about summary refs log tree commit diff
path: root/tvix/eval/src/builtins/to_xml.rs
diff options
context:
space:
mode:
authorProfpatsch <mail@profpatsch.de>2024-05-20T13·50+0200
committerclbot <clbot@tvl.fyi>2024-05-22T10·32+0000
commit5b2ba0efa1b87aa514d665a7f64ada36617c720e (patch)
tree27a4bb08f20e997377cc1bc884ee16c38e203f86 /tvix/eval/src/builtins/to_xml.rs
parente7be3422566b36e5bd3aeaaf7d47537dfd050a5c (diff)
refactor(tvix/eval): rewrite xml emitter to be simple-stupid r/8160
In order to be compatible with the nix XML generator, it’s easier to
generate the XML directly, instead of going through a library which we
have to bend to do what we need.

Removes dependency on `xml-rs`, which came with a full XML parser that
we didn’t use. Only takes a tiny bit of code for the XML escaping,
somewhat simplified.

I add a little escaping value, to make sure we have the same behaviour
as nix proper.

Interestingly enough, we never need to escape XML attribute names,
because the `builtins.toXML` format encodes user-defined values as
attribute keys only. So we only escape attribute values.

Fixes: https://b.tvl.fyi/issues/399
Change-Id: If4d407d324864b3bb9aa3160e2ec6889f7727127
Reviewed-on: https://cl.tvl.fyi/c/depot/+/11697
Tested-by: BuildkiteCI
Reviewed-by: flokli <flokli@flokli.de>
Autosubmit: Profpatsch <mail@profpatsch.de>
Diffstat (limited to 'tvix/eval/src/builtins/to_xml.rs')
-rw-r--r--tvix/eval/src/builtins/to_xml.rs246
1 files changed, 196 insertions, 50 deletions
diff --git a/tvix/eval/src/builtins/to_xml.rs b/tvix/eval/src/builtins/to_xml.rs
index bb12cebfc9d0..10a6d8cdb358 100644
--- a/tvix/eval/src/builtins/to_xml.rs
+++ b/tvix/eval/src/builtins/to_xml.rs
@@ -3,59 +3,43 @@
 //! things in nixpkgs rely on.
 
 use bstr::ByteSlice;
+use std::borrow::Cow;
 use std::{io::Write, rc::Rc};
-use xml::writer::events::XmlEvent;
-use xml::writer::EmitterConfig;
-use xml::writer::EventWriter;
 
 use crate::{ErrorKind, Value};
 
 /// Recursively serialise a value to XML. The value *must* have been
 /// deep-forced before being passed to this function.
 pub fn value_to_xml<W: Write>(mut writer: W, value: &Value) -> Result<(), ErrorKind> {
-    let config = EmitterConfig {
-        perform_indent: true,
-        pad_self_closing: true,
-
-        // Nix uses single-quotes *only* in the document declaration,
-        // so we need to write it manually.
-        write_document_declaration: false,
-        ..Default::default()
-    };
-
     // Write a literal document declaration, using C++-Nix-style
     // single quotes.
     writeln!(writer, "<?xml version='1.0' encoding='utf-8'?>")?;
 
-    let mut writer = EventWriter::new_with_config(writer, config);
-
-    writer.write(XmlEvent::start_element("expr"))?;
-    value_variant_to_xml(&mut writer, value)?;
-    writer.write(XmlEvent::end_element())?;
+    let mut emitter = XmlEmitter::new(writer);
 
-    // Unwrap the writer to add the final newline that C++ Nix adds.
-    writeln!(writer.into_inner())?;
+    emitter.write_open_tag("expr", &[])?;
+    value_variant_to_xml(&mut emitter, value)?;
+    emitter.write_closing_tag("expr")?;
 
     Ok(())
 }
 
 fn write_typed_value<W: Write, V: ToString>(
-    w: &mut EventWriter<W>,
-    name: &str,
+    w: &mut XmlEmitter<W>,
+    name_unescaped: &str,
     value: V,
 ) -> Result<(), ErrorKind> {
-    w.write(XmlEvent::start_element(name).attr("value", &value.to_string()))?;
-    w.write(XmlEvent::end_element())?;
+    w.write_self_closing_tag(name_unescaped, &[("value", &value.to_string())])?;
     Ok(())
 }
 
-fn value_variant_to_xml<W: Write>(w: &mut EventWriter<W>, value: &Value) -> Result<(), ErrorKind> {
+fn value_variant_to_xml<W: Write>(w: &mut XmlEmitter<W>, value: &Value) -> Result<(), ErrorKind> {
     match value {
         Value::Thunk(t) => return value_variant_to_xml(w, &t.value()),
 
         Value::Null => {
-            w.write(XmlEvent::start_element("null"))?;
-            w.write(XmlEvent::end_element())
+            w.write_open_tag("null", &[])?;
+            w.write_closing_tag("null")?;
         }
 
         Value::Bool(b) => return write_typed_value(w, "bool", b),
@@ -65,50 +49,46 @@ fn value_variant_to_xml<W: Write>(w: &mut EventWriter<W>, value: &Value) -> Resu
         Value::Path(p) => return write_typed_value(w, "path", p.to_string_lossy()),
 
         Value::List(list) => {
-            w.write(XmlEvent::start_element("list"))?;
+            w.write_open_tag("list", &[])?;
 
             for elem in list.into_iter() {
                 value_variant_to_xml(w, elem)?;
             }
 
-            w.write(XmlEvent::end_element())
+            w.write_closing_tag("list")?;
         }
 
         Value::Attrs(attrs) => {
-            w.write(XmlEvent::start_element("attrs"))?;
+            w.write_open_tag("attrs", &[])?;
 
             for elem in attrs.iter() {
-                w.write(XmlEvent::start_element("attr").attr("name", &elem.0.to_str_lossy()))?;
+                w.write_open_tag("attr", &[("name", &elem.0.to_str_lossy())])?;
                 value_variant_to_xml(w, elem.1)?;
-                w.write(XmlEvent::end_element())?;
+                w.write_closing_tag("attr")?;
             }
 
-            w.write(XmlEvent::end_element())
+            w.write_closing_tag("attrs")?;
         }
 
         Value::Closure(c) => {
-            w.write(XmlEvent::start_element("function"))?;
+            w.write_open_tag("function", &[])?;
 
             match &c.lambda.formals {
                 Some(formals) => {
-                    let mut attrspat = XmlEvent::start_element("attrspat");
+                    let mut attrs: Vec<(&str, &str)> = Vec::with_capacity(2);
                     if formals.ellipsis {
-                        attrspat = attrspat.attr("ellipsis", "1");
+                        attrs.push(("ellipsis", "1"));
                     }
                     if let Some(ref name) = &formals.name {
-                        attrspat = attrspat.attr("name", name.as_str());
+                        attrs.push(("name", name.as_str()));
                     }
 
-                    w.write(attrspat)?;
-
+                    w.write_open_tag("attrspat", &attrs)?;
                     for arg in formals.arguments.iter() {
-                        w.write(
-                            XmlEvent::start_element("attr").attr("name", &arg.0.to_str_lossy()),
-                        )?;
-                        w.write(XmlEvent::end_element())?;
+                        w.write_self_closing_tag("attr", &[("name", &arg.0.to_str_lossy())])?;
                     }
 
-                    w.write(XmlEvent::end_element())?;
+                    w.write_closing_tag("attrspat")?;
                 }
                 None => {
                     // TODO(tazjin): tvix does not currently persist function
@@ -120,17 +100,16 @@ fn value_variant_to_xml<W: Write>(w: &mut EventWriter<W>, value: &Value) -> Resu
                     // If we don't want to persist the data, we can re-parse the
                     // AST from the spans of the lambda's bytecode and figure it
                     // out that way, but it needs some investigating.
-                    w.write(XmlEvent::start_element("varpat").attr("name", /* fake: */ "x"))?;
-                    w.write(XmlEvent::end_element())?;
+                    w.write_self_closing_tag("varpat", &[("name", /* fake: */ "x")])?;
                 }
             }
 
-            w.write(XmlEvent::end_element())
+            w.write_closing_tag("function")?;
         }
 
         Value::Builtin(_) => {
-            w.write(XmlEvent::start_element("unevaluated"))?;
-            w.write(XmlEvent::end_element())
+            w.write_open_tag("unevaluated", &[])?;
+            w.write_closing_tag("unevaluated")?;
         }
 
         Value::AttrNotFound
@@ -148,7 +127,174 @@ fn value_variant_to_xml<W: Write>(w: &mut EventWriter<W>, value: &Value) -> Resu
         Value::Catchable(_) => {
             panic!("tvix bug: value_to_xml() called on a value which had not been deep-forced")
         }
-    }?;
+    };
 
     Ok(())
 }
+
+/// A simple-stupid XML emitter, which implements only the subset needed for byte-by-byte compat with C++ nix’ `builtins.toXML`.
+struct XmlEmitter<W> {
+    /// The current indentation
+    cur_indent: usize,
+    writer: W,
+}
+
+impl<W: Write> XmlEmitter<W> {
+    pub fn new(writer: W) -> Self {
+        XmlEmitter {
+            cur_indent: 0,
+            writer,
+        }
+    }
+
+    /// Write an open tag with the given name (which is not escaped!)
+    /// and attributes (Keys are not escaped! Only attribute values are.)
+    pub fn write_open_tag(
+        &mut self,
+        name_unescaped: &str,
+        attrs: &[(&str, &str)],
+    ) -> std::io::Result<()> {
+        self.add_indent()?;
+        self.writer.write_all(b"<")?;
+        self.writer.write_all(name_unescaped.as_bytes())?;
+        self.write_attrs_escape_vals(attrs)?;
+        self.writer.write_all(b">\n")?;
+        self.cur_indent += 2;
+        Ok(())
+    }
+
+    /// Write a self-closing open tag with the given name (which is not escaped!)
+    /// and attributes (Keys are not escaped! Only attribute values are.)
+    pub fn write_self_closing_tag(
+        &mut self,
+        name_unescaped: &str,
+        attrs: &[(&str, &str)],
+    ) -> std::io::Result<()> {
+        self.add_indent()?;
+        self.writer.write_all(b"<")?;
+        self.writer.write_all(name_unescaped.as_bytes())?;
+        self.write_attrs_escape_vals(attrs)?;
+        self.writer.write_all(b" />\n")?;
+        Ok(())
+    }
+
+    /// Write a closing tag with the given name (which is not escaped!)
+    pub fn write_closing_tag(&mut self, name_unescaped: &str) -> std::io::Result<()> {
+        self.cur_indent -= 2;
+        self.add_indent()?;
+        self.writer.write_all(b"</")?;
+        self.writer.write_all(name_unescaped.as_bytes())?;
+        self.writer.write_all(b">\n")?;
+        Ok(())
+    }
+
+    #[inline]
+    fn add_indent(&mut self) -> std::io::Result<()> {
+        self.writer.write_all(&b" ".repeat(self.cur_indent))
+    }
+
+    /// Write an attribute list
+    fn write_attrs_escape_vals(&mut self, attrs: &[(&str, &str)]) -> std::io::Result<()> {
+        for (name, val) in attrs {
+            self.writer.write_all(b" ")?;
+            self.writer.write_all(name.as_bytes())?;
+            self.writer.write_all(br#"=""#)?;
+            self.writer
+                .write_all(Self::escape_attr_value(val).as_bytes())?;
+            self.writer.write_all(b"\"")?;
+        }
+        Ok(())
+    }
+
+    /// Escape the given attribute value, making sure we only actually clone the string if we needed to replace something.
+    fn escape_attr_value(s: &str) -> Cow<str> {
+        let mut last_escape: usize = 0;
+        let mut res: Cow<str> = Cow::Borrowed("");
+        // iterating via char_indices gives us the ability to index the original string slice at character boundaries
+        for (idx, c) in s.char_indices() {
+            match Self::should_escape_char(c) {
+                None => {}
+                Some(new) => {
+                    // add characters since the last escape we did
+                    res += &s[last_escape..idx];
+                    // add the escaped value
+                    res += new;
+                    last_escape = idx + 1;
+                }
+            }
+        }
+        // we did not need to escape anything, so borrow original string
+        if last_escape == 0 {
+            Cow::Borrowed(s)
+        } else {
+            // add the remaining characters
+            res += &s[last_escape..];
+            res
+        }
+    }
+
+    fn should_escape_char(c: char) -> Option<&'static str> {
+        match c {
+            '<' => Some("&lt;"),
+            '>' => Some("&gt;"),
+            '"' => Some("&quot;"),
+            '\'' => Some("&apos;"),
+            '&' => Some("&amp;"),
+            '\n' => Some("&#xA;"),
+            '\r' => Some("&#xD;"),
+            _ => None,
+        }
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use bytes::buf::Writer;
+    use pretty_assertions::assert_eq;
+
+    use crate::builtins::to_xml::XmlEmitter;
+    use std::borrow::Cow;
+
+    #[test]
+    fn xml_gen() {
+        let mut buf = Vec::new();
+        let mut x = XmlEmitter::new(&mut buf);
+        x.write_open_tag("hello", &[("hi", "it’s me"), ("no", "<escape>")])
+            .unwrap();
+        x.write_self_closing_tag("self-closing", &[("tag", "yay")])
+            .unwrap();
+        x.write_closing_tag("hello").unwrap();
+
+        assert_eq!(
+            std::str::from_utf8(&buf).unwrap(),
+            r##"<hello hi="it’s me" no="&lt;escape&gt;">
+  <self-closing tag="yay" />
+</hello>
+"##
+        );
+    }
+
+    #[test]
+    fn xml_escape() {
+        match XmlEmitter::<Writer<Vec<u8>>>::escape_attr_value("ab<>c&de") {
+            Cow::Owned(s) => assert_eq!(s, "ab&lt;&gt;c&amp;de".to_string(), "escape stuff"),
+            Cow::Borrowed(s) => panic!("s should be owned {}", s),
+        }
+        match XmlEmitter::<Writer<Vec<u8>>>::escape_attr_value("") {
+            Cow::Borrowed(s) => assert_eq!(s, "", "empty escape is borrowed"),
+            Cow::Owned(s) => panic!("s should be borrowed {}", s),
+        }
+        match XmlEmitter::<Writer<Vec<u8>>>::escape_attr_value("hi!ŷbla") {
+            Cow::Borrowed(s) => assert_eq!(s, "hi!ŷbla", "no escape is borrowed"),
+            Cow::Owned(s) => panic!("s should be borrowed {}", s),
+        }
+        match XmlEmitter::<Writer<Vec<u8>>>::escape_attr_value("hi!<ŷ>bla") {
+            Cow::Owned(s) => assert_eq!(
+                s,
+                "hi!&lt;ŷ&gt;bla".to_string(),
+                "multi-byte chars are correctly used"
+            ),
+            Cow::Borrowed(s) => panic!("s should be owned {}", s),
+        }
+    }
+}