Merge pull request #107 from uklotzde/parse_lazy_url

Parse homepage URLs of entries lazily

src/core/usecases/create_new_entry.rs

@@ -1,5 +1,6 @@
 use chrono::*;
 use core::prelude::*;
+use core::util::parse::parse_url_param;
 use core::util::validate::Validate;
 use uuid::Uuid;
 
@@ -44,12 +45,12 @@ pub fn create_new_entry<D: Db>(db: &mut D, e: NewEntry) -> Result<String> {
         country     :  e.country,
         email       :  e.email,
         telephone   :  e.telephone,
-        homepage    :  e.homepage,
+        homepage    :  e.homepage.map(|ref url| parse_url_param(url)).transpose()?,
         categories  :  e.categories,
         tags,
         license     :  Some(e.license),
-        image_url     : e.image_url,
-        image_link_url: e.image_link_url,
+        image_url     : e.image_url.map(|ref url| parse_url_param(url)).transpose()?,
+        image_link_url: e.image_link_url.map(|ref url| parse_url_param(url)).transpose()?,
     };
     debug!("Creating new entry: {:?}", new_entry);
     new_entry.validate()?;

src/core/usecases/update_entry.rs

@@ -1,5 +1,6 @@
 use chrono::*;
 use core::prelude::*;
+use core::util::parse::parse_url_param;
 
 #[cfg_attr(rustfmt, rustfmt_skip)]
 #[derive(Serialize, Deserialize, Debug, Clone)]
@@ -47,12 +48,12 @@ pub fn update_entry<D: Db>(db: &mut D, e: UpdateEntry) -> Result<()> {
         country     :  e.country,
         email       :  e.email,
         telephone   :  e.telephone,
-        homepage    :  e.homepage,
+        homepage    :  e.homepage.map(|ref url| parse_url_param(url)).transpose()?,
         categories  :  e.categories,
         tags,
         license     :  old.license, // license is immutable
-        image_url     : e.image_url,
-        image_link_url: e.image_link_url,
+        image_url     : e.image_url.map(|ref url| parse_url_param(url)).transpose()?,
+        image_link_url: e.image_link_url.map(|ref url| parse_url_param(url)).transpose()?,
     };
     debug!("Updating existing entry: {:?}", updated_entry);
     for t in &updated_entry.tags {
@@ -77,8 +78,8 @@ mod tests {
             .version(1)
             .title("foo")
             .description("bar")
-            .image_url(Some("img"))
-            .image_link_url(Some("img link"))
+            .image_url(Some("http://img"))
+            .image_link_url(Some("http://imglink"))
             .finish();
 
         #[cfg_attr(rustfmt, rustfmt_skip)]
@@ -113,8 +114,8 @@ mod tests {
         assert_eq!(2, x.version);
         assert!(x.created as i64 >= now.timestamp());
         assert!(Uuid::parse_str(&x.id).is_ok());
-        assert_eq!("img2", x.image_url.as_ref().unwrap());
-        assert_eq!("img link", x.image_link_url.as_ref().unwrap());
+        assert_eq!("https://www.img2/", x.image_url.as_ref().unwrap());
+        assert_eq!("http://imglink/", x.image_link_url.as_ref().unwrap());
     }
 
     #[test]

src/core/util/mod.rs

@@ -1,4 +1,5 @@
 pub mod filter;
 pub mod geo;
+pub mod parse;
 pub mod sort;
 pub mod validate;

src/core/util/parse.rs

@@ -0,0 +1,50 @@
+use core::error::ParameterError;
+use url::{ParseError, Url};
+
+/// Completes incomplete URLs before parsing
+pub fn parse_lazy_url(url: &str) -> Result<Url, ParseError> {
+    let url = url.trim();
+    if url.is_empty() || url.contains("://") {
+        Url::parse(url)
+    } else {
+        // Add the missing protocol by assuming https
+        if url.starts_with("www.") {
+            Url::parse(&format!("https://{}", url))
+        } else {
+            Url::parse(&format!("https://www.{}", url))
+        }
+    }
+}
+
+pub fn parse_url_param(url: &str) -> Result<String, ParameterError> {
+    parse_lazy_url(url)
+        .map(|url| url.into_string())
+        .map_err(|_| ParameterError::Url)
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn parse_url_param() {
+        assert!(super::parse_url_param("").is_err());
+        assert!(super::parse_url_param("\t \n").is_err());
+        assert_eq!(
+            super::parse_url_param("example.com/index.html").unwrap(),
+            "https://www.example.com/index.html"
+        );
+        assert_eq!(
+            super::parse_url_param("www.example.com").unwrap(),
+            "https://www.example.com/"
+        );
+        assert_eq!(
+            super::parse_url_param("http://www.example.com/index.html").unwrap(),
+            "http://www.example.com/index.html"
+        );
+        assert_eq!(
+            super::parse_url_param("https://example.com").unwrap(),
+            "https://example.com/"
+        );
+    }
+}

src/main.rs

@@ -1,6 +1,6 @@
 // Copyright (c) 2015 - 2018 Markus Kohlhase <mail@markus-kohlhase.de>
 
-#![feature(plugin, custom_derive, test)]
+#![feature(plugin, custom_derive, test, transpose_result)]
 #![plugin(rocket_codegen)]
 #![recursion_limit = "256"]