Support @generated marker to skip code formatting

topecongiro · calebcartwright · commit 1b87a8beaeaf · 2021-09-07T20:37:07.000-05:00
This is a copy of #4296 with these changes: * file is not reopened again to find if the file is generated * first five lines are scanned for `@generated` marker instead of one * no attempt is made to only search for marker in comments `@generated` marker is used by certain tools to understand that the file is generated, so it should be treated differently than a file written by a human: * linters should not be invoked on these files, * diffs in these files are less important, * and these files should not be reformatted. This PR proposes builtin support for `@generated` marker. I have not found a standard for a generated file marker, but: * Facebook [uses `@generated` marker](https://tinyurl.com/fb-generated) * Phabricator tool which was spawned from Facebook internal tool [also understands `@generated` marker](https://git.io/JnVHa) * Cargo inserts `@generated` marker into [generated Cargo.lock files](https://git.io/JnVHP) My personal story is that rust-protobuf project which I maintain was broken twice because of incompatibilities/bugs in rustfmt marker handling: [one](stepancheg/rust-protobuf#493), [two](stepancheg/rust-protobuf#551). (Also, rust-protobuf started generating `@generated` marker [6 years ago](https://git.io/JnV5h)). While rustfmt AST markers are useful to apply to a certain AST elements, disable whole-file-at-once all-tools-at-once text level marker might be easier to use and more reliable for generated code.
diff --git a/Configurations.md b/Configurations.md
@@ -924,6 +924,15 @@ fn add_one(x: i32) -> i32 {
 }
 ```
 
+## `format_generated_files`
+
+Format generated files. A file is considered generated
+if any of the first five lines contains `@generated` marker.
+
+- **Default value**: `false`
+- **Possible values**: `true`, `false`
+- **Stable**: No
+
 ## `format_macro_matchers`
 
 Format the metavariable matching patterns in macros.
diff --git a/src/config/mod.rs b/src/config/mod.rs
@@ -136,6 +136,7 @@ create_config! {
     inline_attribute_width: usize, 0, false,
         "Write an item and its attribute on the same line \
         if their combined width is below a threshold";
+    format_generated_files: bool, false, false, "Format generated files";
 
     // Options that can change the source code beyond whitespace/blocks (somewhat linty things)
     merge_derives: bool, true, true, "Merge multiple `#[derive(...)]` into a single one";
@@ -604,6 +605,7 @@ blank_lines_lower_bound = 0
 edition = "2015"
 version = "One"
 inline_attribute_width = 0
+format_generated_files = false
 merge_derives = true
 use_try_shorthand = false
 use_field_init_shorthand = false
diff --git a/src/formatting.rs b/src/formatting.rs
@@ -10,6 +10,7 @@ use rustc_span::Span;
 use self::newline_style::apply_newline_style;
 use crate::comment::{CharClasses, FullCodeCharKind};
 use crate::config::{Config, FileName, Verbosity};
+use crate::formatting::generated::is_generated_file;
 use crate::issues::BadIssueSeeker;
 use crate::modules::Module;
 use crate::syntux::parser::{DirectoryOwnership, Parser, ParserError};
@@ -18,6 +19,7 @@ use crate::utils::count_newlines;
 use crate::visitor::FmtVisitor;
 use crate::{modules, source_file, ErrorKind, FormatReport, Input, Session};
 
+mod generated;
 mod newline_style;
 
 // A map of the files of a crate, with their new content
@@ -103,7 +105,12 @@ fn format_project<T: FormatHandler>(
     context.parse_session.set_silent_emitter();
 
     for (path, module) in files {
-        let should_ignore = !input_is_stdin && context.ignore_file(&path);
+        let source_file = context.parse_session.span_to_file_contents(module.span);
+        let src = source_file.src.as_ref().expect("SourceFile without src");
+
+        let should_ignore = (!input_is_stdin && context.ignore_file(&path))
+            || (!config.format_generated_files() && is_generated_file(src));
+
         if (config.skip_children() && path != main_file) || should_ignore {
             continue;
         }
diff --git a/src/formatting/generated.rs b/src/formatting/generated.rs
@@ -0,0 +1,7 @@
+/// Returns `true` if the given span is a part of generated files.
+pub(super) fn is_generated_file(original_snippet: &str) -> bool {
+    original_snippet
+        .lines()
+        .take(5) // looking for marker only in the beginning of the file
+        .any(|line| line.contains("@generated"))
+}
diff --git a/src/syntux/session.rs b/src/syntux/session.rs
@@ -175,6 +175,12 @@ impl ParseSess {
         self.parse_sess.source_map().span_to_filename(span).into()
     }
 
+    pub(crate) fn span_to_file_contents(&self, span: Span) -> Lrc<rustc_span::SourceFile> {
+        self.parse_sess
+            .source_map()
+            .lookup_source_file(span.data().lo)
+    }
+
     pub(crate) fn span_to_first_line_string(&self, span: Span) -> String {
         let file_lines = self.parse_sess.source_map().span_to_lines(span).ok();
 
diff --git a/src/test/mod.rs b/src/test/mod.rs
@@ -693,7 +693,7 @@ fn read_significant_comments(file_name: &Path) -> HashMap<String, String> {
     reader
         .lines()
         .map(|line| line.expect("failed getting line"))
-        .take_while(|line| line_regex.is_match(line))
+        .filter(|line| line_regex.is_match(line))
         .filter_map(|line| {
             regex.captures_iter(&line).next().map(|capture| {
                 (
diff --git a/tests/source/configs/format_generated_files/false.rs b/tests/source/configs/format_generated_files/false.rs
@@ -0,0 +1,8 @@
+// @generated
+// rustfmt-format_generated_files: false
+
+fn main()
+{
+    println!("hello, world")
+    ;
+}
diff --git a/tests/source/configs/format_generated_files/true.rs b/tests/source/configs/format_generated_files/true.rs
@@ -0,0 +1,8 @@
+// @generated
+// rustfmt-format_generated_files: true
+
+fn main()
+{
+    println!("hello, world")
+    ;
+}
diff --git a/tests/target/configs/format_generated_files/false.rs b/tests/target/configs/format_generated_files/false.rs
@@ -0,0 +1,8 @@
+// @generated
+// rustfmt-format_generated_files: false
+
+fn main()
+{
+    println!("hello, world")
+    ;
+}
diff --git a/tests/target/configs/format_generated_files/true.rs b/tests/target/configs/format_generated_files/true.rs
@@ -0,0 +1,6 @@
+// @generated
+// rustfmt-format_generated_files: true
+
+fn main() {
+    println!("hello, world");
+}

Original file line number	Diff line number	Diff line change
`@@ -693,7 +693,7 @@ fn read_significant_comments(file_name: &Path) -> HashMap<String, String> {`
`693`	`693`	`reader`
`694`	`694`	`.lines()`
`695`	`695`	`.map(\|line\| line.expect("failed getting line"))`
`696`		`- .take_while(\|line\| line_regex.is_match(line))`
	`696`	`+ .filter(\|line\| line_regex.is_match(line))`
`697`	`697`	`.filter_map(\|line\| {`
`698`	`698`	`regex.captures_iter(&line).next().map(\|capture\| {`
`699`	`699`	`(`