Merge pull request #25 from MichaelHatherly/mh/lexbor

Switch HTML parser
MichaelHatherly · Jul 3, 2024 · 2729753 · 2729753
2 parents 43f9142 + ea142ec
commit 2729753
Show file tree

Hide file tree

Showing 37 changed files with 6,350 additions and 204 deletions.
diff --git a/.JuliaFormatter.toml b/.JuliaFormatter.toml
@@ -1 +1,2 @@
 margin = 92
+remove_extra_newlines = true
diff --git a/Project.toml b/Project.toml
@@ -5,15 +5,14 @@ version = "1.3.1"
 
 [deps]
 AbstractTrees = "1520ce14-60c1-5f80-bbc7-55ef81b5835c"
-EzXML = "8f5d6c58-4d21-5cfd-889c-e3ad7ee6a615"
+CEnum = "fa961155-64e5-5f13-b03f-caf6b980ea82"
 InteractiveUtils = "b77e0a4c-d291-57a0-90e8-8db25a27a240"
-Logging = "56ddb016-857b-54e1-b83d-db4d58db5568"
-LoggingExtras = "e6f89c97-d47a-5376-807f-9c37f3926c36"
 MacroTools = "1914dd2f-81c6-5fcd-8719-6d5c9610ff09"
 Match = "7eb4fadd-790c-5f42-8a69-bfa0b872bfbf"
 PackageExtensionCompat = "65ce6f38-6b18-4e1d-a461-8949797d7930"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 TOML = "fa267f1f-6049-4f14-aa54-33bafae1ed76"
+lexbor_jll = "be8d7a73-1782-5ce7-96ff-8fff21e5e970"
 
 [weakdeps]
 CommonMark = "a80b9123-70ca-4bc0-993e-6e3bcb318db6"
@@ -27,13 +26,11 @@ HypertextTemplatesReviseExt = "Revise"
 
 [compat]
 AbstractTrees = "0.4"
-EzXML = "1"
+CEnum = "0.5"
 InteractiveUtils = "1.6"
-Logging = "1.6"
-LoggingExtras = "1"
 MacroTools = "0.5"
 Match = "2"
 PackageExtensionCompat = "1"
-Random = "1.6"
 TOML = "1"
+lexbor_jll = "= 2.3.0"
 julia = "1.6"
diff --git a/docs/src/index.md b/docs/src/index.md
@@ -57,7 +57,7 @@ Let's add a new template function to that `template.html` file:
 <function second-template count="1::Integer" value>
   <ul>
     <for iter="1:count" item>
-      <li><julia value="item" />: <julia value /></li>
+      <li><span .julia=item></span>: <span .julia=value></span></li>
     </for>
   </ul>
 </function>
@@ -109,9 +109,9 @@ You can use one of your template functions in another template with:
 
 ```html
 <function third-template number>
-  <example />
-  <second-template count="number" value='"Hello, world!"' />
-  <second-template count="2 * number" value='"Goodbye, world!"' />
+  <example></example>
+  <second-template count="number" value='"Hello, world!"'></second-template>
+  <second-template count="2 * number" value='"Goodbye, world!"'></second-template>
 </function>
 ```
 
@@ -141,10 +141,10 @@ be called `<base-template>` in other templates that it is used in, and
 ```html
 <html lang="en">
   <head>
-    <title>My Website - <julia value="props.title" /></title>
+    <title $julia="My Website - $(props.title)"></title>
   </head>
   <body>
-    <third-example number="props.number" />
+    <third-example number="props.number"></third-example>
   </body>
 </html>
 ```
@@ -195,7 +195,7 @@ since it uses that package internally.
     <case when="[element]">
       <p>
         The number is actually a single element vector containing:
-        <julia value="element" />
+        <span .julia=element></span>
       </p>
     </case>
     <case when="_">
@@ -219,7 +219,7 @@ For that we use `<slot>` tags.
 <function slot-example>
   <div class="...">
     <h1>Header</h1>
-    <slot />
+    <slot></slot>
     <hr />
     <p>Footer</p>
   </div>
@@ -241,9 +241,9 @@ rendered in a single template function. For that we have named slots:
 <function named-slots-example>
   <div class="...">
     <h1>Header</h1>
-    <slot content />
+    <slot content></slot>
     <hr />
-    <p><slot footer /></p>
+    <p><slot footer></slot></p>
   </div>
 </function>
 ```
@@ -342,12 +342,12 @@ name = "my-template"
 props = ["foo", "bar"]
 +++
 
-<large-custom-template prop="foo" />
+<large-custom-template prop="foo"></large-custom-template>
 
 # My Template
 
 This is a template that takes two props: `foo` and `bar`. Their values are
-<julia value="foo" /> and <julia value="bar" /> respectively.
+<span julia="foo"></span> and <span julia="bar"></span> respectively.
 ```
 
 

diff --git a/gen/Project.toml b/gen/Project.toml
@@ -0,0 +1,5 @@
+[deps]
+Clang = "40e3b903-d033-50b4-a0cc-940c62c95e31"
+JuliaFormatter = "98e50ef6-434e-11e9-1051-2b60c6c9e899"
+Pkg = "44cfe95a-1eb2-52ea-b672-e2afdf69b78f"
+lexbor_jll = "be8d7a73-1782-5ce7-96ff-8fff21e5e970"
diff --git a/gen/generator.jl b/gen/generator.jl
@@ -0,0 +1,82 @@
+import Pkg
+import Pkg.Artifacts
+import Clang.Generators
+import Clang.Generators.JLLEnvs
+import lexbor_jll
+import JuliaFormatter
+
+cd(@__DIR__) do
+    include_dir = normpath(joinpath(lexbor_jll.artifact_dir, "include"))
+    @assert isdir(include_dir)
+
+    options = Generators.load_options(joinpath(@__DIR__, "generator.toml"))
+
+    args = Generators.get_default_args()
+    push!(args, "-I$include_dir")
+
+    header_files = [joinpath(include_dir, "lexbor", "html", "html.h")]
+
+    ctx = Generators.create_context(header_files, args, options)
+
+    Generators.build!(ctx, Generators.BUILDSTAGE_NO_PRINTING)
+
+    allowed_functions = Set([
+        :lxb_html_document_destroy,
+        :lxb_html_tokenizer_tags_noi,
+        :lxb_tag_name_by_id_noi,
+        :lxb_html_tokenizer_destroy,
+        :lxb_html_tokenizer_create,
+        :lxb_html_tokenizer_init,
+        :lxb_html_tokenizer_callback_token_done_set_noi,
+        :lxb_html_tokenizer_begin,
+        :lxb_html_tokenizer_chunk,
+        :lxb_html_tokenizer_end,
+        :lxb_html_parser_create,
+        :lxb_html_parser_init,
+        :lxb_html_parser_destroy,
+        :lxb_html_parse,
+        :lxb_html_document_destroy,
+        :lxb_dom_node_text_content,
+        :lxb_dom_node_first_child_noi,
+        :lxb_dom_node_next_noi,
+        :lxb_dom_node_prev_noi,
+        :lxb_dom_element_qualified_name,
+        :lxb_dom_element_first_attribute_noi,
+        :lxb_dom_attr_value_noi,
+        :lxb_dom_attr_qualified_name,
+        :lxb_dom_element_next_attribute_noi,
+    ])
+
+    function rewrite!(e::Expr)
+        if Meta.isexpr(e, :function)
+            name = e.args[1].args[1]
+            if name === :lxb_css_syntax_token
+                e.args[1].args[2] = :($(e.args[1].args[2])::Ptr{lxb_css_syntax_tokenizer_t})
+            end
+            if isa(name, Symbol) && name ∉ allowed_functions
+                e.head = :(=)
+                e.args = [:_, "skip-this-code"]
+            end
+        end
+        return e
+    end
+
+    function rewrite!(dag::Generators.ExprDAG)
+        for node in Generators.get_nodes(dag)
+            for expr in Generators.get_exprs(node)
+                rewrite!(expr)
+            end
+        end
+    end
+
+    rewrite!(ctx.dag)
+
+    Generators.build!(ctx, Generators.BUILDSTAGE_PRINTING_ONLY)
+
+    generated_file = joinpath(@__DIR__, "..", "src", "liblexbor_api.jl")
+    content = read(generated_file, String)
+    content = replace(content, "_ = \"skip-this-code\"" => "")
+    write(generated_file, content)
+
+    JuliaFormatter.format_file(generated_file)
+end