Test reporting: auto-discover widgets to test

Orange/canvas/report/tests/test_report.py

@@ -1,3 +1,6 @@
+from importlib import import_module
+import os
+
 from PyQt4.QtGui import QFont, QBrush
 from PyQt4.QtCore import Qt
 from Orange.data.table import Table
@@ -8,43 +11,15 @@
 from Orange.distance import Euclidean
 from Orange.canvas.report.owreport import OWReport
 from Orange.widgets import gui
+from Orange.widgets.widget import OWWidget
 from Orange.widgets.tests.base import WidgetTest
-from Orange.widgets.classify.owclassificationtree import OWClassificationTree
 from Orange.widgets.classify.owclassificationtreegraph import OWClassificationTreeGraph
-from Orange.widgets.classify.owknn import OWKNNLearner
-from Orange.widgets.classify.owlogisticregression import OWLogisticRegression
-from Orange.widgets.classify.owmajority import OWMajority
-from Orange.widgets.classify.ownaivebayes import OWNaiveBayes
-from Orange.widgets.classify.owrandomforest import OWRandomForest
-from Orange.widgets.classify.owsvmclassification import OWSVMClassification
-from Orange.widgets.data.owconcatenate import OWConcatenate
-from Orange.widgets.data.owcontinuize import OWContinuize
-from Orange.widgets.data.owdatainfo import OWDataInfo
-from Orange.widgets.data.owdatasampler import OWDataSampler
-from Orange.widgets.data.owdiscretize import OWDiscretize
-from Orange.widgets.data.owfeatureconstructor import OWFeatureConstructor
 from Orange.widgets.data.owfile import OWFile
-from Orange.widgets.data.owimpute import OWImpute
-from Orange.widgets.data.owmergedata import OWMergeData
-from Orange.widgets.data.owoutliers import OWOutliers
-from Orange.widgets.data.owpaintdata import OWPaintData
-from Orange.widgets.data.owpurgedomain import OWPurgeDomain
-from Orange.widgets.data.owrank import OWRank
-from Orange.widgets.data.owselectcolumns import OWSelectAttributes
-from Orange.widgets.data.owselectrows import OWSelectRows
-from Orange.widgets.data.owsql import OWSql
-from Orange.widgets.data.owtable import OWDataTable
-from Orange.widgets.data.owcolor import OWColor
-from Orange.widgets.data.owpreprocess import OWPreprocess
 from Orange.widgets.evaluate.owcalibrationplot import OWCalibrationPlot
 from Orange.widgets.evaluate.owliftcurve import OWLiftCurve
 from Orange.widgets.evaluate.owrocanalysis import OWROCAnalysis
 from Orange.widgets.evaluate.owtestlearners import OWTestLearners
-from Orange.widgets.regression.owregressiontree import OWRegressionTree
 from Orange.widgets.regression.owregressiontreegraph import OWRegressionTreeGraph
-from Orange.widgets.regression.owknnregression import OWKNNRegression
-from Orange.widgets.regression.owmean import OWMean
-from Orange.widgets.regression.owsvmregression import OWSVMRegression
 from Orange.widgets.unsupervised.owcorrespondence import OWCorrespondenceAnalysis
 from Orange.widgets.unsupervised.owdistancemap import OWDistanceMap
 from Orange.widgets.unsupervised.owdistances import OWDistances
@@ -53,15 +28,31 @@
 from Orange.widgets.unsupervised.owmds import OWMDS
 from Orange.widgets.unsupervised.owpca import OWPCA
 from Orange.widgets.utils.itemmodels import PyTableModel
-from Orange.widgets.visualize.owboxplot import OWBoxPlot
-from Orange.widgets.visualize.owdistributions import OWDistributions
-from Orange.widgets.visualize.owheatmap import OWHeatMap
-from Orange.widgets.visualize.owlinearprojection import OWLinearProjection
-from Orange.widgets.visualize.owmosaic import OWMosaicDisplay
-from Orange.widgets.visualize.owscattermap import OWScatterMap
-from Orange.widgets.visualize.owscatterplot import OWScatterPlot
-from Orange.widgets.visualize.owsieve import OWSieveDiagram
-from Orange.widgets.visualize.owvenndiagram import OWVennDiagram
+
+
+def get_owwidgets(top_module_name):
+    top_module = import_module(top_module_name)
+    widgets = []
+    for root, dirs, files in os.walk(top_module.__path__[0]):
+        root = root[len(top_module.__path__[0]):].lstrip(os.path.sep)
+        for file in files:
+            if file.lower().startswith('ow') and file.lower().endswith('.py'):
+                module_name = top_module_name + '.' + os.path.join(root, file).replace(os.path.sep, '.')[:-len('.py')]
+                module = import_module(module_name, top_module_name[:top_module_name.index('.')])
+                for name, value in module.__dict__.items():
+                    if (name.upper().startswith('OW') and
+                            isinstance(value, type) and
+                            issubclass(value, OWWidget) and
+                            getattr(value, 'name', None) and
+                            getattr(value, 'send_report', None)):
+                        widgets.append(value)
+    return list(set(widgets))
+
+
+DATA_WIDGETS = get_owwidgets('Orange.widgets.data')
+VISUALIZATION_WIDGETS = get_owwidgets('Orange.widgets.visualize')
+CLASSIFICATION_WIDGETS = get_owwidgets('Orange.widgets.classify')
+REGRESSION_WIDGETS = get_owwidgets('Orange.widgets.regression')
 
 
 class TestReport(WidgetTest):
@@ -119,32 +110,24 @@ def test_report_table(self):
 
 
 class TestReportWidgets(WidgetTest):
-    clas_widgets = [OWClassificationTree, OWKNNLearner, OWLogisticRegression,
-                    OWMajority, OWNaiveBayes, OWRandomForest,
-                    OWSVMClassification]
-    data_widgets = [OWConcatenate, OWContinuize, OWDataInfo, OWDataSampler,
-                    OWDiscretize, OWFeatureConstructor, OWOutliers, OWImpute,
-                    OWMergeData, OWFile, OWPaintData, OWPurgeDomain, OWRank,
-                    OWSelectAttributes, OWSelectRows, OWSql, OWDataTable,
-                    OWColor, OWPreprocess]
+    clas_widgets = CLASSIFICATION_WIDGETS
+    data_widgets = DATA_WIDGETS
     eval_widgets = [OWCalibrationPlot, OWLiftCurve, OWROCAnalysis]
-    regr_widgets = [OWRegressionTree, OWKNNRegression, OWMean, OWSVMRegression]
+    regr_widgets = REGRESSION_WIDGETS
     unsu_widgets = [OWCorrespondenceAnalysis, OWDistances, OWKMeans,
                     OWMDS, OWPCA]
     dist_widgets = [OWDistanceMap, OWHierarchicalClustering]
-    visu_widgets = [OWBoxPlot, OWDistributions, OWHeatMap, OWLinearProjection,
-                    OWMosaicDisplay, OWScatterPlot,
-                    OWSieveDiagram, OWScatterMap, OWVennDiagram]
+    visu_widgets = VISUALIZATION_WIDGETS
     spec_widgets = [OWClassificationTreeGraph, OWTestLearners,
                     OWRegressionTreeGraph]
 
     def _create_report(self, widgets, rep, data):
         for widget in widgets:
             w = self.create_widget(widget)
-            if w.inputs and data is not None:
+            if w.inputs and isinstance(data, w.inputs[0].type):
                 handler = getattr(w, w.inputs[0].handler)
                 handler(data)
-            w.create_report_html()
+                w.create_report_html()
             rep.make_report(w)
             # rep.show()
 
@@ -160,14 +143,12 @@ def test_report_widgets_classify(self):
         w.create_report_html()
         rep.make_report(w)
 
-        self.assertEqual(len(widgets) + 1, 8)
         self._create_report(widgets, rep, data)
 
     def test_report_widgets_data(self):
         rep = OWReport.get_instance()
         data = Table("zoo")
         widgets = self.data_widgets
-        self.assertEqual(len(widgets), 19)
         self._create_report(widgets, rep, data)
 
     def test_report_widgets_evaluate(self):
@@ -189,7 +170,6 @@ def test_report_widgets_evaluate(self):
         w.create_report_html()
         rep.make_report(w)
 
-        self.assertEqual(len(widgets) + 1, 4)
         self._create_report(widgets, rep, results)
 
     def test_report_widgets_regression(self):
@@ -204,35 +184,30 @@ def test_report_widgets_regression(self):
         w.create_report_html()
         rep.make_report(w)
 
-        self.assertEqual(len(widgets) + 1, 5)
         self._create_report(widgets, rep, data)
 
     def test_report_widgets_unsupervised(self):
         rep = OWReport.get_instance()
         data = Table("zoo")
         widgets = self.unsu_widgets
-        self.assertEqual(len(widgets), 5)
         self._create_report(widgets, rep, data)
 
     def test_report_widgets_unsupervised_dist(self):
         rep = OWReport.get_instance()
         data = Table("zoo")
         dist = Euclidean(data)
         widgets = self.dist_widgets
-        self.assertEqual(len(widgets), 2)
         self._create_report(widgets, rep, dist)
 
     def test_report_widgets_visualize(self):
         rep = OWReport.get_instance()
         data = Table("zoo")
         widgets = self.visu_widgets
-        self.assertEqual(len(widgets), 9)
         self._create_report(widgets, rep, data)
 
     def test_report_widgets_all(self):
         rep = OWReport.get_instance()
         widgets = self.clas_widgets + self.data_widgets + self.eval_widgets + \
                   self.regr_widgets + self.unsu_widgets + self.dist_widgets + \
                   self.visu_widgets + self.spec_widgets
-        self.assertEqual(len(widgets), 52)
         self._create_report(widgets, rep, None)

Orange/widgets/data/oweditdomain.py

@@ -543,9 +543,12 @@ def sizeHint(self):
         return sh.expandedTo(QSize(660, 550))
 
     def send_report(self):
-        self.report_raw("", EditDomainReport(
-            old_domain=chain(self.data.domain.variables, self.data.domain.metas),
-            new_domain=self.domain_model).to_html())
+        if self.data is not None:
+            self.report_raw("", EditDomainReport(
+                old_domain=chain(self.data.domain.variables, self.data.domain.metas),
+                new_domain=self.domain_model).to_html())
+        else:
+            self.report_data(None)
 
 
 class EditDomainReport:

Orange/widgets/visualize/owsilhouetteplot.py

@@ -335,6 +335,9 @@ def commit(self):
         self.send("Other Data", other)
 
     def send_report(self):
+        if not len(self.cluster_var_model):
+            return
+
         self.report_plot()
         caption = "Silhouette plot ({} distance), clustered by '{}'".format(
             self.Distances[self.distance_idx][0],