feedparser-rs 0.6.0__tar.gz → 0.7.0__tar.gz

{feedparser_rs-0.6.0 → feedparser_rs-0.7.0}/Cargo.lock

@@ -155,7 +155,7 @@ dependencies = [
  "addr2line",
  "cfg-if",
  "libc",
- "miniz_oxide",
+ "miniz_oxide 0.8.9",
  "object",
  "rustc-demangle",
  "windows-link",
@@ -582,7 +582,7 @@ checksum = "da7c62ceae207dd37ea5b845da6a0696c799f85e97da1ab5b7910be3c1c80223"
 
 [[package]]
 name = "feedparser-rs"
-version = "0.6.0"
+version = "0.7.0"
 dependencies = [
  "ammonia",
  "chrono",
@@ -592,10 +592,8 @@ dependencies = [
  "encoding_rs",
  "flate2",
  "html-escape",
- "memchr",
  "mockito",
  "quick-xml",
- "regex",
  "reqwest",
  "serde",
  "serde_json",
@@ -606,7 +604,7 @@ dependencies = [
 
 [[package]]
 name = "feedparser-rs-node"
-version = "0.6.0"
+version = "0.7.0"
 dependencies = [
  "chrono",
  "feedparser-rs",
@@ -620,12 +618,12 @@ dependencies = [
 
 [[package]]
 name = "feedparser-rs-py"
-version = "0.6.0"
+version = "0.7.0"
 dependencies = [
  "chrono",
  "feedparser-rs",
- "once_cell",
  "pyo3",
+ "serde_json",
 ]
 
 [[package]]
@@ -636,12 +634,13 @@ checksum = "d45db016d36b838f563236e9193d0ee6ce38f3f68b6c94e914b4929c96bbb890"
 
 [[package]]
 name = "flate2"
-version = "1.1.9"
+version = "1.1.10"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "843fba2746e448b37e26a819579957415c8cef339bf08564fe8b7ddbd959573c"
+checksum = "6e634e2e0ebac1ee034020da1ca582e17ffe4e0f5e985823721e168928136dcb"
 dependencies = [
  "crc32fast",
- "miniz_oxide",
+ "miniz_oxide 0.9.1",
+ "zlib-rs",
 ]
 
 [[package]]
@@ -800,9 +799,9 @@ checksum = "e629b9b98ef3dd8afe6ca2bd0f89306cec16d43d907889945bc5d6687f2f13c7"
 
 [[package]]
 name = "h2"
-version = "0.4.15"
+version = "0.4.16"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6cb093c84e8bd9b188d4c4a8cb6579fc016968d14c99882163cd3ff402a4f155"
+checksum = "a9f37a958b41b3b19ee2707c06439c0e9e547e847223eb791ecb0cb821c65e27"
 dependencies = [
  "atomic-waker",
  "bytes",
@@ -1267,6 +1266,15 @@ name = "miniz_oxide"
 version = "0.8.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "1fa76a2c86f704bdb222d66965fb3d63269ce38518b83cb0575fca855ebb6316"
+dependencies = [
+ "adler2",
+]
+
+[[package]]
+name = "miniz_oxide"
+version = "0.9.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b63fbc4a50860e98e7b2aa7804ded1db5cbc3aff9193adaff57a6931bf7c4b4c"
 dependencies = [
  "adler2",
  "simd-adler32",
@@ -1316,9 +1324,9 @@ dependencies = [
 
 [[package]]
 name = "napi"
-version = "3.12.1"
+version = "3.12.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "459197f1592f4c3dbbf9c1b13f5a4599a343e4ef66b96bc340e2a518b36a6662"
+checksum = "58c5f4d5375213fdb7be2655e152386e82f026f9a5ba36a75556e11359aafe09"
 dependencies = [
  "anyhow",
  "bitflags",
@@ -1329,6 +1337,8 @@ dependencies = [
  "napi-sys",
  "nohash-hasher",
  "rustc-hash 2.1.3",
+ "serde",
+ "serde_json",
 ]
 
 [[package]]
@@ -1638,9 +1648,9 @@ dependencies = [
 
 [[package]]
 name = "quick-xml"
-version = "0.41.0"
+version = "0.42.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e660451e55124f798a69a5af3f49ccfbefbd41910eefd25caf2393e1f3473ec1"
+checksum = "41b1177fdf999d2321d3fb46ff47159d9c1fb9ad66a4879f8c50a0b504615e9b"
 dependencies = [
  "memchr",
 ]
@@ -2898,6 +2908,12 @@ dependencies = [
  "syn 3.0.3",
 ]
 
+[[package]]
+name = "zlib-rs"
+version = "0.6.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "34b31d188d9d685a4f9c7b46d6e36631b07058d2cfe190267adce54dc230bf12"
+
 [[package]]
 name = "zmij"
 version = "1.0.23"

{feedparser_rs-0.6.0 → feedparser_rs-0.7.0}/Cargo.toml

@@ -3,9 +3,9 @@ members = ["crates/feedparser-rs-core", "crates/feedparser-rs-py"]
 resolver = "2"
 
 [workspace.package]
-version = "0.6.0"
+version = "0.7.0"
 edition = "2024"
-rust-version = "1.88.0"
+rust-version = "1.91.0"
 authors = ["bug-ops"]
 license = "MIT OR Apache-2.0"
 repository = "https://github.com/bug-ops/feedparser-rs"
@@ -21,14 +21,11 @@ criterion = "0.8"
 encoding_rs = "0.8"
 flate2 = "1.1"
 html-escape = "0.2"
-memchr = "2.8"
 mockito = "1.7"
 napi = "3.12"
 napi-derive = "3.6"
-once_cell = "1.21"
 pyo3 = "0.29"
-quick-xml = "0.41"
-regex = "1.13"
+quick-xml = "0.42"
 reqwest = { version = "0.13", default-features = false }
 serde = "1.0"
 serde_json = "1.0"

{feedparser_rs-0.6.0 → feedparser_rs-0.7.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: feedparser-rs
-Version: 0.6.0
+Version: 0.7.0
 Classifier: Development Status :: 4 - Beta
 Classifier: Intended Audience :: Developers
 Classifier: License :: OSI Approved :: MIT License
@@ -38,6 +38,7 @@ High-performance RSS/Atom/JSON Feed parser for Python with feedparser-compatible
 - **Sanitized by default**: HTML-bearing fields are sanitized against XSS unless explicitly disabled via `sanitize_html=False`
 - **Multi-format**: RSS 0.9x/1.0/2.0, Atom 0.3/1.0, JSON Feed 1.0/1.1
 - **Podcast support**: iTunes and Podcast 2.0 namespace extensions
+- **JSON Feed extensions**: `_`-prefixed custom objects (e.g. `_cast`) captured as `feed.json_extensions`/`entry.json_extensions` dicts
 - **feedparser-compatible**: Dict-style access, field aliases, same API patterns
 - **DoS protection**: Built-in resource limits
 

{feedparser_rs-0.6.0/crates/feedparser-rs-py → feedparser_rs-0.7.0}/README.md

@@ -15,6 +15,7 @@ High-performance RSS/Atom/JSON Feed parser for Python with feedparser-compatible
 - **Sanitized by default**: HTML-bearing fields are sanitized against XSS unless explicitly disabled via `sanitize_html=False`
 - **Multi-format**: RSS 0.9x/1.0/2.0, Atom 0.3/1.0, JSON Feed 1.0/1.1
 - **Podcast support**: iTunes and Podcast 2.0 namespace extensions
+- **JSON Feed extensions**: `_`-prefixed custom objects (e.g. `_cast`) captured as `feed.json_extensions`/`entry.json_extensions` dicts
 - **feedparser-compatible**: Dict-style access, field aliases, same API patterns
 - **DoS protection**: Built-in resource limits
 

{feedparser_rs-0.6.0 → feedparser_rs-0.7.0}/crates/feedparser-rs-core/Cargo.toml

@@ -16,9 +16,7 @@ chrono = { workspace = true, features = ["std", "clock"] }
 compact_str.workspace = true
 encoding_rs.workspace = true
 html-escape.workspace = true
-memchr.workspace = true
 quick-xml.workspace = true
-regex.workspace = true
 reqwest = { workspace = true, features = [
     "blocking",
     "gzip",

{feedparser_rs-0.6.0 → feedparser_rs-0.7.0}/crates/feedparser-rs-core/README.md

@@ -19,6 +19,7 @@ This is the core parsing library that powers the Python and Node.js bindings.
 - **HTTP support**: Fetch feeds from URLs with compression and conditional GET
 - **Podcast support**: iTunes and Podcast 2.0 namespace extensions
 - **Namespace extensions**: Dublin Core, Media RSS (`media:title`, `media:description`, `media:content`, `media:thumbnail`, `media:credit`, `media:copyright`, `media:rating`, `media:keywords`), Syndication, GeoRSS, Creative Commons
+- **JSON Feed extension objects**: generic capture of underscore-prefixed custom objects (e.g. `_cast`) at feed- and item-level via `json_extensions`
 - **Well-tested**: 91%+ test coverage with real-world feed fixtures
 
 ## Installation
@@ -35,7 +36,7 @@ feedparser-rs = "0.6"
 ```
 
 > [!IMPORTANT]
-> Requires Rust 1.88.0 or later (edition 2024).
+> Requires Rust 1.91.0 or later (edition 2024).
 
 ## Quick Start
 
@@ -190,7 +191,7 @@ cargo bench
 
 ## MSRV Policy
 
-Minimum Supported Rust Version: **1.88.0** (edition 2024).
+Minimum Supported Rust Version: **1.91.0** (edition 2024).
 
 MSRV increases are considered breaking changes and will result in a minor version bump.
 

{feedparser_rs-0.6.0 → feedparser_rs-0.7.0}/crates/feedparser-rs-core/src/http/client.rs

@@ -537,7 +537,7 @@ mod tests {
 
     #[test]
     fn test_custom_timeout() {
-        let timeout = Duration::from_secs(60);
+        let timeout = Duration::from_mins(1);
         let client = FeedHttpClient::new().unwrap().with_timeout(timeout);
         assert_eq!(client.timeout, timeout);
     }
@@ -547,7 +547,7 @@ mod tests {
         // `Duration::MAX` would overflow `Instant::now() + timeout` inside
         // reqwest's blocking wait and panic; it must be clamped instead.
         let client = FeedHttpClient::new().unwrap().with_timeout(Duration::MAX);
-        assert_eq!(client.timeout, Duration::from_secs(3600));
+        assert_eq!(client.timeout, Duration::from_hours(1));
     }
 
     #[test]
@@ -558,7 +558,7 @@ mod tests {
         let timeout = Duration::from_secs(7);
         let client = FeedHttpClient::new().unwrap().with_timeout(timeout);
         let request = client
-            .build_request("http://example.test/feed.xml", HeaderMap::new())
+            .build_request("https://example.test/feed.xml", HeaderMap::new())
             .unwrap();
         assert_eq!(request.timeout(), Some(&timeout));
     }

{feedparser_rs-0.6.0 → feedparser_rs-0.7.0}/crates/feedparser-rs-core/src/limits.rs

@@ -110,7 +110,10 @@ pub struct ParserLimits {
 
     /// Maximum text field length in bytes
     ///
-    /// Prevents excessive memory from huge title/description fields.
+    /// Prevents excessive memory from huge title/description fields. Also bounds
+    /// the per-value serialized size of a captured JSON Feed extension object
+    /// (see `max_json_extensions`); an oversized extension value is dropped
+    /// entirely rather than truncated.
     ///
     /// Default: 10 MB
     pub max_text_length: usize,
@@ -209,6 +212,20 @@ pub struct ParserLimits {
     ///
     /// Default: 20
     pub max_podcast_value_time_splits: usize,
+
+    /// Maximum number of JSON Feed extension objects captured per scope
+    ///
+    /// JSON Feed permits custom object keys (starting with `_` followed by a
+    /// letter) anywhere in a feed; this bounds how many distinct keys are
+    /// captured per feed-root or item-root scope, counted independently. When a
+    /// scope has more extension keys than this limit, the surviving keys are the
+    /// alphabetically-first ones, not the document-order-first ones: `serde_json`
+    /// backs a JSON object with a `BTreeMap` by default (the `preserve_order`
+    /// feature, which would switch it to an `IndexMap` and make truncation
+    /// document-order-first, is not enabled).
+    ///
+    /// Default: 100
+    pub max_json_extensions: usize,
 }
 
 impl Default for ParserLimits {
@@ -245,6 +262,7 @@ impl Default for ParserLimits {
             max_podcast_follow: 20,
             max_podcast_chat: 20,
             max_podcast_value_time_splits: 20,
+            max_json_extensions: 100,
         }
     }
 }
@@ -293,6 +311,7 @@ impl ParserLimits {
             max_podcast_follow: 5,
             max_podcast_chat: 5,
             max_podcast_value_time_splits: 5,
+            max_json_extensions: 20,
         }
     }
 
@@ -339,6 +358,7 @@ impl ParserLimits {
             max_podcast_follow: 100,
             max_podcast_chat: 100,
             max_podcast_value_time_splits: 50,
+            max_json_extensions: 500,
         }
     }
 
@@ -635,4 +655,19 @@ mod tests {
         assert!(result.is_err());
         assert!(matches!(result, Err(LimitError::CollectionTooLarge { .. })));
     }
+
+    #[test]
+    fn test_max_json_extensions_tiers() {
+        assert_eq!(ParserLimits::default().max_json_extensions, 100);
+        assert_eq!(ParserLimits::strict().max_json_extensions, 20);
+        assert_eq!(ParserLimits::permissive().max_json_extensions, 500);
+        assert!(
+            ParserLimits::strict().max_json_extensions
+                < ParserLimits::default().max_json_extensions
+        );
+        assert!(
+            ParserLimits::permissive().max_json_extensions
+                > ParserLimits::default().max_json_extensions
+        );
+    }
 }

{feedparser_rs-0.6.0 → feedparser_rs-0.7.0}/crates/feedparser-rs-core/src/namespace/threading.rs

@@ -69,10 +69,10 @@ pub fn parse_in_reply_to_from_attrs<'a>(
             continue;
         };
         match attr.key.as_ref() {
-            b"ref" => ref_ = non_empty(&value).map(|s| s.to_string().into()),
-            b"href" => href = non_empty(&value).map(|s| s.to_string().into()),
-            b"type" => type_ = non_empty(&value).map(|s| s.to_string().into()),
-            b"source" => source = non_empty(&value).map(|s| s.to_string().into()),
+            "ref" => ref_ = non_empty(&value).map(|s| s.to_string().into()),
+            "href" => href = non_empty(&value).map(|s| s.to_string().into()),
+            "type" => type_ = non_empty(&value).map(|s| s.to_string().into()),
+            "source" => source = non_empty(&value).map(|s| s.to_string().into()),
             _ => {}
         }
     }