Skip to main content

feedparser_rs/parser/
mod.rs

1pub mod atom;
2mod common;
3mod detect;
4pub mod json;
5pub mod namespace_detection;
6pub mod rss;
7pub mod rss10;
8
9use crate::{error::Result, types::ParsedFeed};
10
11pub use common::skip_element;
12pub use detect::detect_format;
13
14/// Parse feed from raw bytes
15///
16/// This is the main entry point for parsing feeds. It automatically detects
17/// the feed format (RSS, Atom, JSON) and parses accordingly.
18///
19/// # Errors
20///
21/// Returns a `FeedError` if the feed cannot be parsed. However, in most cases,
22/// the parser will set the `bozo` flag and return partial results rather than
23/// returning an error.
24///
25/// # Examples
26///
27/// ```
28/// use feedparser_rs::parse;
29///
30/// let xml = r#"
31///     <?xml version="1.0"?>
32///     <rss version="2.0">
33///         <channel>
34///             <title>Example Feed</title>
35///         </channel>
36///     </rss>
37/// "#;
38///
39/// let feed = parse(xml.as_bytes()).unwrap();
40/// assert_eq!(feed.feed.title.as_deref(), Some("Example Feed"));
41/// ```
42pub fn parse(data: &[u8]) -> Result<ParsedFeed> {
43    parse_with_limits(data, crate::ParserLimits::default())
44}
45
46/// Parse feed with custom parser limits
47///
48/// This allows controlling resource usage when parsing untrusted feeds.
49///
50/// # Examples
51///
52/// ```
53/// use feedparser_rs::{parse_with_limits, ParserLimits};
54///
55/// let xml = b"<rss version=\"2.0\"><channel><title>Test</title></channel></rss>";
56/// let limits = ParserLimits::strict();
57/// let feed = parse_with_limits(xml, limits).unwrap();
58/// ```
59///
60/// # Errors
61///
62/// Returns an error if:
63/// - Feed size exceeds limits
64/// - Format is unknown or unsupported
65/// - Fatal parsing error occurs
66pub fn parse_with_limits(data: &[u8], limits: crate::ParserLimits) -> Result<ParsedFeed> {
67    use crate::types::FeedVersion;
68    use crate::util::encoding::detect_and_convert;
69
70    // Detect encoding and convert to UTF-8 before parsing.
71    // This handles ISO-8859-1, Windows-1252, UTF-16, and BOM-prefixed feeds.
72    let (utf8_string, detected_encoding) = detect_and_convert(data)
73        .unwrap_or_else(|_| (String::from_utf8_lossy(data).into_owned(), "UTF-8"));
74
75    let utf8_bytes = utf8_string.as_bytes();
76    let encoding_label = detected_encoding.to_lowercase();
77
78    // Detect format on UTF-8 data (required for correct UTF-16 detection)
79    let version = detect_format(utf8_bytes);
80
81    // Parse based on detected format, then update the encoding field
82    let mut feed = match version {
83        // RSS variants (all use RSS 2.0 parser for now)
84        FeedVersion::Rss20 | FeedVersion::Rss092 | FeedVersion::Rss091 | FeedVersion::Rss090 => {
85            rss::parse_rss20_with_limits(utf8_bytes, limits)
86        }
87
88        // Atom variants
89        FeedVersion::Atom10 | FeedVersion::Atom03 => {
90            atom::parse_atom10_with_limits(utf8_bytes, limits)
91        }
92
93        // RSS 1.0 (RDF)
94        FeedVersion::Rss10 => rss10::parse_rss10_with_limits(utf8_bytes, limits),
95
96        // JSON Feed
97        FeedVersion::JsonFeed10 | FeedVersion::JsonFeed11 => {
98            json::parse_json_feed_with_limits(utf8_bytes, limits)
99        }
100
101        // Unknown format - return a bozo feed since the format is unrecognized.
102        // The bozo pattern requires we never panic and always return partial data,
103        // but unrecognizable input must signal the caller via bozo=true.
104        FeedVersion::Unknown => {
105            let mut feed = crate::types::ParsedFeed::new();
106            feed.version = FeedVersion::Unknown;
107            feed.bozo = true;
108            feed.bozo_exception = Some("Feed format not recognized".to_string());
109            Ok(feed)
110        }
111    }?;
112
113    feed.encoding = encoding_label;
114    Ok(feed)
115}
116
117#[cfg(test)]
118mod tests {
119    use super::*;
120
121    #[test]
122    fn test_parse_returns_ok_bozo_for_garbage() {
123        let feed = parse(b"test").unwrap();
124        assert!(feed.bozo, "unrecognized input must set bozo");
125        assert_eq!(feed.version, crate::types::FeedVersion::Unknown);
126        assert!(feed.entries.is_empty());
127    }
128}