feat: support for doctype retrieval (#26)

src/reader.rs

@@ -95,6 +95,11 @@ impl<R: Read> EventReader<R> {
     pub fn into_inner(self) -> R {
         self.source
     }
+
+    /// Retrieves the DOCTYPE from the document if any
+    pub fn doctype(&self) -> Option<&String> {
+        self.parser.doctype()
+    }
 }
 
 impl<B: Read> Position for EventReader<B> {

src/reader/parser.rs

@@ -122,6 +122,7 @@ impl PullParser {
 
             data: MarkupData {
                 name: String::new(),
+                doctype: None,
                 version: None,
                 encoding: None,
                 standalone: None,
@@ -146,6 +147,12 @@ impl PullParser {
     /// Checks if this parser ignores the end of stream errors.
     pub fn is_ignoring_end_of_stream(&self) -> bool { self.config.c.ignore_end_of_stream }
 
+    /// Retrieves the Doctype from the document if any
+    #[inline]
+    pub fn doctype(&self) -> Option<&String> {
+        self.data.doctype.as_ref()
+    }
+
     #[inline(never)]
     fn set_encountered(&mut self, new_encounter: Encountered) -> Option<Result> {
         if new_encounter <= self.encountered {
@@ -299,6 +306,7 @@ struct MarkupData {
     name: String,     // used for processing instruction name
     ref_data: String,  // used for reference content
 
+    doctype: Option<String>, // keeps a copy of the original doctype
     version: Option<XmlVersion>,  // used for XML declaration version
     encoding: Option<String>,  // used for XML declaration encoding
     standalone: Option<bool>,  // used for XML declaration standalone parameter

src/reader/parser/inside_doctype.rs

@@ -1,3 +1,5 @@
+use std::fmt::Write;
+
 use crate::common::{is_name_char, is_name_start_char, is_whitespace_char};
 use crate::reader::error::SyntaxError;
 use crate::reader::lexer::Token;
@@ -6,6 +8,10 @@ use super::{DoctypeSubstate, PullParser, QuoteToken, Result, State};
 
 impl PullParser {
     pub fn inside_doctype(&mut self, t: Token, substate: DoctypeSubstate) -> Option<Result> {
+        if let Some(ref mut doctype) = self.data.doctype {
+            write!(doctype, "{t}").ok()?;
+        }
+
         match substate {
             DoctypeSubstate::Outside => match t {
                 Token::TagEnd => self.into_state_continue(State::OutsideTag),

tests/event_reader.rs

@@ -714,6 +714,26 @@ fn push_pos_issue() {
     parser.into_iter().for_each(|e| { e.unwrap(); });
 }
 
+#[test]
+fn retrieve_doctype() {
+    let source = r#"<?xml version="1.0" encoding="UTF-8"?>
+    <!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"
+      "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
+    <svg xmlns="http://www.w3.org/2000/svg" version="1.1"
+         width="120" height="120">
+      <rect x="14" y="23" width="200" height="50" fill="lime"
+          stroke="black" />
+    </svg>"#;
+
+    let parser = ParserConfig::new()
+        .cdata_to_characters(true)
+        .ignore_comments(true)
+        .coalesce_characters(false)
+        .create_reader(std::io::Cursor::new(source));
+
+    assert_eq!(parser.doctype(), Some(&String::from(r#"<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN" "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">"#)));
+}
+
 // clones a lot but that's fine
 fn trim_until_bar(s: String) -> String {
     match s.trim() {