Auto merge of #54861 - rep-nop:find_main_in_doctest, r=estebank

rustdoc: Replaces fn main search and extern crate search with proper parsing during doctests.

Fixes #21299.
Fixes #33731.

Let me know if there's any additional changes you'd like made!

src/librustdoc/test.rs

@@ -378,7 +378,7 @@ pub fn make_test(s: &str,
                  dont_insert_main: bool,
                  opts: &TestOptions)
                  -> (String, usize) {
-    let (crate_attrs, everything_else) = partition_source(s);
+    let (crate_attrs, everything_else, crates) = partition_source(s);
     let everything_else = everything_else.trim();
     let mut line_offset = 0;
     let mut prog = String::new();
@@ -402,10 +402,84 @@ pub fn make_test(s: &str,
     // are intended to be crate attributes.
     prog.push_str(&crate_attrs);
 
+    // Uses libsyntax to parse the doctest and find if there's a main fn and the extern
+    // crate already is included.
+    let (already_has_main, already_has_extern_crate) = crate::syntax::with_globals(|| {
+        use crate::syntax::{ast, parse::{self, ParseSess}, source_map::FilePathMapping};
+        use crate::syntax_pos::FileName;
+        use errors::emitter::EmitterWriter;
+        use errors::Handler;
+
+        let filename = FileName::Anon;
+        let source = crates + &everything_else;
+
+        // any errors in parsing should also appear when the doctest is compiled for real, so just
+        // send all the errors that libsyntax emits directly into a Sink instead of stderr
+        let cm = Lrc::new(SourceMap::new(FilePathMapping::empty()));
+        let emitter = EmitterWriter::new(box io::sink(), None, false, false);
+        let handler = Handler::with_emitter(false, false, box emitter);
+        let sess = ParseSess::with_span_handler(handler, cm);
+
+        debug!("about to parse: \n{}", source);
+
+        let mut found_main = false;
+        let mut found_extern_crate = cratename.is_none();
+
+        let mut parser = match parse::maybe_new_parser_from_source_str(&sess, filename, source) {
+            Ok(p) => p,
+            Err(errs) => {
+                for mut err in errs {
+                    err.cancel();
+                }
+
+                return (found_main, found_extern_crate);
+            }
+        };
+
+        loop {
+            match parser.parse_item() {
+                Ok(Some(item)) => {
+                    if !found_main {
+                        if let ast::ItemKind::Fn(..) = item.node {
+                            if item.ident.as_str() == "main" {
+                                found_main = true;
+                            }
+                        }
+                    }
+
+                    if !found_extern_crate {
+                        if let ast::ItemKind::ExternCrate(original) = item.node {
+                            // This code will never be reached if `cratename` is none because
+                            // `found_extern_crate` is initialized to `true` if it is none.
+                            let cratename = cratename.unwrap();
+
+                            match original {
+                                Some(name) => found_extern_crate = name.as_str() == cratename,
+                                None => found_extern_crate = item.ident.as_str() == cratename,
+                            }
+                        }
+                    }
+
+                    if found_main && found_extern_crate {
+                        break;
+                    }
+                }
+                Ok(None) => break,
+                Err(mut e) => {
+                    e.cancel();
+                    break;
+                }
+            }
+        }
+
+        (found_main, found_extern_crate)
+    });
+
     // Don't inject `extern crate std` because it's already injected by the
     // compiler.
-    if !s.contains("extern crate") && !opts.no_crate_inject && cratename != Some("std") {
+    if !already_has_extern_crate && !opts.no_crate_inject && cratename != Some("std") {
         if let Some(cratename) = cratename {
+            // Make sure its actually used if not included.
             if s.contains(cratename) {
                 prog.push_str(&format!("extern crate {};\n", cratename));
                 line_offset += 1;
@@ -413,19 +487,6 @@ pub fn make_test(s: &str,
         }
     }
 
-    // FIXME (#21299): prefer libsyntax or some other actual parser over this
-    // best-effort ad hoc approach
-    let already_has_main = s.lines()
-        .map(|line| {
-            let comment = line.find("//");
-            if let Some(comment_begins) = comment {
-                &line[0..comment_begins]
-            } else {
-                line
-            }
-        })
-        .any(|code| code.contains("fn main"));
-
     if dont_insert_main || already_has_main {
         prog.push_str(everything_else);
     } else {
@@ -441,9 +502,10 @@ pub fn make_test(s: &str,
 }
 
 // FIXME(aburka): use a real parser to deal with multiline attributes
-fn partition_source(s: &str) -> (String, String) {
+fn partition_source(s: &str) -> (String, String, String) {
     let mut after_header = false;
     let mut before = String::new();
+    let mut crates = String::new();
     let mut after = String::new();
 
     for line in s.lines() {
@@ -457,12 +519,17 @@ fn partition_source(s: &str) -> (String, String) {
             after.push_str(line);
             after.push_str("\n");
         } else {
+            if trimline.starts_with("#[macro_use] extern crate")
+                || trimline.starts_with("extern crate") {
+                crates.push_str(line);
+                crates.push_str("\n");
+            }
             before.push_str(line);
             before.push_str("\n");
         }
     }
 
-    (before, after)
+    (before, after, crates)
 }
 
 pub trait Tester {
@@ -1014,4 +1081,38 @@ assert_eq!(2+2, 4);
         let output = make_test(input, None, false, &opts);
         assert_eq!(output, (expected, 1));
     }
+
+    #[test]
+    fn make_test_issues_21299_33731() {
+        let opts = TestOptions::default();
+
+        let input =
+"// fn main
+assert_eq!(2+2, 4);";
+
+        let expected =
+"#![allow(unused)]
+fn main() {
+// fn main
+assert_eq!(2+2, 4);
+}".to_string();
+
+        let output = make_test(input, None, false, &opts);
+        assert_eq!(output, (expected, 2));
+
+        let input =
+"extern crate hella_qwop;
+assert_eq!(asdf::foo, 4);";
+
+        let expected =
+"#![allow(unused)]
+extern crate hella_qwop;
+extern crate asdf;
+fn main() {
+assert_eq!(asdf::foo, 4);
+}".to_string();
+
+        let output = make_test(input, Some("asdf"), false, &opts);
+        assert_eq!(output, (expected, 3));
+    }
 }

src/libsyntax/lib.rs

@@ -70,6 +70,23 @@ macro_rules! panictry {
     })
 }
 
+// A variant of 'panictry!' that works on a Vec<Diagnostic> instead of a single DiagnosticBuilder.
+macro_rules! panictry_buffer {
+    ($handler:expr, $e:expr) => ({
+        use std::result::Result::{Ok, Err};
+        use errors::{FatalError, DiagnosticBuilder};
+        match $e {
+            Ok(e) => e,
+            Err(errs) => {
+                for e in errs {
+                    DiagnosticBuilder::new_diagnostic($handler, e).emit();
+                }
+                FatalError.raise()
+            }
+        }
+    })
+}
+
 #[macro_export]
 macro_rules! unwrap_or {
     ($opt:expr, $default:expr) => {

src/libsyntax/parse/lexer/mod.rs

@@ -11,7 +11,7 @@
 use ast::{self, Ident};
 use syntax_pos::{self, BytePos, CharPos, Pos, Span, NO_EXPANSION};
 use source_map::{SourceMap, FilePathMapping};
-use errors::{Applicability, FatalError, DiagnosticBuilder};
+use errors::{Applicability, FatalError, Diagnostic, DiagnosticBuilder};
 use parse::{token, ParseSess};
 use str::char_at;
 use symbol::{Symbol, keywords};
@@ -175,6 +175,16 @@ impl<'a> StringReader<'a> {
         self.fatal_errs.clear();
     }
 
+    pub fn buffer_fatal_errors(&mut self) -> Vec<Diagnostic> {
+        let mut buffer = Vec::new();
+
+        for err in self.fatal_errs.drain(..) {
+            err.buffer(&mut buffer);
+        }
+
+        buffer
+    }
+
     pub fn peek(&self) -> TokenAndSpan {
         // FIXME(pcwalton): Bad copy!
         TokenAndSpan {
@@ -251,6 +261,17 @@ impl<'a> StringReader<'a> {
         Ok(sr)
     }
 
+    pub fn new_or_buffered_errs(sess: &'a ParseSess,
+                                source_file: Lrc<syntax_pos::SourceFile>,
+                                override_span: Option<Span>) -> Result<Self, Vec<Diagnostic>> {
+        let mut sr = StringReader::new_raw(sess, source_file, override_span);
+        if sr.advance_token().is_err() {
+            Err(sr.buffer_fatal_errors())
+        } else {
+            Ok(sr)
+        }
+    }
+
     pub fn retokenize(sess: &'a ParseSess, mut span: Span) -> Self {
         let begin = sess.source_map().lookup_byte_offset(span.lo());
         let end = sess.source_map().lookup_byte_offset(span.hi());

src/libsyntax/parse/mod.rs

@@ -15,7 +15,7 @@ use ast::{self, CrateConfig, NodeId};
 use early_buffered_lints::{BufferedEarlyLint, BufferedEarlyLintId};
 use source_map::{SourceMap, FilePathMapping};
 use syntax_pos::{Span, SourceFile, FileName, MultiSpan};
-use errors::{Handler, ColorConfig, DiagnosticBuilder};
+use errors::{Handler, ColorConfig, Diagnostic, DiagnosticBuilder};
 use feature_gate::UnstableFeatures;
 use parse::parser::Parser;
 use ptr::P;
@@ -174,12 +174,21 @@ pub fn parse_stream_from_source_str(name: FileName, source: String, sess: &Parse
     source_file_to_stream(sess, sess.source_map().new_source_file(name, source), override_span)
 }
 
-// Create a new parser from a source string
+/// Create a new parser from a source string
 pub fn new_parser_from_source_str(sess: &ParseSess, name: FileName, source: String)
                                       -> Parser {
-    let mut parser = source_file_to_parser(sess, sess.source_map().new_source_file(name, source));
+    panictry_buffer!(&sess.span_diagnostic, maybe_new_parser_from_source_str(sess, name, source))
+}
+
+/// Create a new parser from a source string. Returns any buffered errors from lexing the initial
+/// token stream.
+pub fn maybe_new_parser_from_source_str(sess: &ParseSess, name: FileName, source: String)
+    -> Result<Parser, Vec<Diagnostic>>
+{
+    let mut parser = maybe_source_file_to_parser(sess,
+                                                 sess.source_map().new_source_file(name, source))?;
     parser.recurse_into_file_modules = false;
-    parser
+    Ok(parser)
 }
 
 /// Create a new parser, handling errors as appropriate
@@ -204,14 +213,23 @@ crate fn new_sub_parser_from_file<'a>(sess: &'a ParseSess,
 
 /// Given a source_file and config, return a parser
 fn source_file_to_parser(sess: & ParseSess, source_file: Lrc<SourceFile>) -> Parser {
+    panictry_buffer!(&sess.span_diagnostic,
+                     maybe_source_file_to_parser(sess, source_file))
+}
+
+/// Given a source_file and config, return a parser. Returns any buffered errors from lexing the
+/// initial token stream.
+fn maybe_source_file_to_parser(sess: &ParseSess, source_file: Lrc<SourceFile>)
+    -> Result<Parser, Vec<Diagnostic>>
+{
     let end_pos = source_file.end_pos;
-    let mut parser = stream_to_parser(sess, source_file_to_stream(sess, source_file, None));
+    let mut parser = stream_to_parser(sess, maybe_file_to_stream(sess, source_file, None)?);
 
     if parser.token == token::Eof && parser.span.is_dummy() {
         parser.span = Span::new(end_pos, end_pos, parser.span.ctxt());
     }
 
-    parser
+    Ok(parser)
 }
 
 // must preserve old name for now, because quote! from the *existing*
@@ -243,9 +261,25 @@ fn file_to_source_file(sess: &ParseSess, path: &Path, spanopt: Option<Span>)
 pub fn source_file_to_stream(sess: &ParseSess,
                              source_file: Lrc<SourceFile>,
                              override_span: Option<Span>) -> TokenStream {
-    let mut srdr = lexer::StringReader::new(sess, source_file, override_span);
+    panictry_buffer!(&sess.span_diagnostic, maybe_file_to_stream(sess, source_file, override_span))
+}
+
+/// Given a source file, produce a sequence of token-trees. Returns any buffered errors from
+/// parsing the token tream.
+pub fn maybe_file_to_stream(sess: &ParseSess,
+                            source_file: Lrc<SourceFile>,
+                            override_span: Option<Span>) -> Result<TokenStream, Vec<Diagnostic>> {
+    let mut srdr = lexer::StringReader::new_or_buffered_errs(sess, source_file, override_span)?;
     srdr.real_token();
-    panictry!(srdr.parse_all_token_trees())
+
+    match srdr.parse_all_token_trees() {
+        Ok(stream) => Ok(stream),
+        Err(err) => {
+            let mut buffer = Vec::with_capacity(1);
+            err.buffer(&mut buffer);
+            Err(buffer)
+        }
+    }
 }
 
 /// Given stream and the `ParseSess`, produce a parser

src/tools/error_index_generator/main.rs

@@ -10,6 +10,7 @@
 
 #![feature(rustc_private)]
 
+extern crate env_logger;
 extern crate syntax;
 extern crate rustdoc;
 extern crate serialize as rustc_serialize;
@@ -264,6 +265,7 @@ fn parse_args() -> (OutputFormat, PathBuf) {
 }
 
 fn main() {
+    env_logger::init();
     PLAYGROUND.with(|slot| {
         *slot.borrow_mut() = Some((None, String::from("https://play.rust-lang.org/")));
     });