diff --git a/Cargo.lock b/Cargo.lock
index de5b71b..3727b00 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -120,6 +120,18 @@ version = "1.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "baf1de4339761588bc0619e3cbc0120ee582ebb74b53b4efbf79117bd2da40fd"
+[[package]]
+name = "console"
+version = "0.15.8"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0e1f83fc076bd6dd27517eacdf25fef6c4dfe5f1d7448bafaaf3a26f13b5e4eb"
+dependencies = [
+ "encode_unicode",
+ "lazy_static",
+ "libc",
+ "windows-sys 0.52.0",
+]
+
[[package]]
name = "core-foundation"
version = "0.9.4"
@@ -191,6 +203,12 @@ version = "0.6.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3a68a4904193147e0a8dec3314640e6db742afd5f6e634f428a6af230d9b3591"
+[[package]]
+name = "encode_unicode"
+version = "0.3.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a357d28ed41a50f9c765dbfe56cbc04a64e53e5fc58ba79fbc34c10ef3df831f"
+
[[package]]
name = "encoding_rs"
version = "0.8.34"
@@ -564,6 +582,19 @@ dependencies = [
"hashbrown",
]
+[[package]]
+name = "insta"
+version = "1.39.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "810ae6042d48e2c9e9215043563a58a80b877bc863228a74cf10c49d4620a6f5"
+dependencies = [
+ "console",
+ "lazy_static",
+ "linked-hash-map",
+ "serde",
+ "similar",
+]
+
[[package]]
name = "ipnet"
version = "2.9.0"
@@ -585,12 +616,24 @@ dependencies = [
"wasm-bindgen",
]
+[[package]]
+name = "lazy_static"
+version = "1.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "bbd2bcb4c963f2ddae06a2efc7e9f3591312473c50c6685e1f298068316e66fe"
+
[[package]]
name = "libc"
version = "0.2.155"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "97b3888a4aecf77e811145cadf6eef5901f4782c53886191b2f693f24761847c"
+[[package]]
+name = "linked-hash-map"
+version = "0.5.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0717cef1bc8b636c6e1c1bbdefc09e6322da8a9321966e8928ef80d20f7f770f"
+
[[package]]
name = "linux-raw-sys"
version = "0.4.14"
@@ -1146,6 +1189,7 @@ version = "0.2.0"
dependencies = [
"anyhow",
"futures",
+ "insta",
"regex",
"reqwest",
"scrapelect-filter-proc-macro",
@@ -1160,7 +1204,6 @@ dependencies = [
name = "scrapelect-filter-proc-macro"
version = "0.2.0"
dependencies = [
- "insta",
"quote",
"syn 2.0.64",
]
@@ -1284,6 +1327,12 @@ dependencies = [
"libc",
]
+[[package]]
+name = "similar"
+version = "2.6.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1de1d4f81173b03af4c0cbed3c898f6bff5b870e4a7f5d6f4057d62a7a4b686e"
+
[[package]]
name = "siphasher"
version = "0.3.11"
diff --git a/Cargo.toml b/Cargo.toml
index 6ef7135..3984260 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -21,3 +21,13 @@ url = "2.5"
[workspace]
members = [".", "filter-proc-macro"]
+
+[dev-dependencies]
+insta = { version = "1.39.0", features = ["json"] }
+
+# compile insta and similar in release mode to improve testing times
+[profile.dev.package.insta]
+opt-level = 3
+
+[profile.dev.package.similar]
+opt-level = 3
diff --git a/examples/inputs/recurser.html b/examples/inputs/recurser.html
new file mode 100644
index 0000000..20ee881
--- /dev/null
+++ b/examples/inputs/recurser.html
@@ -0,0 +1,6 @@
+
+
+
+ Cool link!
+
+
diff --git a/examples/outputs/abc.json b/examples/outputs/abc.snap
similarity index 63%
rename from examples/outputs/abc.json
rename to examples/outputs/abc.snap
index d081400..f20a0ad 100644
--- a/examples/outputs/abc.json
+++ b/examples/outputs/abc.snap
@@ -1,3 +1,7 @@
+---
+source: src/interpreter/mod.rs
+expression: result
+---
{
"header": {
"link": {
diff --git a/examples/outputs/attr.json b/examples/outputs/attr.snap
similarity index 68%
rename from examples/outputs/attr.json
rename to examples/outputs/attr.snap
index cecc0a0..505aaa7 100644
--- a/examples/outputs/attr.json
+++ b/examples/outputs/attr.snap
@@ -1,3 +1,7 @@
+---
+source: src/interpreter/mod.rs
+expression: result
+---
{
"link": [
{
diff --git a/examples/outputs/qualifiers.json b/examples/outputs/qualifiers.snap
similarity index 91%
rename from examples/outputs/qualifiers.json
rename to examples/outputs/qualifiers.snap
index 5e4ec0b..e4cb25e 100644
--- a/examples/outputs/qualifiers.json
+++ b/examples/outputs/qualifiers.snap
@@ -1,3 +1,7 @@
+---
+source: src/interpreter/mod.rs
+expression: result
+---
{
"items": [
{
diff --git a/examples/outputs/recurser.snap b/examples/outputs/recurser.snap
new file mode 100644
index 0000000..23047a4
--- /dev/null
+++ b/examples/outputs/recurser.snap
@@ -0,0 +1,10 @@
+---
+source: src/interpreter/mod.rs
+expression: result
+---
+{
+ "link": "https://example.com",
+ "new_page": {
+ "content": "Example Domain"
+ }
+}
diff --git a/examples/outputs/relative.json b/examples/outputs/relative.snap
similarity index 52%
rename from examples/outputs/relative.json
rename to examples/outputs/relative.snap
index 25e6547..1bb52a1 100644
--- a/examples/outputs/relative.json
+++ b/examples/outputs/relative.snap
@@ -1,3 +1,7 @@
+---
+source: src/interpreter/mod.rs
+expression: result
+---
{
"href": "./relative2.html",
"relative": "You found me!"
diff --git a/examples/scrps/recurser.scrp b/examples/scrps/recurser.scrp
new file mode 100644
index 0000000..27b05f7
--- /dev/null
+++ b/examples/scrps/recurser.scrp
@@ -0,0 +1,7 @@
+link: a {
+ href: $element | attrs() | take(key: "href");
+} | take(key: "href");
+
+new_page: <$link> h1 {
+ content: $text;
+};
diff --git a/filter-proc-macro/Cargo.toml b/filter-proc-macro/Cargo.toml
index c237657..8f2bf94 100644
--- a/filter-proc-macro/Cargo.toml
+++ b/filter-proc-macro/Cargo.toml
@@ -1,7 +1,11 @@
[package]
-name = "filter-proc-macro"
+name = "scrapelect-filter-proc-macro"
version = "0.2.0"
edition = "2021"
+authors = ["Max Carr "]
+license = "Apache-2.0 OR MIT"
+repository = "https://github.com/suaviloquence/scrapelect"
+description = "proc-macros for making scrapelect Filters"
[lib]
proc-macro = true
diff --git a/src/interpreter/mod.rs b/src/interpreter/mod.rs
index a16f0b6..1cacdd5 100644
--- a/src/interpreter/mod.rs
+++ b/src/interpreter/mod.rs
@@ -346,9 +346,6 @@ mod tests {
async fn integration_test(filename: &str) -> anyhow::Result<()> {
let input = std::fs::read_to_string(format!("examples/inputs/{filename}.html"))?;
let script = std::fs::read_to_string(format!("examples/scrps/{filename}.scrp"))?;
- let output: serde_json::Value = serde_json::from_reader(std::fs::File::open(format!(
- "examples/outputs/{filename}.json"
- ))?)?;
let (ast, head) = crate::frontend::Parser::new(&script)
.parse()
@@ -369,9 +366,15 @@ mod tests {
.parse()
.expect("parse URL failed"),
)
- .await?;
- let result = serde_json::to_value(result.0)?;
- assert_eq!(output, result);
+ .await?.0;
+
+ let mut settings = insta::Settings::clone_current();
+ settings.set_snapshot_path("../../examples/outputs");
+ settings.set_prepend_module_to_snapshot(false);
+
+ settings.bind(|| {
+ insta::assert_json_snapshot!(filename, result);
+ });
Ok(())
}
@@ -454,5 +457,6 @@ mod tests {
attr,
qualifiers,
relative,
+ recurser,
}
}