From 2137e8474365b1346ac436422a4804e872e80d37 Mon Sep 17 00:00:00 2001
From: Jan Lukas Gernert <jangernert@gmail.com>
Date: Thu, 26 Sep 2019 21:28:05 +0200
Subject: [PATCH] update to new serialization api of libxml

---
 Cargo.toml        |  2 +-
 src/images/mod.rs | 17 +++++++++++++++--
 src/lib.rs        | 15 +++++++++++++--
 3 files changed, 29 insertions(+), 5 deletions(-)

diff --git a/Cargo.toml b/Cargo.toml
index 506acbc..f732ba8 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -6,7 +6,7 @@ edition = "2018"
 
 [dependencies]
 failure = "0.1"
-libxml = "0.2"
+libxml = { git = "https://github.com/KWARC/rust-libxml.git" }
 reqwest = "0.9"
 url = "1.7"
 regex = "1.3"
diff --git a/src/images/mod.rs b/src/images/mod.rs
index 427054e..efe6e75 100644
--- a/src/images/mod.rs
+++ b/src/images/mod.rs
@@ -5,7 +5,10 @@ use log::{
 };
 use libxml::parser::Parser;
 use libxml::xpath::Context;
-use libxml::tree::Node;
+use libxml::tree::{
+    Node,
+    SaveOptions,
+};
 use url;
 use failure::ResultExt;
 use std::error::Error;
@@ -45,7 +48,17 @@ impl ImageDownloader {
 
         self.download_images_from_context(&xpath_ctx)?;
 
-        Ok(doc.to_string(/*format:*/ false))
+        let options = SaveOptions {
+            format: false,
+            no_declaration: false,
+            no_empty_tags: true,
+            no_xhtml: false,
+            xhtml: false,
+            as_xml: false,
+            as_html: true,
+            non_significant_whitespace: false,
+        };
+        Ok(doc.to_string_with_options(options))
     }
 
     pub fn download_images_from_context(&self, context: &Context) -> Result<(), ImageDownloadError> {
diff --git a/src/lib.rs b/src/lib.rs
index 20597ef..963a69e 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -19,7 +19,8 @@ use libxml::parser::Parser;
 use libxml::xpath::Context;
 use libxml::tree::{
     Document,
-    Node
+    Node,
+    SaveOptions,
 };
 use std::path::PathBuf;
 use std::ops::Index;
@@ -128,7 +129,17 @@ impl ArticleScraper {
         }
 
         // serialize content
-        let html = document.to_string(/*format:*/ false);
+        let options = SaveOptions {
+            format: false,
+            no_declaration: false,
+            no_empty_tags: true,
+            no_xhtml: false,
+            xhtml: false,
+            as_xml: false,
+            as_html: true,
+            non_significant_whitespace: false,
+        };
+        let html = document.to_string_with_options(options);
         article.html = Some(html);
         
         Ok(article)