Add tool that tags collection elements from a file

This tags collection elements (but not collections).
galaxyproject · Feb 6, 2018 · b114f3f · b114f3f
1 parent 3864a17
commit b114f3f
Show file tree

Hide file tree

Showing 4 changed files with 142 additions and 0 deletions.
diff --git a/config/tool_conf.xml.sample b/config/tool_conf.xml.sample
@@ -36,6 +36,7 @@
     <tool file="${model_tools_path}/filter_failed_collection.xml" />
     <tool file="${model_tools_path}/flatten_collection.xml" />
     <tool file="${model_tools_path}/merge_collection.xml" />
+    <tool file="${model_tools_path}/tag_collection_from_file.xml" />
     <tool file="${model_tools_path}/relabel_from_file.xml" />
     <tool file="${model_tools_path}/filter_from_file.xml" />
     <tool file="${model_tools_path}/sort_collection_list.xml" />

diff --git a/lib/galaxy/tools/__init__.py b/lib/galaxy/tools/__init__.py
@@ -28,6 +28,7 @@
 from galaxy.datatypes.metadata import JobExternalOutputMetadataWrapper
 from galaxy.managers import histories
 from galaxy.managers.jobs import JobSearch
+from galaxy.managers.tags import GalaxyTagManager
 from galaxy.queue_worker import send_control_task
 from galaxy.tools.actions import DefaultToolAction
 from galaxy.tools.actions.data_manager import DataManagerToolAction
@@ -2605,6 +2606,59 @@ def add_copied_value_to_new_elements(new_label, dce_object):
         )
 
 
+class TagFromFileTool(DatabaseOperationTool):
+    tool_type = 'tag_from_file'
+
+    def produce_outputs(self, trans, out_data, output_collections, incoming, history, **kwds):
+        hdca = incoming["input"]
+        set_tags = incoming['set_tags']
+        new_tags_dataset_assoc = incoming["tags"]
+        new_elements = odict()
+        tags_manager = GalaxyTagManager(trans.app.model.context)
+
+        def add_copied_value_to_new_elements(new_tags_dict, dce):
+            if getattr(dce.element_object, "history_content_type", None) == "dataset":
+                copied_value = dce.element_object.copy()
+                # copy should never be visible, since part of a collection
+                copied_value.visble = False
+                history.add_dataset(copied_value, copied_value, set_hid=False)
+                new_tags = new_tags_dict.get(dce.element_identifier)
+                if new_tags:
+                    if not set_tags and dce.element_object.tags:
+                        # We need get the original tags and update them with the new tags
+                        old_tags = set(tag for tag in tags_manager.get_tags_str(dce.element_object.tags).split(',') if tag)
+                        old_tags.update(set(new_tags))
+                        new_tags = old_tags
+                    tags_manager.add_tags_from_list(user=history.user, item=copied_value, new_tags_list=new_tags)
+            else:
+                # We have a collection, and we copy the elements so that we don't manipulate the original tags
+                copied_value = dce.element_object.copy(element_destination=history)
+                for new_element, old_element in zip(copied_value.dataset_elements, dce.element_object.dataset_elements):
+                    # TODO: This should be eliminated, but collections created by the collection builder
+                    # don't set `visible` to `False` if you don't hide the original elements.
+                    new_element.element_object.visible = False
+                    new_tags = new_tags_dict.get(new_element.element_identifier)
+                    if not set_tags:
+                        old_tags = set(tag for tag in tags_manager.get_tags_str(old_element.element_object.tags).split(',') if tag)
+                        if new_tags:
+                            old_tags.update(set(new_tags))
+                        new_tags = old_tags
+                    tags_manager.add_tags_from_list(user=history.user, item=new_element.element_object, new_tags_list=new_tags)
+            new_elements[dce.element_identifier] = copied_value
+
+        new_tags_path = new_tags_dataset_assoc.file_name
+        new_tags = open(new_tags_path, "r").readlines(1024 * 1000000)
+        # We have a tabular file, where the first column is an existing element identifier,
+        # and the remaining columns represent new tags.
+        source_new_tags = (line.strip().split('\t') for line in new_tags)
+        new_tags_dict = {item[0]: item[1:] for item in source_new_tags}
+        for i, dce in enumerate(hdca.collection.elements):
+            add_copied_value_to_new_elements(new_tags_dict, dce)
+        output_collections.create_collection(
+            next(iter(self.outputs.values())), "output", elements=new_elements
+        )
+
+
 class FilterFromFileTool(DatabaseOperationTool):
     tool_type = 'filter_from_file'
 

diff --git a/lib/galaxy/tools/tag_collection_from_file.xml b/lib/galaxy/tools/tag_collection_from_file.xml
@@ -0,0 +1,85 @@
+<tool id="__TAG_FROM_FILE__"
+      name="Tag elements from file"
+      version="1.0.0"
+      tool_type="tag_from_file">
+    <description>from contents of a file</description>
+    <type class="TagFromFileTool" module="galaxy.tools" />
+    <action module="galaxy.tools.actions.model_operations"
+            class="ModelOperationToolAction"/>
+    <inputs>
+        <param type="data_collection" name="input" label="Input Collection" help="A tabular file indicating how to tag collection elements."/>
+           <param type="data" name="tags" format="tabular" label="Add tags from this file" />
+        <param name="set_tags" type="boolean" label="Set tags instead of updating tags?" help="If you select yes existing tags will be removed, otherwise new tags and existing tags will be merged."/>
+    </inputs>
+    <outputs>
+        <collection name="output" format_source="input" type_source="input" label="${on_string} (Tagged)" >
+        </collection>
+    </outputs>
+    <tests>
+        <test>
+            <param name="input">
+                <collection type="list">
+                    <element name="forward" value="simple_line.txt" />
+                    <element name="reverse" value="simple_line_alternative.txt" />
+                </collection>
+            </param>
+            <param name="tags" value="new_tags_1.txt" ftype="txt" />
+            <output_collection name="output" type="list">
+                <element name="forward">
+                  <assert_contents>
+                    <has_text_matching expression="^This is a line of text.\n$" />
+                  </assert_contents>
+                  <metadata name="tags" value="alias:r1,orientation:forward" />
+                </element>
+                <element name="reverse">
+                  <assert_contents>
+                    <has_text_matching expression="^This is a different line of text.\n$" />
+                  </assert_contents>
+                  <metadata name="tags" value="alias:r2,orientation:reverse" />
+                </element>
+            </output_collection>
+        </test>
+        <test>
+            <param name="input">
+                <collection type="list:paired">
+                    <element name="i1">
+                        <collection type="paired">
+                            <element name="forward" value="simple_line.txt" />
+                            <element name="reverse" value="simple_line_alternative.txt" />
+                        </collection>
+                    </element>
+                </collection>
+            </param>
+            <param name="tags" value="new_tags_1.txt" ftype="txt" />
+            <output_collection name="output" type="list:paired">
+              <element name="i1">
+                <element name="forward">
+                  <assert_contents>
+                    <has_text_matching expression="^This is a line of text.\n$" />
+                  </assert_contents>
+                  <metadata name="tags" value="alias:r1,orientation:forward" />
+                </element>
+                <element name="reverse">
+                  <assert_contents>
+                    <has_text_matching expression="^This is a different line of text.\n$" />
+                  </assert_contents>
+                  <metadata name="tags" value="alias:r2,orientation:reverse" />
+                </element>
+              </element>
+            </output_collection>
+        </test>
+    </tests>
+    <help><![CDATA[
+
+.. class:: infomark
+
+        This tool will take an input collection and a tabular file,
+        where the first column indicates an element identifier and the
+        remaining columns contains the new tags. This file may contain
+        less entries than elements in the collection.
+        In that case only matching list identifiers will be tagged.
+
+        This tool will create new history datasets from your collection
+        but your quota usage will not increase.
+    ]]></help>
+</tool>
diff --git a/test-data/new_tags_1.txt b/test-data/new_tags_1.txt
@@ -0,0 +1,2 @@
+forward	orientation:forward	alias:r1
+reverse	orientation:reverse	alias:r2