Implement streaming writers (#765)

* Make XMLBuilder generic * Reduce allocations at XmlData display impl * Implement streaming writers - Extend BuildXML trait, add the streaming method - Remove impls for Box<Ty>, as they can be implemented on the trait level - Rewrite build() methods in chaining style, backed by apply_* helpers - Remove quite a few allocations, though numbers still produce them - Add spaces between children nodes, fix tests * Add rustfmt.toml and format code * Fix clippy warnings * Expose the BuildXML trait without displaying its methods in hints
bokuweb · Nov 5, 2024 · 238d2bc · 238d2bc
1 parent 72637a4
commit 238d2bc
Show file tree

Hide file tree

Showing 186 changed files with 3,502 additions and 4,407 deletions.
diff --git a/docx-core/src/documents/build_xml.rs b/docx-core/src/documents/build_xml.rs
@@ -1,3 +1,38 @@
+use crate::xml_builder::XMLBuilder;
+use std::io::Write;
+
 pub trait BuildXML {
-    fn build(&self) -> Vec<u8>;
+    /// Write XML to the output stream.
+    #[doc(hidden)]
+    fn build_to<W: Write>(
+        &self,
+        stream: xml::writer::EventWriter<W>,
+    ) -> xml::writer::Result<xml::writer::EventWriter<W>>;
+
+    #[doc(hidden)]
+    fn build(&self) -> Vec<u8> {
+        self.build_to(XMLBuilder::new(Vec::new()).into_inner().unwrap())
+            .expect("should write to buf")
+            .into_inner()
+    }
+}
+
+impl<'a, T: BuildXML> BuildXML for &'a T {
+    /// Building XML from `&T` is the same as from `T`.
+    fn build_to<W: Write>(
+        &self,
+        stream: xml::writer::EventWriter<W>,
+    ) -> xml::writer::Result<xml::writer::EventWriter<W>> {
+        (*self).build_to(stream)
+    }
+}
+
+impl<T: BuildXML> BuildXML for Box<T> {
+    /// Building XML from `Box<T>` is the same as from `T`.
+    fn build_to<W: Write>(
+        &self,
+        stream: xml::writer::EventWriter<W>,
+    ) -> xml::writer::Result<xml::writer::EventWriter<W>> {
+        (**self).build_to(stream)
+    }
 }
diff --git a/docx-core/src/documents/comments.rs b/docx-core/src/documents/comments.rs
@@ -1,6 +1,7 @@
 use super::Comment;
 use crate::documents::BuildXML;
 use crate::xml_builder::*;
+use std::io::Write;
 
 use serde::Serialize;
 
@@ -29,12 +30,16 @@ impl Comments {
 }
 
 impl BuildXML for Comments {
-    fn build(&self) -> Vec<u8> {
-        let mut b = XMLBuilder::new().declaration(Some(true)).open_comments();
-        for c in &self.comments {
-            b = b.add_child(c)
-        }
-        b.close().build()
+    fn build_to<W: Write>(
+        &self,
+        stream: xml::writer::EventWriter<W>,
+    ) -> xml::writer::Result<xml::writer::EventWriter<W>> {
+        XMLBuilder::from(stream)
+            .declaration(Some(true))?
+            .open_comments()?
+            .add_children(&self.comments)?
+            .close()?
+            .into_inner()
     }
 }
 
@@ -51,8 +56,7 @@ mod tests {
         let b = Comments::new().build();
         assert_eq!(
             str::from_utf8(&b).unwrap(),
-            r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
-<w:comments xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships" xmlns:v="urn:schemas-microsoft-com:vml" xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:w10="urn:schemas-microsoft-com:office:word" xmlns:wp="http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing" xmlns:wps="http://schemas.microsoft.com/office/word/2010/wordprocessingShape" xmlns:wpg="http://schemas.microsoft.com/office/word/2010/wordprocessingGroup" xmlns:mc="http://schemas.openxmlformats.org/markup-compatibility/2006" xmlns:wp14="http://schemas.microsoft.com/office/word/2010/wordprocessingDrawing" xmlns:w14="http://schemas.microsoft.com/office/word/2010/wordml" mc:Ignorable="w14 wp14" />"#
+            r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?><w:comments xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships" xmlns:v="urn:schemas-microsoft-com:vml" xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:w10="urn:schemas-microsoft-com:office:word" xmlns:wp="http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing" xmlns:wps="http://schemas.microsoft.com/office/word/2010/wordprocessingShape" xmlns:wpg="http://schemas.microsoft.com/office/word/2010/wordprocessingGroup" xmlns:mc="http://schemas.openxmlformats.org/markup-compatibility/2006" xmlns:wp14="http://schemas.microsoft.com/office/word/2010/wordprocessingDrawing" xmlns:w14="http://schemas.microsoft.com/office/word/2010/wordml" mc:Ignorable="w14 wp14" />"#
         );
     }
 }
diff --git a/docx-core/src/documents/comments_extended.rs b/docx-core/src/documents/comments_extended.rs
@@ -1,4 +1,5 @@
 use serde::Serialize;
+use std::io::Write;
 
 use super::*;
 use crate::documents::BuildXML;
@@ -22,14 +23,15 @@ impl CommentsExtended {
 }
 
 impl BuildXML for CommentsExtended {
-    fn build(&self) -> Vec<u8> {
-        let mut b = XMLBuilder::new();
-        b = b.open_comments_extended();
-
-        for c in &self.children {
-            b = b.add_child(c)
-        }
-        b.close().build()
+    fn build_to<W: Write>(
+        &self,
+        stream: xml::writer::EventWriter<W>,
+    ) -> xml::writer::Result<xml::writer::EventWriter<W>> {
+        XMLBuilder::from(stream)
+            .open_comments_extended()?
+            .add_children(&self.children)?
+            .close()?
+            .into_inner()
     }
 }
 

diff --git a/docx-core/src/documents/content_types.rs b/docx-core/src/documents/content_types.rs
@@ -1,6 +1,6 @@
 use serde::{Deserialize, Serialize};
 use std::collections::BTreeMap;
-use std::io::Read;
+use std::io::{Read, Write};
 use xml::reader::{EventReader, XmlEvent};
 
 use crate::documents::BuildXML;
@@ -153,28 +153,26 @@ impl Default for ContentTypes {
 }
 
 impl BuildXML for ContentTypes {
-    fn build(&self) -> Vec<u8> {
-        let b = XMLBuilder::new();
-        let mut b = b
-            .declaration(None)
-            .open_types("http://schemas.openxmlformats.org/package/2006/content-types");
-
-        b = b
-            .add_default("png", "image/png")
-            .add_default("jpeg", "image/jpeg")
-            .add_default("jpg", "image/jpg")
-            .add_default("bmp", "image/bmp")
-            .add_default("gif", "image/gif")
+    fn build_to<W: Write>(
+        &self,
+        stream: xml::writer::EventWriter<W>,
+    ) -> xml::writer::Result<xml::writer::EventWriter<W>> {
+        XMLBuilder::from(stream)
+            .declaration(None)?
+            .open_types("http://schemas.openxmlformats.org/package/2006/content-types")?
+            .add_default("png", "image/png")?
+            .add_default("jpeg", "image/jpeg")?
+            .add_default("jpg", "image/jpg")?
+            .add_default("bmp", "image/bmp")?
+            .add_default("gif", "image/gif")?
             .add_default(
                 "rels",
                 "application/vnd.openxmlformats-package.relationships+xml",
-            )
-            .add_default("xml", "application/xml");
-
-        for (k, v) in self.types.iter() {
-            b = b.add_override(k, v);
-        }
-        b.close().build()
+            )?
+            .add_default("xml", "application/xml")?
+            .apply_each(self.types.iter(), |(k, v), b| b.add_override(k, v))?
+            .close()?
+            .into_inner()
     }
 }
 
@@ -213,8 +211,7 @@ mod tests {
 
     #[test]
     fn test_from_xml() {
-        let xml = r#"<Types xmlns="http://schemas.openxmlformats.org/package/2006/content-types">
-        <Override ContentType="application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml" PartName="/word/document.xml"></Override></Types>"#;
+        let xml = r#"<Types xmlns="http://schemas.openxmlformats.org/package/2006/content-types"><Override ContentType="application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml" PartName="/word/document.xml"></Override></Types>"#;
         let c = ContentTypes::from_xml(xml.as_bytes()).unwrap();
         let mut types = BTreeMap::new();
         types.insert(

diff --git a/docx-core/src/documents/custom_item.rs b/docx-core/src/documents/custom_item.rs
@@ -1,7 +1,9 @@
 use crate::documents::BuildXML;
+use crate::xml_builder::XMLBuilder;
 use crate::{ParseXmlError, XmlDocument};
 use serde::ser::SerializeSeq;
 use serde::Serialize;
+use std::io::Write;
 use std::str::FromStr;
 
 #[derive(Debug, Clone)]
@@ -29,8 +31,13 @@ impl Serialize for CustomItem {
 }
 
 impl BuildXML for CustomItem {
-    fn build(&self) -> Vec<u8> {
-        self.0.to_string().as_bytes().to_vec()
+    fn build_to<W: Write>(
+        &self,
+        stream: xml::writer::EventWriter<W>,
+    ) -> xml::writer::Result<xml::writer::EventWriter<W>> {
+        let mut b = XMLBuilder::from(stream);
+        write!(b.inner_mut()?, "{}", self.0)?;
+        b.into_inner()
     }
 }
 
@@ -44,17 +51,13 @@ mod tests {
     #[test]
     fn test_custom_xml() {
         let c = CustomItem::from_str(
-            r#"<ds:datastoreItem ds:itemID="{06AC5857-5C65-A94A-BCEC-37356A209BC3}" xmlns:ds="http://schemas.openxmlformats.org/officeDocument/2006/customXml">
-<ds:schemaRefs>
-<ds:schemaRef ds:uri="https://hoge.com"></ds:schemaRef></ds:schemaRefs></ds:datastoreItem>"#,
+            r#"<ds:datastoreItem ds:itemID="{06AC5857-5C65-A94A-BCEC-37356A209BC3}" xmlns:ds="http://schemas.openxmlformats.org/officeDocument/2006/customXml"><ds:schemaRefs><ds:schemaRef ds:uri="https://hoge.com"></ds:schemaRef></ds:schemaRefs></ds:datastoreItem>"#,
         )
         .unwrap();
 
         assert_eq!(
             c.0.to_string(),
-            r#"<ds:datastoreItem ds:itemID="{06AC5857-5C65-A94A-BCEC-37356A209BC3}" xmlns:ds="http://schemas.openxmlformats.org/officeDocument/2006/customXml">
-<ds:schemaRefs>
-<ds:schemaRef ds:uri="https://hoge.com"></ds:schemaRef></ds:schemaRefs></ds:datastoreItem>"#
+            r#"<ds:datastoreItem ds:itemID="{06AC5857-5C65-A94A-BCEC-37356A209BC3}" xmlns:ds="http://schemas.openxmlformats.org/officeDocument/2006/customXml"><ds:schemaRefs><ds:schemaRef ds:uri="https://hoge.com"></ds:schemaRef></ds:schemaRefs></ds:datastoreItem>"#
         );
         assert_eq!(
             serde_json::to_string(&c).unwrap(),

diff --git a/docx-core/src/documents/custom_item_property.rs b/docx-core/src/documents/custom_item_property.rs
@@ -1,4 +1,5 @@
 use serde::Serialize;
+use std::io::Write;
 
 use crate::documents::BuildXML;
 use crate::xml_builder::*;
@@ -15,16 +16,19 @@ impl CustomItemProperty {
 }
 
 impl BuildXML for CustomItemProperty {
-    fn build(&self) -> Vec<u8> {
-        let mut b = XMLBuilder::new();
-        b = b.declaration(Some(false));
-        b = b
+    fn build_to<W: Write>(
+        &self,
+        stream: xml::writer::EventWriter<W>,
+    ) -> xml::writer::Result<xml::writer::EventWriter<W>> {
+        XMLBuilder::from(stream)
+            .declaration(Some(false))?
             .open_data_store_item(
                 "http://schemas.openxmlformats.org/officeDocument/2006/customXml",
                 &format!("{{{}}}", self.id),
-            )
-            .open_data_store_schema_refs()
-            .close();
-        b.close().build()
+            )?
+            .open_data_store_schema_refs()?
+            .close()?
+            .close()?
+            .into_inner()
     }
 }
diff --git a/docx-core/src/documents/custom_item_rels.rs b/docx-core/src/documents/custom_item_rels.rs
@@ -1,4 +1,5 @@
 use serde::Serialize;
+use std::io::Write;
 
 use crate::documents::BuildXML;
 use crate::xml_builder::*;
@@ -21,21 +22,21 @@ impl CustomItemRels {
 }
 
 impl BuildXML for CustomItemRels {
-    fn build(&self) -> Vec<u8> {
-        let mut b = XMLBuilder::new();
-        b = b
-            .declaration(Some(true))
-            .open_relationships("http://schemas.openxmlformats.org/package/2006/relationships");
-
-        for id in 0..self.custom_item_count {
-            let id = id + 1;
-            b = b.relationship(
-                &format!("rId{}", id),
-                "http://schemas.openxmlformats.org/officeDocument/2006/relationships/customXmlProps",
-                &format!("itemProps{}.xml", id),
-            )
-        }
-
-        b.close().build()
+    fn build_to<W: Write>(
+        &self,
+        stream: xml::writer::EventWriter<W>,
+    ) -> xml::writer::Result<xml::writer::EventWriter<W>> {
+        XMLBuilder::from(stream)
+            .declaration(Some(true))?
+            .open_relationships("http://schemas.openxmlformats.org/package/2006/relationships")?
+            .apply_each(0..self.custom_item_count, |id, b| {
+                b.relationship(
+          &format!("rId{}", id + 1),
+          "http://schemas.openxmlformats.org/officeDocument/2006/relationships/customXmlProps",
+          &format!("itemProps{}.xml", id + 1),
+        )
+            })?
+            .close()?
+            .into_inner()
     }
 }
diff --git a/docx-core/src/documents/doc_props/app.rs b/docx-core/src/documents/doc_props/app.rs
@@ -1,4 +1,5 @@
 use serde::Serialize;
+use std::io::Write;
 
 use crate::documents::BuildXML;
 use crate::xml_builder::*;
@@ -14,13 +15,18 @@ impl AppProps {
 }
 
 impl BuildXML for AppProps {
-    fn build(&self) -> Vec<u8> {
-        let b = XMLBuilder::new();
-        let base = b.declaration(Some(true)).open_properties(
-            "http://schemas.openxmlformats.org/officeDocument/2006/extended-properties",
-            "http://schemas.openxmlformats.org/officeDocument/2006/docPropsVTypes",
-        );
-        base.close().build()
+    fn build_to<W: Write>(
+        &self,
+        stream: xml::writer::EventWriter<W>,
+    ) -> xml::writer::Result<xml::writer::EventWriter<W>> {
+        XMLBuilder::from(stream)
+            .declaration(Some(true))?
+            .open_properties(
+                "http://schemas.openxmlformats.org/officeDocument/2006/extended-properties",
+                "http://schemas.openxmlformats.org/officeDocument/2006/docPropsVTypes",
+            )?
+            .close()?
+            .into_inner()
     }
 }
 
@@ -38,8 +44,7 @@ mod tests {
         let b = c.build();
         assert_eq!(
             str::from_utf8(&b).unwrap(),
-            r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
-<Properties xmlns="http://schemas.openxmlformats.org/officeDocument/2006/extended-properties" xmlns:vt="http://schemas.openxmlformats.org/officeDocument/2006/docPropsVTypes" />"#
+            r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?><Properties xmlns="http://schemas.openxmlformats.org/officeDocument/2006/extended-properties" xmlns:vt="http://schemas.openxmlformats.org/officeDocument/2006/docPropsVTypes" />"#
         );
     }
 }