feat: Add domain rewrite in body sample plugin

plugins/README.md

@@ -138,6 +138,9 @@ for your own plugin. Extend them to fit your particular use case.
     initial numbers will be masked.
 *   [Validate client token on query string using HMAC](samples/hmac_authtoken):
     Check the client request URL for a valid token signed using HMAC.
+*   [Rewrite domains in html response body](samples/html_domain_rewrite/): Parse
+    html in response body chunks and replace insances of "foo.com" with
+    "bar.com" in `<a href=***>`.
 
 # Feature set / ABI
 

plugins/bazel/cargo/BUILD

@@ -32,5 +32,5 @@ crates_vendor(
     repository_name = "cu",  # shorten generated paths for Windows...
     tags = ["manual"],
     vendor_path = "remote",
-    supported_platform_triples = ["x86_64-unknown-linux-gnu", "wasm32-wasi"],
+    supported_platform_triples = ["x86_64-unknown-linux-gnu", "wasm32-wasi"], # needed x86_64-unknown-linux-gnu to fix issue with getrandom building for wrong environment
 )

plugins/bazel/cargo/Cargo.Bazel.lock

@@ -143,7 +143,18 @@ checksum = "8fc3cb4d91f53b50155bdcfd23f6a4c39ae1969c2ae85982b135750cccaf5fce"
 dependencies = [
  "cfg-if",
  "libc",
- "wasi",
+ "wasi 0.9.0+wasi-snapshot-preview1",
+]
+
+[[package]]
+name = "getrandom"
+version = "0.2.15"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c4567c8db10ae91089c99af84c68c38da3ec2f087c3f82960bcdbf3656b6f4d7"
+dependencies = [
+ "cfg-if",
+ "libc",
+ "wasi 0.11.0+wasi-snapshot-preview1",
 ]
 
 [[package]]
@@ -359,7 +370,7 @@ version = "0.7.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "6a6b1679d49b24bbfe0c803429aa1874472f50d9b363131f0e89fc356b544d03"
 dependencies = [
- "getrandom",
+ "getrandom 0.1.16",
  "libc",
  "rand_chacha",
  "rand_core",
@@ -383,7 +394,7 @@ version = "0.5.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "90bde5296fc891b0cef12a6d03ddccc162ce7b2aff54160af9338f8d40df6d19"
 dependencies = [
- "getrandom",
+ "getrandom 0.1.16",
 ]
 
 [[package]]
@@ -474,7 +485,6 @@ version = "0.0.0"
 dependencies = [
  "log",
  "lol_html",
- "ppv-lite86",
  "proxy-wasm",
  "regex",
  "url",
@@ -610,7 +620,7 @@ version = "1.12.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b3758f5e68192bb96cc8f9b7e2c2cfdabb435499a28499a42f8f984092adad4b"
 dependencies = [
- "getrandom",
+ "getrandom 0.2.15",
 ]
 
 [[package]]
@@ -625,6 +635,12 @@ version = "0.9.0+wasi-snapshot-preview1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "cccddf32554fecc6acb585f82a32a72e28b48f8c4c1883ddfeeeaa96f7d8e519"
 
+[[package]]
+name = "wasi"
+version = "0.11.0+wasi-snapshot-preview1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9c8d87e72b64a3b4db28d11ce29237c246188f4f51057d65a7eab63b7987e423"
+
 [[package]]
 name = "zerocopy"
 version = "0.6.6"

plugins/bazel/cargo/Cargo.toml

@@ -10,8 +10,9 @@ proxy-wasm = "0.2.1"
 log = "0.4"
 regex = "1.9"
 url = "2.4"
+# Pin version 2.0.0 until rust-version includes https://github.com/rust-lang/rust/issues/119128
 lol_html = "=2.0.0"
-ppv-lite86 = "=0.2.18"
+uuid = { version = "1.12.1", features = [ "v4" ] }
 
 [lib]
 crate-type = ["cdylib"]

plugins/bazel/cargo/remote/BUILD.bazel

@@ -43,12 +43,6 @@ alias(
     tags = ["manual"],
 )
 
-alias(
-    name = "ppv-lite86",
-    actual = "@cu__ppv-lite86-0.2.18//:ppv_lite86",
-    tags = ["manual"],
-)
-
 alias(
     name = "proxy-wasm",
     actual = "@cu__proxy-wasm-0.2.1//:proxy_wasm",

plugins/bazel/cargo/remote/BUILD.getrandom-0.2.15.bazel

@@ -47,6 +47,9 @@ rust_library(
         "@rules_rust//rust/platform:wasm32-wasi": [
             "@cu__wasi-0.11.0-wasi-snapshot-preview1//:wasi",  # cfg(target_os = "wasi")
         ],
+        "@rules_rust//rust/platform:x86_64-unknown-linux-gnu": [
+            "@cu__libc-0.2.169//:libc",  # cfg(unix)
+        ],
         "//conditions:default": [],
     }),
 )

plugins/bazel/cargo/remote/BUILD.ppv-lite86-0.2.18.bazel

@@ -29,7 +29,6 @@ rust_library(
         ],
     ),
     crate_features = [
-        "default",
         "simd",
         "std",
     ],

plugins/bazel/cargo/remote/defs.bzl

@@ -297,7 +297,6 @@ _NORMAL_DEPENDENCIES = {
         _COMMON_CONDITION: {
             "log": Label("@cu__log-0.4.25//:log"),
             "lol_html": Label("@cu__lol_html-2.0.0//:lol_html"),
-            "ppv-lite86": Label("@cu__ppv-lite86-0.2.18//:ppv_lite86"),
             "proxy-wasm": Label("@cu__proxy-wasm-0.2.1//:proxy_wasm"),
             "regex": Label("@cu__regex-1.9.6//:regex"),
             "url": Label("@cu__url-2.4.1//:url"),
@@ -542,6 +541,16 @@ def crate_repositories():
         build_file = Label("@//bazel/cargo/remote:BUILD.getrandom-0.1.16.bazel"),
     )
 
+    maybe(
+        http_archive,
+        name = "cu__getrandom-0.2.15",
+        sha256 = "c4567c8db10ae91089c99af84c68c38da3ec2f087c3f82960bcdbf3656b6f4d7",
+        type = "tar.gz",
+        urls = ["https://static.crates.io/crates/getrandom/0.2.15/download"],
+        strip_prefix = "getrandom-0.2.15",
+        build_file = Label("@//bazel/cargo/remote:BUILD.getrandom-0.2.15.bazel"),
+    )
+
     maybe(
         http_archive,
         name = "cu__hashbrown-0.13.2",
@@ -1072,6 +1081,16 @@ def crate_repositories():
         build_file = Label("@//bazel/cargo/remote:BUILD.version_check-0.9.5.bazel"),
     )
 
+    maybe(
+        http_archive,
+        name = "cu__wasi-0.11.0-wasi-snapshot-preview1",
+        sha256 = "9c8d87e72b64a3b4db28d11ce29237c246188f4f51057d65a7eab63b7987e423",
+        type = "tar.gz",
+        urls = ["https://static.crates.io/crates/wasi/0.11.0+wasi-snapshot-preview1/download"],
+        strip_prefix = "wasi-0.11.0+wasi-snapshot-preview1",
+        build_file = Label("@//bazel/cargo/remote:BUILD.wasi-0.11.0+wasi-snapshot-preview1.bazel"),
+    )
+
     maybe(
         http_archive,
         name = "cu__wasi-0.9.0-wasi-snapshot-preview1",
@@ -1125,7 +1144,6 @@ def crate_repositories():
     return [
         struct(repo = "cu__log-0.4.25", is_dev_dep = False),
         struct(repo = "cu__lol_html-2.0.0", is_dev_dep = False),
-        struct(repo = "cu__ppv-lite86-0.2.18", is_dev_dep = False),
         struct(repo = "cu__proxy-wasm-0.2.1", is_dev_dep = False),
         struct(repo = "cu__regex-1.9.6", is_dev_dep = False),
         struct(repo = "cu__url-2.4.1", is_dev_dep = False),

plugins/samples/domain_rewrite/plugin.rs → plugins/samples/html_domain_rewrite/plugin.rs

@@ -12,7 +12,7 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-// [START serviceextensions_plugin_domain_rewrite]
+// [START serviceextensions_plugin_html_domain_rewrite]
 use lol_html::*;
 use proxy_wasm::traits::*;
 use proxy_wasm::types::*;
@@ -27,14 +27,20 @@ proxy_wasm::main! {{
 
 struct MyHttpContext<'a> {
     // Some member variables need to be wrapped in Rc<Refcell<T>> so that we can
-    // modify the variable in multiple places as well as share ownership. By
-    // using Rc, we are commiting to not modfying the variable in multiple places
-    // simultaneously, otherwise the code will panic and plugin will crash.
+    // modify the variable in multiple places as well as share ownership.
+    // By using Rc, we are allowing multiple references to the same obeject to
+    // exist at the same time.
+    // By using Refcell, we are evaluating borrow checks at runtime, thus we are
+    // commiting to not modfying the variable in multiple places simultaneously,
+    // otherwise the code will panic and plugin will crash.
 
     // Stores HTML as rewriter parses and modifies HTML.
     // Only stores completed sections(e.g., when rewriter parses "<di", output
     // will be empty until the next chunk comes in. If the next chunk was
     // "v> <h1>foo</h", the output will then contain "<div><h1>foo").
+    // Only stores sections of HTML that have not yet been seen back to client.
+    // After sending data to client, we clear output to avoid unnecessary memory
+    // growth.
     output: Rc<RefCell<Vec<u8>>>,
     // HTML rewriter. Member of MyHttpContext a.k.a "StreamContext" so that the
     // rewriter persists across multiple body callbacks.
@@ -50,7 +56,7 @@ impl<'a> MyHttpContext<'a> {
                 Settings {
                     element_content_handlers: vec![element!("a[href]", move |el| {
                         let href = el.get_attribute("href").unwrap();
-                        let modified_href = href.replace("foo", "bar");
+                        let modified_href = href.replace("foo.com", "bar.com");
                         if modified_href != href {
                             el.set_attribute("href", &modified_href).unwrap();
                         }
@@ -100,4 +106,4 @@ impl<'a> HttpContext for MyHttpContext<'a> {
         return Action::Continue;
     }
 }
-// [END serviceextensions_plugin_domain_rewrite]
+// [END serviceextensions_plugin_html_domain_rewrite]