Return partial RSS feeds, rss_loose feature, resolves #22

.travis.yml

@@ -4,6 +4,9 @@ rust:
   - nightly
   - beta
   - stable
+script:
+    - cargo test --verbose
+    - cargo test --verbose --features rss_loose
 after_success: |
   cargo doc && \
   echo '<meta http-equiv=refresh content=0;url=rss/index.html>' > target/doc/index.html && \

Cargo.toml

@@ -1,5 +1,4 @@
 [package]
-
 name = "rss"
 version = "0.3.1"
 authors = ["Corey Farwell <[email protected]>"]
@@ -10,6 +9,9 @@ description = "Library for serializing the RSS web content syndication format"
 keywords = ["rss", "feed", "blog", "web", "news"]
 exclude = ["test-data/"]
 
-
 [dependencies]
 RustyXML = "0.1"
+
+[features]
+rss_loose = []
+

README.md

@@ -57,3 +57,24 @@ let rss_str = r#"
 
 let rss = rss_str.parse::<Rss>().unwrap();
 ```
+
+### Partial Feeds
+
+In some cases, the RSS source may not return a standards-compliant RSS such as a missing description tag. The library
+is designed to return an error in such cases, however this behaviour can be loosened by using the feature
+flag `rss_loose`.
+
+Using this flag changes what would normally be a `String` type to a `Option<String>`, just like other fields.
+
+In your `Cargo.toml`, add the following:
+```toml
+[dependencies.rss]
+features = ["rss_loose"]
+```
+
+## Contributors & License
+
+- Michael Yoo [GitHub](https://github.com/sekjun9878) [Web](https://www.michael.yoo.id.au/)
+
+Released under The Apache License 2.0
+

src/category.rs

@@ -45,3 +45,4 @@ impl ViaXml for Category {
         })
     }
 }
+

src/channel.rs

@@ -23,21 +23,42 @@ use ::{Category, ElementUtils, Item, Image, ReadError, TextInput, ViaXml};
 /// # Examples
 ///
 /// ```
+///# #![feature(stmt_expr_attributes)]
 /// use rss::Channel;
 ///
+///# #[cfg(not(feature = "rss_loose"))]
 /// let channel = Channel {
 ///     title: String::from("My Blog"),
 ///     link: String::from("http://myblog.com"),
 ///     description: String::from("My thoughts on life, the universe, and everything"),
 ///     items: vec![],
 ///     ..Default::default()
 /// };
+///# #[cfg(feature = "rss_loose")]
+///# let channel = Channel {
+///#     title: Some(String::from("My Blog")),
+///#     link: Some(String::from("http://myblog.com")),
+///#     description: Some(String::from("My thoughts on life, the universe, and everything")),
+///#     items: vec![],
+///#     ..Default::default()
+///# };
 /// ```
 #[derive(Default, Debug, Clone)]
 pub struct Channel {
+    #[cfg(not(feature = "rss_loose"))]
     pub title: String,
+    #[cfg(not(feature = "rss_loose"))]
     pub link: String,
+    #[cfg(not(feature = "rss_loose"))]
     pub description: String,
+
+    #[cfg(feature = "rss_loose")]
+    pub title: Option<String>,
+    #[cfg(feature = "rss_loose")]
+    pub link: Option<String>,
+    #[cfg(feature = "rss_loose")]
+    pub description: Option<String>,
+
     pub items: Vec<Item>,
     pub language: Option<String>,
     pub copyright: Option<String>,
@@ -61,10 +82,20 @@ impl ViaXml for Channel {
     fn to_xml(&self) -> Element {
         let mut channel = Element::new("channel".to_owned(), None, vec![]);
 
+        #[cfg(not(feature = "rss_loose"))]
         channel.tag_with_text("title", self.title.clone());
+        #[cfg(not(feature = "rss_loose"))]
         channel.tag_with_text("link", self.link.clone());
+        #[cfg(not(feature = "rss_loose"))]
         channel.tag_with_text("description", self.description.clone());
 
+        #[cfg(feature = "rss_loose")]
+        channel.tag_with_optional_text("title", self.title.clone());
+        #[cfg(feature = "rss_loose")]
+        channel.tag_with_optional_text("link", self.link.clone());
+        #[cfg(feature = "rss_loose")]
+        channel.tag_with_optional_text("description", self.description.clone());
+
         for item in &self.items {
             channel.tag(item.to_xml());
         }
@@ -95,21 +126,29 @@ impl ViaXml for Channel {
     }
 
     fn from_xml(elem: Element) -> Result<Self, ReadError> {
+        #[cfg(not(feature = "rss_loose"))]
         let title = match elem.get_child("title", None) {
             Some(elem) => elem.content_str(),
             None => return Err(ReadError::ChannelMissingTitle),
         };
-
+        #[cfg(not(feature = "rss_loose"))]
         let link = match elem.get_child("link", None) {
             Some(elem) => elem.content_str(),
             None => return Err(ReadError::ChannelMissingLink),
         };
-
+        #[cfg(not(feature = "rss_loose"))]
         let description = match elem.get_child("description", None) {
             Some(elem) => elem.content_str(),
             None => return Err(ReadError::ChannelMissingDescription),
         };
 
+        #[cfg(feature = "rss_loose")]
+        let title = elem.get_child("title", None).map(Element::content_str);
+        #[cfg(feature = "rss_loose")]
+        let link = elem.get_child("link", None).map(Element::content_str);
+        #[cfg(feature = "rss_loose")]
+        let description = elem.get_child("description", None).map(Element::content_str);
+
         let items = match elem.get_children("item", None)
                               .map(|e| ViaXml::from_xml(e.clone()))
                               .collect::<Result<Vec<_>, _>>()

src/image.rs

@@ -8,19 +8,42 @@ use ::{ElementUtils, ReadError, ViaXml};
 /// (http://cyber.law.harvard.edu/rss/rss.html#ltimagegtSubelementOfLtchannelgt)
 #[derive(Default, Debug, Clone)]
 pub struct Image {
+    #[cfg(not(feature = "rss_loose"))]
     pub url: String,
+    #[cfg(not(feature = "rss_loose"))]
     pub title: String,
+    #[cfg(not(feature = "rss_loose"))]
     pub link: String,
+
+    #[cfg(feature = "rss_loose")]
+    pub url: Option<String>,
+    #[cfg(feature = "rss_loose")]
+    pub title: Option<String>,
+    #[cfg(feature = "rss_loose")]
+    pub link: Option<String>,
+
     pub width: Option<u32>,
     pub height: Option<u32>,
 }
 
 impl ViaXml for Image {
     fn to_xml(&self) -> Element {
         let mut elem = Element::new("image".to_owned(), None, vec![]);
+
+        #[cfg(not(feature = "rss_loose"))]
         elem.tag_with_text("url", self.url.clone());
+        #[cfg(not(feature = "rss_loose"))]
         elem.tag_with_text("title", self.title.clone());
+        #[cfg(not(feature = "rss_loose"))]
         elem.tag_with_text("link", self.link.clone());
+
+        #[cfg(feature = "rss_loose")]
+        elem.tag_with_optional_text("url", self.url.clone());
+        #[cfg(feature = "rss_loose")]
+        elem.tag_with_optional_text("title", self.title.clone());
+        #[cfg(feature = "rss_loose")]
+        elem.tag_with_optional_text("link", self.link.clone());
+
         if let Some(ref n) = self.width {
             elem.tag_with_text("width", n.to_string());
         }
@@ -31,21 +54,31 @@ impl ViaXml for Image {
     }
 
     fn from_xml(elem: Element) -> Result<Self, ReadError> {
+        #[cfg(not(feature = "rss_loose"))]
         let url = match elem.get_child("url", None) {
             Some(elem) => elem.content_str(),
             None => return Err(ReadError::ImageMissingUrl),
         };
 
+        #[cfg(not(feature = "rss_loose"))]
         let title = match elem.get_child("title", None) {
             Some(elem) => elem.content_str(),
             None => return Err(ReadError::ImageMissingTitle),
         };
 
+        #[cfg(not(feature = "rss_loose"))]
         let link = match elem.get_child("link", None) {
             Some(elem) => elem.content_str(),
             None => return Err(ReadError::ImageMissingLink),
         };
 
+        #[cfg(feature = "rss_loose")]
+        let url = elem.get_child("url", None).map(Element::content_str);
+        #[cfg(feature = "rss_loose")]
+        let title = elem.get_child("title", None).map(Element::content_str);
+        #[cfg(feature = "rss_loose")]
+        let link = elem.get_child("link", None).map(Element::content_str);
+
         let height = match elem.get_child("height", None)
                                .map(|h| u32::from_str(&h.content_str()))
         {

src/lib.rs

@@ -19,6 +19,7 @@
 //! ## Writing
 //!
 //! ```
+//!# #![feature(stmt_expr_attributes)]
 //! use rss::{Channel, Item, Rss};
 //!
 //! let item = Item {
@@ -28,13 +29,23 @@
 //!     ..Default::default()
 //! };
 //!
+//!# #[cfg(not(feature = "rss_loose"))]
 //! let channel = Channel {
 //!     title: String::from("TechCrunch"),
 //!     link: String::from("http://techcrunch.com"),
 //!     description: String::from("The latest technology news and information on startups"),
 //!     items: vec![item],
 //!     ..Default::default()
 //! };
+//!#
+//!# #[cfg(feature = "rss_loose")]
+//!# let channel = Channel {
+//!#     title: Some(String::from("TechCrunch")),
+//!#     link: Some(String::from("http://techcrunch.com")),
+//!#     description: Some(String::from("The latest technology news and information on startups")),
+//!#     items: vec![item],
+//!#     ..Default::default()
+//!# };
 //!
 //! let rss = Rss(channel);
 //!
@@ -64,6 +75,22 @@
 //!
 //! let rss = rss_str.parse::<Rss>().unwrap();
 //! ```
+//!
+//! ### Partial Feeds
+//!
+//! In some cases, the RSS source may not return a standards-compliant RSS such as a missing description tag. The library
+//! is designed to return an error in such cases, however this behaviour can be loosened by using the feature
+//! flag `rss_loose`.
+//!
+//! Using this flag changes what would normally be a `String` type to a `Option<String>`, just like other fields.
+//!
+//! In your `Cargo.toml`, add the following:
+//! ```toml
+//! [dependencies.rss]
+//! features = ["rss_loose"]
+//! ```
+
+#![feature(stmt_expr_attributes)]
 
 mod category;
 mod guid;
@@ -246,6 +273,7 @@ mod test {
             ..Default::default()
         };
 
+        #[cfg(not(feature = "rss_loose"))]
         let channel = Channel {
             title: "My Blog".to_owned(),
             link: "http://myblog.com".to_owned(),
@@ -254,6 +282,15 @@ mod test {
             ..Default::default()
         };
 
+        #[cfg(feature = "rss_loose")]
+        let channel = Channel {
+            title: Some("My Blog".to_owned()),
+            link: Some("http://myblog.com".to_owned()),
+            description: Some("Where I write stuff".to_owned()),
+            items: vec![item],
+            ..Default::default()
+        };
+
         let rss = Rss(channel);
         assert_eq!(rss.to_string(), "<?xml version=\'1.0\' encoding=\'UTF-8\'?><rss version=\'2.0\'><channel><title>My Blog</title><link>http://myblog.com</link><description>Where I write stuff</description><item><title>My first post!</title><link>http://myblog.com/post1</link><description>This is my first post</description></item></channel></rss>");
     }
@@ -274,6 +311,7 @@ mod test {
     }
 
     #[test]
+    #[cfg_attr(feature = "rss_loose", ignore)]
     fn test_read_one_channel_no_properties() {
         let rss_str = "\
             <rss>\
@@ -294,7 +332,11 @@ mod test {
                 </channel>\
             </rss>";
         let Rss(channel) = Rss::from_str(rss_str).unwrap();
+
+        #[cfg(not(feature = "rss_loose"))]
         assert_eq!("Hello world!", channel.title);
+        #[cfg(feature = "rss_loose")]
+        assert_eq!(Some("Hello world!".to_owned()), channel.title); // How come &str is dereferenced but Some(&str) is not?
     }
 
     #[test]
@@ -340,7 +382,11 @@ mod test {
                 </channel>\
             </rss>";
         let Rss(channel) = Rss::from_str(rss_str).unwrap();
+
+        #[cfg(not(feature = "rss_loose"))]
         assert_eq!("Foobar", channel.text_input.unwrap().title);
+        #[cfg(feature = "rss_loose")]
+        assert_eq!(Some("Foobar".to_owned()), channel.text_input.unwrap().title);
     }
 
     // Ensure reader ignores the PI XML node and continues to parse the RSS
@@ -356,7 +402,11 @@ mod test {
                 </channel>\
             </rss>";
         let Rss(channel) = Rss::from_str(rss_str).unwrap();
+
+        #[cfg(not(feature = "rss_loose"))]
         assert_eq!("Title", channel.title);
+        #[cfg(feature = "rss_loose")]
+        assert_eq!(Some("Title".to_owned()), channel.title);
     }
 
     #[test]
@@ -379,14 +429,27 @@ mod test {
             </rss>";
         let rss = Rss::from_str(rss_str).unwrap();
         let image = rss.0.image.unwrap();
-        assert_eq!(image.url, "a url");
-        assert_eq!(image.title, "a title");
-        assert_eq!(image.link, "a link");
+
+        #[cfg(not(feature = "rss_loose"))]
+        assert_eq!(image.url, "a url".to_owned());
+        #[cfg(not(feature = "rss_loose"))]
+        assert_eq!(image.title, "a title".to_owned());
+        #[cfg(not(feature = "rss_loose"))]
+        assert_eq!(image.link, "a link".to_owned());
+
+        #[cfg(feature = "rss_loose")]
+        assert_eq!(image.url, Some("a url".to_owned()));
+        #[cfg(feature = "rss_loose")]
+        assert_eq!(image.title, Some("a title".to_owned()));
+        #[cfg(feature = "rss_loose")]
+        assert_eq!(image.link, Some("a link".to_owned()));
+
         assert_eq!(image.height, Some(140));
         assert_eq!(image.width, Some(280));
     }
 
     #[test]
+    #[cfg_attr(feature = "rss_loose", ignore)]
     fn test_read_image_no_url() {
         let rss_str = "\
             <?xml version=\'1.0\' encoding=\'UTF-8\'?>\
@@ -405,6 +468,7 @@ mod test {
     }
 
     #[test]
+    #[cfg_attr(feature = "rss_loose", ignore)]
     fn test_read_image_no_title() {
         let rss_str = "\
             <?xml version=\'1.0\' encoding=\'UTF-8\'?>\
@@ -423,6 +487,7 @@ mod test {
     }
 
     #[test]
+    #[cfg_attr(feature = "rss_loose", ignore)]
     fn test_read_image_no_link() {
         let rss_str = "\
             <?xml version=\'1.0\' encoding=\'UTF-8\'?>\