Rollup merge of rust-lang#65074 - Rantanen:json-byte-pos, r=matklad

Fix the start/end byte positions in the compiler JSON output

Track the changes made during normalization in the `SourceFile` and use this information to correct the `start_byte` and `end_byte` fields in the JSON output.

This should ensure the start/end byte fields can be used to index the original file, even if Rust normalized the source code for parsing purposes. Both CRLF to LF and BOM removal are handled with this one.

The rough plan was discussed with @matklad in rust-lang/rustfix#176 - although I ended up going with `u32` offset tracking so I wouldn't need to deal with `u32 + i32` arithmetics when applying the offset to the span byte positions.

Fixes rust-lang#65029

Author: Centril

src/librustc/ich/impls_syntax.rs

@@ -424,6 +424,7 @@ impl<'a> HashStable<StableHashingContext<'a>> for SourceFile {
             ref lines,
             ref multibyte_chars,
             ref non_narrow_chars,
+            ref normalized_pos,
         } = *self;
 
         (name_hash as u64).hash_stable(hcx, hasher);
@@ -452,6 +453,12 @@ impl<'a> HashStable<StableHashingContext<'a>> for SourceFile {
         for &char_pos in non_narrow_chars.iter() {
             stable_non_narrow_char(char_pos, start_pos).hash_stable(hcx, hasher);
         }
+
+        normalized_pos.len().hash_stable(hcx, hasher);
+        for &char_pos in normalized_pos.iter() {
+            stable_normalized_pos(char_pos, start_pos).hash_stable(hcx, hasher);
+        }
+
     }
 }
 
@@ -481,6 +488,18 @@ fn stable_non_narrow_char(swc: ::syntax_pos::NonNarrowChar,
     (pos.0 - source_file_start.0, width as u32)
 }
 
+fn stable_normalized_pos(np: ::syntax_pos::NormalizedPos,
+                         source_file_start: ::syntax_pos::BytePos)
+                         -> (u32, u32) {
+    let ::syntax_pos::NormalizedPos {
+        pos,
+        diff
+    } = np;
+
+    (pos.0 - source_file_start.0, diff)
+}
+
+
 impl<'tcx> HashStable<StableHashingContext<'tcx>> for feature_gate::Features {
     fn hash_stable(&self, hcx: &mut StableHashingContext<'tcx>, hasher: &mut StableHasher) {
         // Unfortunately we cannot exhaustively list fields here, since the

src/librustc_metadata/decoder.rs

@@ -1317,6 +1317,7 @@ impl<'a, 'tcx> CrateMetadata {
                                       mut lines,
                                       mut multibyte_chars,
                                       mut non_narrow_chars,
+                                      mut normalized_pos,
                                       name_hash,
                                       .. } = source_file_to_import;
 
@@ -1336,6 +1337,9 @@ impl<'a, 'tcx> CrateMetadata {
             for swc in &mut non_narrow_chars {
                 *swc = *swc - start_pos;
             }
+            for np in &mut normalized_pos {
+                np.pos = np.pos - start_pos;
+            }
 
             let local_version = local_source_map.new_imported_source_file(name,
                                                                    name_was_remapped,
@@ -1345,7 +1349,8 @@ impl<'a, 'tcx> CrateMetadata {
                                                                    source_length,
                                                                    lines,
                                                                    multibyte_chars,
-                                                                   non_narrow_chars);
+                                                                   non_narrow_chars,
+                                                                   normalized_pos);
             debug!("CrateMetaData::imported_source_files alloc \
                     source_file {:?} original (start_pos {:?} end_pos {:?}) \
                     translated (start_pos {:?} end_pos {:?})",

src/libsyntax/json.rs

@@ -25,6 +25,9 @@ use std::sync::{Arc, Mutex};
 
 use rustc_serialize::json::{as_json, as_pretty_json};
 
+#[cfg(test)]
+mod tests;
+
 pub struct JsonEmitter {
     dst: Box<dyn Write + Send>,
     registry: Option<Registry>,
@@ -332,8 +335,8 @@ impl DiagnosticSpan {
 
         DiagnosticSpan {
             file_name: start.file.name.to_string(),
-            byte_start: span.lo().0 - start.file.start_pos.0,
-            byte_end: span.hi().0 - start.file.start_pos.0,
+            byte_start: start.file.original_relative_byte_pos(span.lo()).0,
+            byte_end: start.file.original_relative_byte_pos(span.hi()).0,
             line_start: start.line,
             line_end: end.line,
             column_start: start.col.0 + 1,

src/libsyntax/json/tests.rs

@@ -0,0 +1,186 @@
+use super::*;
+
+use crate::json::JsonEmitter;
+use crate::source_map::{FilePathMapping, SourceMap};
+use crate::tests::Shared;
+use crate::with_default_globals;
+
+use errors::emitter::{ColorConfig, HumanReadableErrorType};
+use errors::Handler;
+use rustc_serialize::json::decode;
+use syntax_pos::{BytePos, Span};
+
+use std::str;
+
+#[derive(RustcDecodable, Debug, PartialEq, Eq)]
+struct TestData {
+    spans: Vec<SpanTestData>,
+}
+
+#[derive(RustcDecodable, Debug, PartialEq, Eq)]
+struct SpanTestData {
+    pub byte_start: u32,
+    pub byte_end: u32,
+    pub line_start: u32,
+    pub column_start: u32,
+    pub line_end: u32,
+    pub column_end: u32,
+}
+
+/// Test the span yields correct positions in JSON.
+fn test_positions(code: &str, span: (u32, u32), expected_output: SpanTestData) {
+    let expected_output = TestData { spans: vec![expected_output] };
+
+    with_default_globals(|| {
+        let sm = Lrc::new(SourceMap::new(FilePathMapping::empty()));
+        sm.new_source_file(Path::new("test.rs").to_owned().into(), code.to_owned());
+
+        let output = Arc::new(Mutex::new(Vec::new()));
+        let je = JsonEmitter::new(
+            Box::new(Shared { data: output.clone() }),
+            None,
+            sm,
+            true,
+            HumanReadableErrorType::Short(ColorConfig::Never),
+            false,
+        );
+
+        let span = Span::with_root_ctxt(BytePos(span.0), BytePos(span.1));
+        let handler = Handler::with_emitter(true, None, Box::new(je));
+        handler.span_err(span, "foo");
+
+        let bytes = output.lock().unwrap();
+        let actual_output = str::from_utf8(&bytes).unwrap();
+        let actual_output: TestData = decode(actual_output).unwrap();
+
+        assert_eq!(expected_output, actual_output)
+    })
+}
+
+#[test]
+fn empty() {
+    test_positions(
+        " ",
+        (0, 1),
+        SpanTestData {
+            byte_start: 0,
+            byte_end: 1,
+            line_start: 1,
+            column_start: 1,
+            line_end: 1,
+            column_end: 2,
+        },
+    )
+}
+
+#[test]
+fn bom() {
+    test_positions(
+        "\u{feff} ",
+        (0, 1),
+        SpanTestData {
+            byte_start: 3,
+            byte_end: 4,
+            line_start: 1,
+            column_start: 1,
+            line_end: 1,
+            column_end: 2,
+        },
+    )
+}
+
+#[test]
+fn lf_newlines() {
+    test_positions(
+        "\nmod foo;\nmod bar;\n",
+        (5, 12),
+        SpanTestData {
+            byte_start: 5,
+            byte_end: 12,
+            line_start: 2,
+            column_start: 5,
+            line_end: 3,
+            column_end: 3,
+        },
+    )
+}
+
+#[test]
+fn crlf_newlines() {
+    test_positions(
+        "\r\nmod foo;\r\nmod bar;\r\n",
+        (5, 12),
+        SpanTestData {
+            byte_start: 6,
+            byte_end: 14,
+            line_start: 2,
+            column_start: 5,
+            line_end: 3,
+            column_end: 3,
+        },
+    )
+}
+
+#[test]
+fn crlf_newlines_with_bom() {
+    test_positions(
+        "\u{feff}\r\nmod foo;\r\nmod bar;\r\n",
+        (5, 12),
+        SpanTestData {
+            byte_start: 9,
+            byte_end: 17,
+            line_start: 2,
+            column_start: 5,
+            line_end: 3,
+            column_end: 3,
+        },
+    )
+}
+
+#[test]
+fn span_before_crlf() {
+    test_positions(
+        "foo\r\nbar",
+        (2, 3),
+        SpanTestData {
+            byte_start: 2,
+            byte_end: 3,
+            line_start: 1,
+            column_start: 3,
+            line_end: 1,
+            column_end: 4,
+        },
+    )
+}
+
+#[test]
+fn span_on_crlf() {
+    test_positions(
+        "foo\r\nbar",
+        (3, 4),
+        SpanTestData {
+            byte_start: 3,
+            byte_end: 5,
+            line_start: 1,
+            column_start: 4,
+            line_end: 2,
+            column_end: 1,
+        },
+    )
+}
+
+#[test]
+fn span_after_crlf() {
+    test_positions(
+        "foo\r\nbar",
+        (4, 5),
+        SpanTestData {
+            byte_start: 5,
+            byte_end: 6,
+            line_start: 2,
+            column_start: 1,
+            line_end: 2,
+            column_end: 2,
+        },
+    )
+}

src/libsyntax/source_map.rs

@@ -283,6 +283,7 @@ impl SourceMap {
         mut file_local_lines: Vec<BytePos>,
         mut file_local_multibyte_chars: Vec<MultiByteChar>,
         mut file_local_non_narrow_chars: Vec<NonNarrowChar>,
+        mut file_local_normalized_pos: Vec<NormalizedPos>,
     ) -> Lrc<SourceFile> {
         let start_pos = self.next_start_pos();
 
@@ -301,6 +302,10 @@ impl SourceMap {
             *swc = *swc + start_pos;
         }
 
+        for nc in &mut file_local_normalized_pos {
+            nc.pos = nc.pos + start_pos;
+        }
+
         let source_file = Lrc::new(SourceFile {
             name: filename,
             name_was_remapped,
@@ -314,6 +319,7 @@ impl SourceMap {
             lines: file_local_lines,
             multibyte_chars: file_local_multibyte_chars,
             non_narrow_chars: file_local_non_narrow_chars,
+            normalized_pos: file_local_normalized_pos,
             name_hash,
         });
 

src/libsyntax/tests.rs

@@ -110,8 +110,8 @@ struct SpanLabel {
     label: &'static str,
 }
 
-struct Shared<T: Write> {
-    data: Arc<Mutex<T>>,
+crate struct Shared<T: Write> {
+    pub data: Arc<Mutex<T>>,
 }
 
 impl<T: Write> Write for Shared<T> {