Read previous Xrefs of linearized or incremental updated document

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lopdf"
-version = "0.5.0"
+version = "0.6.0"
 authors = ["Junfeng Liu <[email protected]>"]
 homepage = "https://github.com/J-F-Liu/lopdf"
 documentation = "https://docs.rs/crate/lopdf/"
@@ -10,7 +10,7 @@ readme = "README.md"
 description = "A Rust library for PDF document manipulation."
 
 [dependencies]
-pom = "0.9.0"
+pom = "1.0.0"
 flate2 = "0.2.14"
 linked-hash-map = "0.3.0"
 

src/object.rs

@@ -146,6 +146,14 @@ impl Object {
 			_ => None
 		}
 	}
+
+	pub fn type_name(&self) -> Option<&str> {
+		match *self {
+			Object::Dictionary(ref dict) => dict.type_name(),
+			Object::Stream(ref stream) => stream.dict.type_name(),
+			_ => None
+		}
+	}
 }
 
 impl Dictionary {
@@ -176,6 +184,10 @@ impl Dictionary {
 		self.0.remove(&key.into())
 	}
 
+	pub fn type_name(&self) -> Option<&str> {
+		self.0.get("Type").and_then(|obj|obj.as_name()).or(self.0.get("Linearized").and(Some("Linearized")))
+	}
+
 	pub fn type_is(&self, type_name: &str) -> bool {
 		self.0.get("Type").and_then(|obj|obj.as_name()) == Some(type_name)
 	}

src/reader.rs

@@ -43,9 +43,19 @@ impl Reader {
 		let xref_start = Self::get_xref_start(&self.buffer, &mut input)?;
 		input.jump_to(xref_start);
 
-		let (xref, trailer) = parser::xref_and_trailer(&self).parse(&mut input)
+		let (mut xref, mut trailer) = parser::xref_and_trailer(&self).parse(&mut input)
 			.map_err(|_|Error::new(ErrorKind::InvalidData, "Not a valid PDF file (xref_and_trailer)."))?;
 
+		// Read previous Xrefs of linearized or incremental updated document.
+		let mut prev_xref_start = trailer.remove("Prev");
+		while let Some(prev) = prev_xref_start.and_then(|offset|offset.as_i64()) {
+			input.jump_to(prev as usize);
+			let (prev_xref, mut prev_trailer) = parser::xref_and_trailer(&self).parse(&mut input)
+				.map_err(|_|Error::new(ErrorKind::InvalidData, "Not a valid PDF file (prev xref_and_trailer)."))?;
+			xref.extend(prev_xref);
+			prev_xref_start = prev_trailer.remove("Prev");
+		}
+
 		self.document.version = version;
 		self.document.max_id = xref.size - 1;
 		self.document.trailer = trailer;

src/writer.rs

@@ -14,7 +14,7 @@ impl Document {
 		file.write_all(format!("%PDF-{}\n", self.version).as_bytes())?;
 
 		for (&(id, generation), object) in &self.objects {
-			if object.as_dict().map(|dict| dict.type_is("ObjStm")) != Some(true) {
+			if object.type_name().map(|name| ["ObjStm", "XRef", "Linearized"].contains(&name)) != Some(true) {
 				Writer::write_indirect_object(&mut file, id, generation, object, &mut xref)?;
 			}
 		}

src/xref.rs

@@ -30,6 +30,10 @@ impl Xref {
 		self.entries.insert(id, entry);
 	}
 
+	pub fn extend(&mut self, xref: Xref) {
+		self.entries.extend(xref.entries);
+	}
+
 	pub fn clear(&mut self) {
 		self.entries.clear()
 	}
@@ -106,6 +110,8 @@ pub fn decode_xref_stream(mut stream: Stream) -> (Xref, Dictionary) {
 		}
 	}
 	dict.remove("Length");
+	dict.remove("W");
+	dict.remove("Index");
 	(xref, dict)
 }