Rollup merge of #141399 - GuillaumeGomez:extracted-doctest, r=aDotInTheVoid

[rustdoc] Give more information into extracted doctest information

Follow-up of #134531.

This update fragment the doctest code into its sub-parts to give more control to the end users on how they want to use it.

The new JSON looks like this:

```json
{
  "format_version":2,
  "doctests":[
    {
      "file":"$DIR/extract-doctests-result.rs",
      "line":8,
      "doctest_attributes":{
        "original":"",
        "should_panic":false,
        "no_run":false,
        "ignore":"None",
        "rust":true,
        "test_harness":false,
        "compile_fail":false,
        "standalone_crate":false,
        "error_codes":[],
        "edition":null,
        "added_css_classes":[],
        "unknown":[]
      },
      "original_code":"let x = 12;\nOk(())",
      "doctest_code":{
        "crate_level":"#![allow(unused)]\n",
        "code":"let x = 12;\nOk(())",
        "wrapper":{
          "before":"fn main() { fn _inner() -> core::result::Result<(), impl core::fmt::Debug> {\n",
          "after":"\n} _inner().unwrap() }",
          "returns_result":true
        }
      },
      "name":"$DIR/extract-doctests-result.rs - (line 8)"
    }
  ]
}
```

for this doctest:

```rust
let x = 12;
Ok(())
```

With this, I think it matches what you need ``@ojeda?`` If so, once merged I'll update the patch I sent to RfL.

r? ``@aDotInTheVoid``

Author: matthiaskrgr

src/doc/rustdoc/src/unstable-features.md

@@ -581,7 +581,9 @@ For this rust code:
 
 ```rust
 /// ```
+/// #![allow(dead_code)]
 /// let x = 12;
+/// Ok(())
 /// ```
 pub trait Trait {}
 ```
@@ -590,10 +592,10 @@ The generated output (formatted) will look like this:
 
 ```json
 {
-  "format_version": 1,
+  "format_version": 2,
   "doctests": [
     {
-      "file": "foo.rs",
+      "file": "src/lib.rs",
       "line": 1,
       "doctest_attributes": {
         "original": "",
@@ -609,9 +611,17 @@ The generated output (formatted) will look like this:
         "added_css_classes": [],
         "unknown": []
       },
-      "original_code": "let x = 12;",
-      "doctest_code": "#![allow(unused)]\nfn main() {\nlet x = 12;\n}",
-      "name": "foo.rs - Trait (line 1)"
+      "original_code": "#![allow(dead_code)]\nlet x = 12;\nOk(())",
+      "doctest_code": {
+        "crate_level": "#![allow(unused)]\n#![allow(dead_code)]\n\n",
+        "code": "let x = 12;\nOk(())",
+        "wrapper": {
+          "before": "fn main() { fn _inner() -> core::result::Result<(), impl core::fmt::Debug> {\n",
+          "after": "\n} _inner().unwrap() }",
+          "returns_result": true
+        }
+      },
+      "name": "src/lib.rs - (line 1)"
     }
   ]
 }
@@ -624,6 +634,10 @@ The generated output (formatted) will look like this:
    * `doctest_attributes` contains computed information about the attributes used on the doctests. For more information about doctest attributes, take a look [here](write-documentation/documentation-tests.html#attributes).
    * `original_code` is the code as written in the source code before rustdoc modifies it.
    * `doctest_code` is the code modified by rustdoc that will be run. If there is a fatal syntax error, this field will not be present.
+     * `crate_level` is the crate level code (like attributes or `extern crate`) that will be added at the top-level of the generated doctest.
+     * `code` is "naked" doctest without anything from `crate_level` and `wrapper` content.
+     * `wrapper` contains extra code that will be added before and after `code`.
+       * `returns_result` is a boolean. If `true`, it means that the doctest returns a `Result` type.
    * `name` is the name generated by rustdoc which represents this doctest.
 
 ### html

src/librustdoc/doctest.rs

@@ -1053,14 +1053,14 @@ fn doctest_run_fn(
     let report_unused_externs = |uext| {
         unused_externs.lock().unwrap().push(uext);
     };
-    let (full_test_code, full_test_line_offset) = doctest.generate_unique_doctest(
+    let (wrapped, full_test_line_offset) = doctest.generate_unique_doctest(
         &scraped_test.text,
         scraped_test.langstr.test_harness,
         &global_opts,
         Some(&global_opts.crate_name),
     );
     let runnable_test = RunnableDocTest {
-        full_test_code,
+        full_test_code: wrapped.to_string(),
         full_test_line_offset,
         test_opts,
         global_opts,

src/librustdoc/doctest/extracted.rs

@@ -3,8 +3,10 @@
 //! This module contains the logic to extract doctests and output a JSON containing this
 //! information.
 
+use rustc_span::edition::Edition;
 use serde::Serialize;
 
+use super::make::DocTestWrapResult;
 use super::{BuildDocTestBuilder, ScrapedDocTest};
 use crate::config::Options as RustdocOptions;
 use crate::html::markdown;
@@ -14,7 +16,7 @@ use crate::html::markdown;
 /// This integer is incremented with every breaking change to the API,
 /// and is returned along with the JSON blob into the `format_version` root field.
 /// Consuming code should assert that this value matches the format version(s) that it supports.
-const FORMAT_VERSION: u32 = 1;
+const FORMAT_VERSION: u32 = 2;
 
 #[derive(Serialize)]
 pub(crate) struct ExtractedDocTests {
@@ -34,7 +36,16 @@ impl ExtractedDocTests {
         options: &RustdocOptions,
     ) {
         let edition = scraped_test.edition(options);
+        self.add_test_with_edition(scraped_test, opts, edition)
+    }
 
+    /// This method is used by unit tests to not have to provide a `RustdocOptions`.
+    pub(crate) fn add_test_with_edition(
+        &mut self,
+        scraped_test: ScrapedDocTest,
+        opts: &super::GlobalTestOptions,
+        edition: Edition,
+    ) {
         let ScrapedDocTest { filename, line, langstr, text, name, global_crate_attrs, .. } =
             scraped_test;
 
@@ -44,8 +55,7 @@ impl ExtractedDocTests {
             .edition(edition)
             .lang_str(&langstr)
             .build(None);
-
-        let (full_test_code, size) = doctest.generate_unique_doctest(
+        let (wrapped, _size) = doctest.generate_unique_doctest(
             &text,
             langstr.test_harness,
             opts,
@@ -55,11 +65,46 @@ impl ExtractedDocTests {
             file: filename.prefer_remapped_unconditionaly().to_string(),
             line,
             doctest_attributes: langstr.into(),
-            doctest_code: if size != 0 { Some(full_test_code) } else { None },
+            doctest_code: match wrapped {
+                DocTestWrapResult::Valid { crate_level_code, wrapper, code } => Some(DocTest {
+                    crate_level: crate_level_code,
+                    code,
+                    wrapper: wrapper.map(
+                        |super::make::WrapperInfo { before, after, returns_result, .. }| {
+                            WrapperInfo { before, after, returns_result }
+                        },
+                    ),
+                }),
+                DocTestWrapResult::SyntaxError { .. } => None,
+            },
             original_code: text,
             name,
         });
     }
+
+    #[cfg(test)]
+    pub(crate) fn doctests(&self) -> &[ExtractedDocTest] {
+        &self.doctests
+    }
+}
+
+#[derive(Serialize)]
+pub(crate) struct WrapperInfo {
+    before: String,
+    after: String,
+    returns_result: bool,
+}
+
+#[derive(Serialize)]
+pub(crate) struct DocTest {
+    crate_level: String,
+    code: String,
+    /// This field can be `None` if one of the following conditions is true:
+    ///
+    /// * The doctest's codeblock has the `test_harness` attribute.
+    /// * The doctest has a `main` function.
+    /// * The doctest has the `![no_std]` attribute.
+    pub(crate) wrapper: Option<WrapperInfo>,
 }
 
 #[derive(Serialize)]
@@ -69,7 +114,7 @@ pub(crate) struct ExtractedDocTest {
     doctest_attributes: LangString,
     original_code: String,
     /// `None` if the code syntax is invalid.
-    doctest_code: Option<String>,
+    pub(crate) doctest_code: Option<DocTest>,
     name: String,
 }
 

src/librustdoc/doctest/make.rs

@@ -196,6 +196,80 @@ pub(crate) struct DocTestBuilder {
     pub(crate) can_be_merged: bool,
 }
 
+/// Contains needed information for doctest to be correctly generated with expected "wrapping".
+pub(crate) struct WrapperInfo {
+    pub(crate) before: String,
+    pub(crate) after: String,
+    pub(crate) returns_result: bool,
+    insert_indent_space: bool,
+}
+
+impl WrapperInfo {
+    fn len(&self) -> usize {
+        self.before.len() + self.after.len()
+    }
+}
+
+/// Contains a doctest information. Can be converted into code with the `to_string()` method.
+pub(crate) enum DocTestWrapResult {
+    Valid {
+        crate_level_code: String,
+        /// This field can be `None` if one of the following conditions is true:
+        ///
+        /// * The doctest's codeblock has the `test_harness` attribute.
+        /// * The doctest has a `main` function.
+        /// * The doctest has the `![no_std]` attribute.
+        wrapper: Option<WrapperInfo>,
+        /// Contains the doctest processed code without the wrappers (which are stored in the
+        /// `wrapper` field).
+        code: String,
+    },
+    /// Contains the original source code.
+    SyntaxError(String),
+}
+
+impl std::string::ToString for DocTestWrapResult {
+    fn to_string(&self) -> String {
+        match self {
+            Self::SyntaxError(s) => s.clone(),
+            Self::Valid { crate_level_code, wrapper, code } => {
+                let mut prog_len = code.len() + crate_level_code.len();
+                if let Some(wrapper) = wrapper {
+                    prog_len += wrapper.len();
+                    if wrapper.insert_indent_space {
+                        prog_len += code.lines().count() * 4;
+                    }
+                }
+                let mut prog = String::with_capacity(prog_len);
+
+                prog.push_str(crate_level_code);
+                if let Some(wrapper) = wrapper {
+                    prog.push_str(&wrapper.before);
+
+                    // add extra 4 spaces for each line to offset the code block
+                    if wrapper.insert_indent_space {
+                        write!(
+                            prog,
+                            "{}",
+                            fmt::from_fn(|f| code
+                                .lines()
+                                .map(|line| fmt::from_fn(move |f| write!(f, "    {line}")))
+                                .joined("\n", f))
+                        )
+                        .unwrap();
+                    } else {
+                        prog.push_str(code);
+                    }
+                    prog.push_str(&wrapper.after);
+                } else {
+                    prog.push_str(code);
+                }
+                prog
+            }
+        }
+    }
+}
+
 impl DocTestBuilder {
     fn invalid(
         global_crate_attrs: Vec<String>,
@@ -228,50 +302,49 @@ impl DocTestBuilder {
         dont_insert_main: bool,
         opts: &GlobalTestOptions,
         crate_name: Option<&str>,
-    ) -> (String, usize) {
+    ) -> (DocTestWrapResult, usize) {
         if self.invalid_ast {
             // If the AST failed to compile, no need to go generate a complete doctest, the error
             // will be better this way.
             debug!("invalid AST:\n{test_code}");
-            return (test_code.to_string(), 0);
+            return (DocTestWrapResult::SyntaxError(test_code.to_string()), 0);
         }
         let mut line_offset = 0;
-        let mut prog = String::new();
-        let everything_else = self.everything_else.trim();
-
+        let mut crate_level_code = String::new();
+        let processed_code = self.everything_else.trim();
         if self.global_crate_attrs.is_empty() {
             // If there aren't any attributes supplied by #![doc(test(attr(...)))], then allow some
             // lints that are commonly triggered in doctests. The crate-level test attributes are
             // commonly used to make tests fail in case they trigger warnings, so having this there in
             // that case may cause some tests to pass when they shouldn't have.
-            prog.push_str("#![allow(unused)]\n");
+            crate_level_code.push_str("#![allow(unused)]\n");
             line_offset += 1;
         }
 
         // Next, any attributes that came from #![doc(test(attr(...)))].
         for attr in &self.global_crate_attrs {
-            prog.push_str(&format!("#![{attr}]\n"));
+            crate_level_code.push_str(&format!("#![{attr}]\n"));
             line_offset += 1;
         }
 
         // Now push any outer attributes from the example, assuming they
         // are intended to be crate attributes.
         if !self.crate_attrs.is_empty() {
-            prog.push_str(&self.crate_attrs);
+            crate_level_code.push_str(&self.crate_attrs);
             if !self.crate_attrs.ends_with('\n') {
-                prog.push('\n');
+                crate_level_code.push('\n');
             }
         }
         if !self.maybe_crate_attrs.is_empty() {
-            prog.push_str(&self.maybe_crate_attrs);
+            crate_level_code.push_str(&self.maybe_crate_attrs);
             if !self.maybe_crate_attrs.ends_with('\n') {
-                prog.push('\n');
+                crate_level_code.push('\n');
             }
         }
         if !self.crates.is_empty() {
-            prog.push_str(&self.crates);
+            crate_level_code.push_str(&self.crates);
             if !self.crates.ends_with('\n') {
-                prog.push('\n');
+                crate_level_code.push('\n');
             }
         }
 
@@ -289,17 +362,20 @@ impl DocTestBuilder {
         {
             // rustdoc implicitly inserts an `extern crate` item for the own crate
             // which may be unused, so we need to allow the lint.
-            prog.push_str("#[allow(unused_extern_crates)]\n");
+            crate_level_code.push_str("#[allow(unused_extern_crates)]\n");
 
-            prog.push_str(&format!("extern crate r#{crate_name};\n"));
+            crate_level_code.push_str(&format!("extern crate r#{crate_name};\n"));
             line_offset += 1;
         }
 
         // FIXME: This code cannot yet handle no_std test cases yet
-        if dont_insert_main || self.has_main_fn || prog.contains("![no_std]") {
-            prog.push_str(everything_else);
+        let wrapper = if dont_insert_main
+            || self.has_main_fn
+            || crate_level_code.contains("![no_std]")
+        {
+            None
         } else {
-            let returns_result = everything_else.ends_with("(())");
+            let returns_result = processed_code.ends_with("(())");
             // Give each doctest main function a unique name.
             // This is for example needed for the tooling around `-C instrument-coverage`.
             let inner_fn_name = if let Some(ref test_id) = self.test_id {
@@ -333,28 +409,22 @@ impl DocTestBuilder {
             // /// ``` <- end of the inner main
             line_offset += 1;
 
-            prog.push_str(&main_pre);
-
-            // add extra 4 spaces for each line to offset the code block
-            if opts.insert_indent_space {
-                write!(
-                    prog,
-                    "{}",
-                    fmt::from_fn(|f| everything_else
-                        .lines()
-                        .map(|line| fmt::from_fn(move |f| write!(f, "    {line}")))
-                        .joined("\n", f))
-                )
-                .unwrap();
-            } else {
-                prog.push_str(everything_else);
-            };
-            prog.push_str(&main_post);
-        }
-
-        debug!("final doctest:\n{prog}");
+            Some(WrapperInfo {
+                before: main_pre,
+                after: main_post,
+                returns_result,
+                insert_indent_space: opts.insert_indent_space,
+            })
+        };
 
-        (prog, line_offset)
+        (
+            DocTestWrapResult::Valid {
+                code: processed_code.to_string(),
+                wrapper,
+                crate_level_code,
+            },
+            line_offset,
+        )
     }
 }