add data table filter

lib/galaxy/tool_util/xsd/galaxy.xsd

@@ -4287,7 +4287,7 @@ only used if ``multiple`` is set to ``true``.]]></xs:documentation>
       <xs:annotation>
         <xs:documentation xml:lang="en">If ``true``, keep columns matching the
 value, if ``false`` discard columns matching the value. Used when ``type`` is
-either ``static_value``, ``regexp`` or ``param_value``.</xs:documentation>
+either ``static_value``, ``regexp``, ``param_value`` or ``data_table``.</xs:documentation>
       </xs:annotation>
     </xs:attribute>
     <xs:attribute name="value" type="xs:string">
@@ -4319,6 +4319,18 @@ added to the end of the list.</xs:documentation>
 from the list.</xs:documentation>
       </xs:annotation>
     </xs:attribute>
+    <xs:attribute name="table_name" type="xs:string">
+      <xs:annotation>
+        <xs:documentation xml:lang="en">Only used when ``type`` is
+``data_table``. The name of the data table to use.</xs:documentation>
+      </xs:annotation>
+    </xs:attribute>
+    <xs:attribute name="data_table_column" type="xs:string">
+      <xs:annotation>
+        <xs:documentation xml:lang="en">Only used when ``type`` is
+``data_table``. The column of the data table to use (0 based index or column name).</xs:documentation>
+      </xs:annotation>
+    </xs:attribute>
   </xs:complexType>
   <xs:complexType name="Outputs">
     <xs:annotation>
@@ -6192,6 +6204,7 @@ and ``bibtex`` are the only supported options.</xs:documentation>
       <xs:enumeration value="add_value"/>
       <xs:enumeration value="remove_value"/>
       <xs:enumeration value="sort_by"/>
+      <xs:enumeration value="data_table"/>
     </xs:restriction>
   </xs:simpleType>
   <xs:simpleType name="ActionsConditionalFilterType">

lib/galaxy/tools/parameters/dynamic_options.py

@@ -499,6 +499,58 @@ def filter_options(self, options, trans, other_values):
         return sorted(options, key=lambda x: x[self.column])
 
 
+class DataTableFilter(Filter):
+    """
+    Filters a list of options by entries present in a data table, i.e.
+    option[column] needs to be in the specified data table column
+
+    Type: data_table
+
+    Required Attributes:
+
+        - column: column in options to compare with
+        - table_name: data table to use
+        - data_table_column: data table column to use
+
+    Optional Attributes:
+
+        - keep: Keep options where option[column] is in the data table column (True)
+                Discard columns matching value (False)
+
+    """
+    def __init__(self, d_option, elem):
+        Filter.__init__(self, d_option, elem)
+        self.table_name = elem.get("table_name", None)
+        assert self.table_name is not None, "Required 'table_name' attribute missing from filter"
+        column = elem.get("column", None)
+        assert column is not None, "Required 'column' attribute missing from filter"
+        self.column = d_option.column_spec_to_index(column)
+        self.data_table_column = elem.get("data_table_column", None)
+        assert self.data_table_column is not None, "Required 'data_table_column' attribute missing from filter"
+        self.keep = string_as_bool(elem.get("keep", 'True'))
+
+    def filter_options(self, options, trans, other_values):
+        # get column from data table, by index or column name
+        entries = None
+        try:
+            entries = set([f[int(self.data_table_column)] for f in trans.app.tool_data_tables[self.table_name].get_fields()])
+        except TypeError:
+            pass
+        try:
+            entries = set([f[self.data_table_column] for f in trans.app.tool_data_tables[self.table_name].get_named_fields_list()])
+        except KeyError:
+            pass
+        if entries is None:
+            log.error(f"could not get data from column {self.data_table_column} from data_table {self.table_name}")
+            return options
+
+        rval = []
+        for o in options:
+            log.error(f"o {o}")
+            if self.keep == (o[self.column] in entries):
+                rval.append(o)
+        return rval
+
 filter_types = dict(data_meta=DataMetaFilter,
                     param_value=ParamValueFilter,
                     static_value=StaticValueFilter,
@@ -508,7 +560,8 @@ def filter_options(self, options, trans, other_values):
                     attribute_value_splitter=AttributeValueSplitterFilter,
                     add_value=AdditionalValueFilter,
                     remove_value=RemoveValueFilter,
-                    sort_by=SortByColumnFilter)
+                    sort_by=SortByColumnFilter,
+                    data_table=DataTableFilter)
 
 
 class DynamicOptions:
@@ -557,10 +610,12 @@ def load_from_parameter(from_parameter, transform_lines=None):
                 data_file = data_file.strip()
                 if not os.path.isabs(data_file):
                     full_path = os.path.join(self.tool_param.tool.app.config.tool_data_path, data_file)
+                    log.error(f"full_path {full_path}")
                     if os.path.exists(full_path):
                         self.index_file = data_file
                         with open(full_path) as fh:
                             self.file_fields = self.parse_file_fields(fh)
+                        log.error(f"exists file_fields {self.file_fields}")
                     else:
                         self.missing_index_file = data_file
             elif dataset_file is not None:

test/functional/tool-data/test_file.tsv

@@ -0,0 +1,2 @@
+hg19_value	hg19	hg19_name	hg19_path
+absent_value	absent	absent_name	absent_path

test/functional/tools/filter_data_table.xml

@@ -0,0 +1,45 @@
+<tool id="filter_data_table" name="filter data table" version="0.1.0">
+    <description>Filter on datatable entries</description>
+    <command><![CDATA[
+       echo $dynamic_select > '$output'
+    ]]></command>
+    <inputs>
+        <param name="dynamic_select" type="select">
+            <!-- initialize options from file (contains all entries that could be added to the data table) -->
+            <options from_file="../test/functional/tool-data/test_file.tsv">
+                <column name="name" index="0"/>
+                <column name="value" index="0"/>
+                <!-- filter options that are already in the data table -->
+                <filter type="data_table"  column="0" table_name="test_fasta_indexes" data_table_column="0" keep="false"/>
+                <validator type="no_options" message="No options available" />
+            </options>
+        </param>
+    </inputs>
+
+    <outputs>
+        <data name="output" format="txt" />
+    </outputs>
+
+    <tests>
+        <!-- select the value that is absent from the data table -->
+        <test expect_failure="false">
+            <param name="dynamic_select" value="absent_value" />
+            <output name="output">
+                <assert_contents>
+                    <has_text text="absent_value"/>
+                </assert_contents>
+            </output>
+        </test>
+        <!-- selecting the value that is already in the data table (and therefor filtered) fails -->
+        <test expect_failure="true">
+            <param name="dynamic_select" value="hg19_value" />
+            <output name="output">
+                <assert_contents>
+                    <has_text text="hg19_value"/>
+                </assert_contents>
+            </output>
+        </test>
+    </tests>
+    <help>
+    </help>
+</tool>

test/functional/tools/samples_tool_conf.xml

@@ -53,6 +53,7 @@
   <tool file="dbkey_filter_input.xml" />
   <tool file="dbkey_filter_multi_input.xml" />
   <tool file="dbkey_filter_collection.xml" />
+  <tool file="filter_data_table.xml" />
   <tool file="dbkey_filter_paired_collection.xml" />
   <tool file="dbkey_output_action.xml" />
   <tool file="composite_output.xml" />