Deduplicate dependencies for HTMLTextDocument

CHANGELOG.md

@@ -9,9 +9,12 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ## [UNRELEASED]
 
 * Changed the type annotation of `_add_ws` from `bool` to `TagAttrValue`. This makes it easier to write functions which call `Tag` functions and pass along `**kwargs`. (#67)
+
 * Changed the type annotation of `collapse_` from `str` to `str | float | None`. This makes it easier to write calls to `css()` pass along `**kwargs`. (#68)
+
 * Enhanced the type definition of `TagAttrs` to include `TagAttrDict`, the type of a `Tag`'s `attrs` property. (#55)
 
+* For `HTMLTextDocument` objects, deduplicate HTML dependencies. (#72)
 
 ## [0.4.1] 2023-10-30
 

htmltools/_core.py

@@ -1157,7 +1157,13 @@ def _static_extract_serialized_html_deps(
         # HTMLdependency.get_tag_representation()
         pattern = r'<script type="application/json" data-html-dependency="">((?:.|\r|\n)*?)</script>'
         dep_strs = re.findall(pattern, html)
-        # html = re.sub(pattern, "", html)
+        # Deduplicate dependencies. htmltools normally would dedupe dependencies, but
+        # with HTMLTextDocuments, the input HTML would usually have been generated by
+        # something else (like Quarto) and may not have the dependencies deduped.
+        dep_strs = list(set(dep_strs))
+
+        # Remove the serialized HTML dependencies from the HTML string
+        html = re.sub(pattern, "", html)
 
         deps: list[HTMLDependency] = []
         for dep_str in dep_strs:

tests/test_html_document.py

@@ -291,14 +291,50 @@ def test_json_roundtrip():
                 div("hello world", testdep),
                 # Also make sure it would work even with indents
                 ht.HTML(testdep2.serialize_to_script_json(indent=2)),
+                # Add another copy of testdep, explicitly serialized to script json.
+                # Normally htmltools will dedupe dependencies when .render() is called,
+                # but we do this here because when these deps are embedded in a Quarto
+                # document, Quarto can add each dep independent of the others and
+                # therefore have duplicates. Since we're using .render, to get
+                # duplicates, we need to force the duplication.
+                #
+                div(
+                    "hello again",
+                    ht.HTML(testdep.serialize_to_script_json()),
+                ),
             ]
         )
+
+        # Get a string representation which hasn't been passed through
+        # HTMLTextDocument().
         x_str = str(x)
+
+        # Make sure that we successfully forced testdep to show up twice in the HTML,
+        # before we pass it to HTMLTextDocument() and call .render().
+        assert x_str.count('"name": "testdep"') == 2
+
+        # Make sure that there are three of these HTML dependency script tags.
+        assert (
+            x_str.count('<script type="application/json" data-html-dependency="">') == 3
+        )
+
         rendered = ht.HTMLTextDocument(
             x_str, deps_replace_pattern='<meta data-foo="">'
         ).render()
+
+        # Make sure both deps are present.
         assert "testdep" in [d.name for d in rendered["dependencies"]]
         assert "testdep2" in [d.name for d in rendered["dependencies"]]
 
+        # Make sure testdep was deduplicated by HTMLTextDocument().render().
+        assert rendered["dependencies"].count(testdep) == 1
+        assert len(rendered["dependencies"]) == 2
+
+        # Make sure the HTML dependency script tags were stripped out.
+        assert (
+            '<script type="application/json" data-html-dependency="">'
+            not in rendered["html"]
+        )
+
     finally:
         ht.html_dependency_render_mode = old_mode