URL links fix to open them in new tab (#1369)

* 34067 url links fix to open in new tab

* 34067 fixed test for pyright

* 34067 fixed test for pyright

Signed-off-by: Marta Tomzik <marta.tomzik@ibm.com>

* 34067 removed egg info

Signed-off-by: Marta Tomzik <marta.tomzik@ibm.com>

* 34067 isort fix

Signed-off-by: Marta Tomzik <marta.tomzik@ibm.com>

---------

Signed-off-by: Marta Tomzik <marta.tomzik@ibm.com>

lale/visualize.py

@@ -173,7 +173,7 @@ def _json_to_graphviz_rec(uid, jsn, cluster2reps, is_root, dot_graph_attr):
                 tooltip=_indiv_op_tooltip(uid, jsn),
             )
             if "documentation_url" in jsn:
-                dot.attr("graph", URL=jsn["documentation_url"])
+                dot.attr("graph", URL=jsn["documentation_url"], target="_blank")
             nodes = jsn["steps"]
             if jsn["class"] == _LALE_SKL_PIPELINE:
                 names = list(nodes.keys())
@@ -192,6 +192,7 @@ def _json_to_graphviz_rec(uid, jsn, cluster2reps, is_root, dot_graph_attr):
                 "style": "filled",
                 "fillcolor": _STATE2COLOR[step_jsn["state"]],
                 "tooltip": tooltip,
+                "target": "_blank",
             }
             if "documentation_url" in step_jsn:
                 attrs["URL"] = step_jsn["documentation_url"]

test/test_json_pretty_viz.py

@@ -36,6 +36,19 @@ def test_with_operator_choice(self):
         optimizable = kernel_tfm_or_not >> tfm >> clf
         optimizable.visualize(ipython_display=False)
 
+    def test_url_generation(self):
+        from graphviz import Source
+
+        from lale.lib.lale import NoOp
+        from lale.lib.sklearn import PCA, LogisticRegression
+
+        sample = NoOp >> PCA >> LogisticRegression
+        result = Source(str(sample.visualize(ipython_display=False)))
+        for i in result:
+            if "URL" in i:
+                print(i)
+                self.assertTrue("target=_blank" in i)
+
     def test_invalid_input(self):
         from sklearn.linear_model import LogisticRegression as SklearnLR