))
elif param_type != "boolean":
param_node.attrib["value"] = str(param.default)
else:
# simple boolean with a default
if param.default is True:
param_node.attrib["checked"] = "true"
else:
if param.type is int or param.type is float:
# galaxy requires "value" to be included for int/float
# since no default was included, we need to figure out one in a clever way... but let the user know
# that we are "thinking" for him/her
warning("Generating default value for parameter [%s]. "
"Galaxy requires the attribute 'value' to be set for integer/floats. "
"Edit the CTD file and provide a suitable default value." % param.name, 1)
# check if there's a min/max and try to use them
default_value = None
if param.restrictions is not None:
if type(param.restrictions) is _NumericRange:
default_value = param.restrictions.n_min
if default_value is None:
default_value = param.restrictions.n_max
if default_value is None:
# no min/max provided... just use 0 and see what happens
default_value = 0
else:
# should never be here, since we have validated this anyway...
# this code is here just for documentation purposes
# however, better safe than sorry!
# (it could be that the code changes and then we have an ugly scenario)
raise InvalidModelException("Expected either a numeric range for parameter [%(name)s], "
"but instead got [%(type)s]"
% {"name": param.name, "type": type(param.restrictions)})
else:
# no restrictions and no default value provided...
# make up something
default_value = 0
param_node.attrib["value"] = str(default_value)
label = "%s parameter" % param.name
help_text = ""
if param.description is not None:
label, help_text = generate_label_and_help(param.description)
param_node.attrib["label"] = label
param_node.attrib["help"] = "(-%s)" % param.name + " " + help_text
def generate_label_and_help(desc):
label = ""
help_text = ""
# This tag is found in some descriptions
desc = str(desc).replace("#br#", " <br>")
# Get rid of dots in the end
if desc.endswith("."):
desc = desc.rstrip(".")
# Check if first word is a normal word and make it uppercase
if str(desc).find(" ") > -1:
first_word, rest = str(desc).split(" ", 1)
if str(first_word).islower():
# check if label has a quotient of the form a/b
if first_word.find("/") != 1 :
first_word.capitalize()
desc = first_word + " " + rest
label = desc
# Try to split the label if it is too long
if len(desc) > 50:
# find an example and put everything before in the label and the e.g. in the help
if desc.find("e.g.") > 1 :
label, help_text = desc.split("e.g.",1)
help_text = "e.g." + help_text
else:
# find the end of the first sentence
# look for ". " because some labels contain .file or something similar
delimiter = ""
if desc.find(". ") > 1 and desc.find("? ") > 1:
if desc.find(". ") < desc.find("? "):
delimiter = ". "
else:
delimiter = "? "
elif desc.find(". ") > 1:
delimiter = ". "
elif desc.find("? ") > 1:
delimiter = "? "
if delimiter != "":
label, help_text = desc.split(delimiter, 1)
# add the question mark back
if delimiter == "? ":
label += "? "
# remove all linebreaks
label = label.rstrip().rstrip('<br>').rstrip()
return label, help_text
def get_indented_text(text, indentation_level):
return ("%(indentation)s%(text)s" %
{"indentation": " " * (MESSAGE_INDENTATION_INCREMENT * indentation_level),
"text": text})
def warning(warning_text, indentation_level):
sys.stdout.write(get_indented_text("WARNING: %s\n" % warning_text, indentation_level))
def error(error_text, indentation_level):
sys.stderr.write(get_indented_text("ERROR: %s\n" % error_text, indentation_level))
def info(info_text, indentation_level):
sys.stdout.write(get_indented_text("INFO: %s\n" % info_text, indentation_level))
# determines if the given choices are boolean (basically, if the possible values are yes/no, true/false)
def is_boolean_parameter(param):
## detect boolean selects of OpenMS
if is_selection_parameter(param):
if len(param.restrictions.choices) == 2:
# check that default value is false to make sure it is an actual flag
if "false" in param.restrictions.choices and \
"true" in param.restrictions.choices and \
param.default == "false":
return True
else:
return param.type is bool
# determines if there are choices for the parameter
def is_selection_parameter(param):
return type(param.restrictions) is _Choices
def get_lowercase_list(some_list):
lowercase_list = map(str, some_list)
lowercase_list = map(string.lower, lowercase_list)
lowercase_list = map(strip, lowercase_list)
return lowercase_list
# creates a galaxy boolean parameter type
# this method assumes that param has restrictions, and that only two restictions are present
# (either yes/no or true/false)
def create_boolean_parameter(param_node, param):
# first, determine the 'truevalue' and the 'falsevalue'
"""TODO: true and false values can be way more than 'true' and 'false'
but for that we need CTD support
"""
# by default, 'true' and 'false' are handled as flags, like the verbose flag (i.e., -v)
true_value = "-%s" % get_param_name(param)
false_value = ""
choices = get_lowercase_list(param.restrictions.choices)
if "yes" in choices:
true_value = "yes"
false_value = "no"
param_node.attrib["truevalue"] = true_value
param_node.attrib["falsevalue"] = false_value
# set the checked attribute
if param.default is not None:
checked_value = "false"
default = strip(string.lower(param.default))
if default == "yes" or default == "true":
checked_value = "true"
#attribute_list["checked"] = checked_value
param_node.attrib["checked"] = checked_value
def create_outputs(parent, model, **kwargs):
outputs_node = add_child_node(parent, "outputs")
parameter_hardcoder = kwargs["parameter_hardcoder"]
for param in extract_parameters(model):
# no need to show hardcoded parameters
hardcoded_value = parameter_hardcoder.get_hardcoded_value(param.name, model.name)
if param.name in kwargs["blacklisted_parameters"] or hardcoded_value:
# let's not use an extra level of indentation and use NOP
continue
if param.type is _OutFile:
create_output_node(outputs_node, param, model, kwargs["supported_file_formats"])
# If there are no outputs defined in the ctd the node will have no children
# and the stdout will be used as output
if len(outputs_node) == 0:
add_child_node(outputs_node, "data",
OrderedDict([("name", "param_stdout"), ("format", "text"), ("label", "Output from stdout")]))
def create_output_node(parent, param, model, supported_file_formats):
data_node = add_child_node(parent, "data")
data_node.attrib["name"] = get_galaxy_parameter_name(param)
data_format = "data"
if param.restrictions is not None:
if type(param.restrictions) is _FileFormat:
# set the first data output node to the first file format
# check if there are formats that have not been registered yet...
output = list()
for format_name in param.restrictions.formats:
if not format_name in supported_file_formats.keys():
output.append(str(format_name))
# warn only if there's about to complain
if output:
warning("Parameter " + param.name + " has the following unsupported format(s):" + ','.join(output), 1)
data_format = ','.join(output)
formats = get_supported_file_types(param.restrictions.formats, supported_file_formats)
try:
data_format = formats.pop()
except KeyError:
# there is not much we can do, other than catching the exception
pass
# if there are more than one output file formats try to take the format from the input parameter
if formats:
corresponding_input = get_input_with_same_restrictions(param, model, supported_file_formats)
if corresponding_input is not None:
data_format = "input"
data_node.attrib["metadata_source"] = get_galaxy_parameter_name(corresponding_input)
else:
raise InvalidModelException("Unrecognized restriction type [%(type)s] "
"for output [%(name)s]" % {"type": type(param.restrictions),
"name": param.name})
data_node.attrib["format"] = data_format
#TODO: find a smarter label ?
#if param.description is not None:
# data_node.setAttribute("label", param.description)
return data_node
def get_supported_file_types(formats, supported_file_formats):
return set([supported_file_formats.get(format_name, DataType(format_name, format_name)).galaxy_extension
for format_name in formats if format_name in supported_file_formats.keys()])
def create_change_format_node(parent, data_formats, input_ref):
# <change_format>
# <when input="secondary_structure" value="true" format="text"/>
# </change_format>
change_format_node = add_child_node(parent, "change_format")
for data_format in data_formats:
add_child_node(change_format_node, "when",
OrderedDict([("input", input_ref), ("value", data_format), ("format", data_format)]))
# Shows basic information about the file, such as data ranges and file type.
def create_help(tool, model):
manual = ''
doc_url = None
if 'manual' in model.opt_attribs.keys():
manual += '%s\n\n' % model.opt_attribs["manual"]
if 'docurl' in model.opt_attribs.keys():
doc_url = model.opt_attribs["docurl"]
help_text = "No help available"
if manual is not None:
help_text = manual
if doc_url is not None:
help_text = ("" if manual is None else manual) + "\nFor more information, visit %s" % doc_url
help_node = add_child_node(tool, "help")
# TODO: do we need CDATA Section here?
help_node.text = help_text
# since a model might contain several ParameterGroup elements,
# we want to simply 'flatten' the parameters to generate the Galaxy wrapper
def extract_parameters(model):
parameters = []
if len(model.parameters.parameters) > 0:
# use this to put parameters that are to be processed
# we know that CTDModel has one parent ParameterGroup
pending = [model.parameters]
while len(pending) > 0:
# take one element from 'pending'
parameter = pending.pop()
if type(parameter) is not ParameterGroup:
parameters.append(parameter)
else:
# append the first-level children of this ParameterGroup
pending.extend(parameter.parameters.values())
# returned the reversed list of parameters (as it is now,
# we have the last parameter in the CTD as first in the list)
return reversed(parameters)
# adds and returns a child node using the given name to the given parent node
def add_child_node(parent_node, child_node_name, attributes=OrderedDict([])):
child_node = SubElement(parent_node, child_node_name, attributes)
return child_node
if __name__ == "__main__":
sys.exit(main())