Add tool for exporting individual files to Galaxy.

Stolen wholesale from EU. Trying to make it a bit more general though.

lib/galaxy/config/sample/tool_conf.xml.sample

@@ -23,7 +23,8 @@
     <tool file="data_source/hbvar.xml" />
   </section>
   <section id="send" name="Send Data">
-    <tool file="cloud/send.xml" />
+    <tool file="data_export/send.xml" />
+    <tool file="data_export/export_remote.xml" />
   </section>
   <section id="collection_operations" name="Collection Operations">
     <tool file="${model_tools_path}/unzip_collection.xml" />

lib/galaxy/tools/__init__.py

@@ -126,6 +126,7 @@ GALAXY_LIB_TOOLS_UNVERSIONED = [
     "send_to_cloud",
     "__DATA_FETCH__",
     "directory_uri",
+    "export_remote",
     # Legacy tools bundled with Galaxy.
     "laj_1",
     "gff2bed1",

test/functional/tools/samples_tool_conf.xml

@@ -1,6 +1,7 @@
 <?xml version="1.0"?>
 <toolbox tool_path="${tool_conf_dir}" is_shed_conf="false">
   <tool file="upload.xml"/>
+  <tool file="export_remote.xml"/>
   <section id="test" name="Test Section">
     <tool file="multi_data_optional.xml" />
     <tool file="paths_as_file.xml" />

test/integration/test_remote_files.py

@@ -135,6 +135,37 @@ class RemoteFilesIntegrationTestCase(ConfiguresRemoteFilesIntegrationTestCase):
             with open(os.path.join(ftp_dir, 'helloworld')) as f:
                 assert 'hello world!\n' == f.read()
 
+    def test_export_remote_tool_default(self):
+        dataset_populator = self.dataset_populator
+        ftp_dir = self.user_ftp_dir
+        _write_file_fixtures(self.root, ftp_dir)
+        with dataset_populator.test_history() as history_id:
+            dataset = dataset_populator.new_dataset(history_id, content="example content", wait=True, name="foo")
+            infile = {
+                "src": "hda",
+                "id": dataset["id"]
+            }
+            inputs = {
+                "d_uri": "gxftp://",
+                "export_type|export_type_selector": "datasets_auto",
+                "export_type|infiles": [infile],
+            }
+            response = dataset_populator.run_tool("export_remote", inputs, history_id)
+            response.raise_for_status()
+            with open(os.path.join(ftp_dir, 'foo.txt')) as f:
+                assert 'example content\n' == f.read()
+
+            inputs = {
+                "d_uri": "gxftp://",
+                "export_type|export_type_selector": "datasets_named",
+                "export_type|datasets_0|infile": infile,
+                "export_type|datasets_0|name": "my_cool_name.txt",
+            }
+            response = dataset_populator.run_tool("export_remote", inputs, history_id)
+            response.raise_for_status()
+            with open(os.path.join(ftp_dir, 'my_cool_name.txt')) as f:
+                assert 'example content\n' == f.read()
+
     def _assert_index_empty(self, index):
         assert len(index) == 0
 

tools/data_export/export_remote.py

@@ -0,0 +1,53 @@
+import argparse
+import json
+import os
+import sys
+
+from galaxy.files import ConfiguredFileSources
+
+
+def get_file_sources(file_sources_path):
+    assert os.path.exists(file_sources_path), "file sources path [%s] does not exist" % file_sources_path
+    with open(file_sources_path, "r") as f:
+        file_sources_as_dict = json.load(f)
+    file_sources = ConfiguredFileSources.from_dict(file_sources_as_dict)
+    return file_sources
+
+
+def main(argv=None):
+    if argv is None:
+        argv = sys.argv[1:]
+    args = _parser().parse_args(argv)
+    exit_code = 0
+    file_sources = get_file_sources(args.file_sources)
+    directory_uri = args.directory_uri
+    counter = 0
+    for real_data_path, name in zip(args.infiles, args.names):
+        if directory_uri.endswith("/"):
+            target_uri = directory_uri + name
+        else:
+            target_uri = directory_uri + "/" + name
+        file_source_path = file_sources.get_file_source_path(target_uri)
+        if os.path.exists(file_source_path.path):
+            print(f'Error: File "{file_source_path.path}" already exists. Skipping.')
+            exit_code = 1
+            continue
+        file_source = file_source_path.file_source
+        file_source.write_from(file_source_path.path, real_data_path)
+        counter += 1
+    print(f"'{counter}' out of '{len(args.infiles)}' files have been exported.\n")
+    sys.exit(exit_code)
+
+
+def _parser():
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--directory-uri', type=str,
+                        help='directory target URI')
+    parser.add_argument('--file-sources', type=str, help='file sources json')
+    parser.add_argument('--infiles', type=str, nargs='*', required=True)
+    parser.add_argument('--names', type=str, nargs='*', required=True)
+    return parser
+
+
+if __name__ == "__main__":
+    main()

tools/data_export/export_remote.xml

@@ -0,0 +1,60 @@
+<tool id="export_remote" name="Export datasets" version="0.1.0">
+    <description>to remote files source</description>
+    <command><![CDATA[
+        python '$__tool_directory__/export_remote.py'
+            --file-sources '$file_sources'
+            --directory-uri '$d_uri'
+            #if $export_type.export_type_selector == "datasets_auto"
+                --infiles
+                #for $filepath in $export_type.infiles:
+                    '$filepath'
+                #end for
+                --names
+                #for $name in $export_type.infiles:
+                    #if $name.ext == 'vcf_bgzip':
+                        #set file_ext = 'vcf.bz'
+                    #else:
+                        #set file_ext = $name.ext
+                    #end if
+                    '${name.element_identifier}.${file_ext}'
+                #end for
+            #else
+                --infiles
+                #for $dataset in $export_type.datasets:
+                    '$dataset.infile'
+                #end for
+                --names
+                #for $dataset in $export_type.datasets:
+                    '$dataset.name'
+                #end for
+            #end if
+            > '$out'
+    ]]></command>
+    <inputs>
+        <conditional name="export_type">
+            <param name="export_type_selector" type="select" label="What would you like to export?">
+                <option value="datasets_auto" selected="True">Datasets using their Galaxy name and extension.</option>
+                <option value="datasets_named">Datasets using specified name.</option>
+                <!-- Allow more options - e.g. HTML, collections preserving their structure... -->
+            </param>
+            <when value="datasets_auto">
+                <param name="infiles" type="data" format="txt,binary" multiple="true" label="Choose your datasets"/>
+            </when>
+            <when value="datasets_named">
+                <repeat name="datasets" title="Dataset">
+                    <param type="data" name="infile" label="Input dataset" />
+                    <param type="text" name="name" label="Name" />
+                </repeat>
+            </when>
+        </conditional>
+        <param type="directory_uri" name="d_uri" label="Directory URI" />
+    </inputs>
+    <outputs>
+        <data name="out" format="txt" label="Export logs"/>
+    </outputs>
+    <configfiles>
+        <file_sources name="file_sources" />
+    </configfiles>
+    <tests>
+    </tests>
+</tool>