Move checking for beginning-of-file indent to a preprocessing step

Author: msimacek

graalpython/com.oracle.graal.python.test/src/com/oracle/graal/python/test/parser/BasicTests.java

@@ -84,6 +84,16 @@ public void moduleWithLincense() throws Exception {
                                         "print(\"module A\")");
     }
 
+    @Test
+    public void leadingIndent1() throws Exception {
+        checkSyntaxErrorMessage(" 1", "IndentationError: unexpected indent");
+    }
+
+    @Test
+    public void leadingIndent2() throws Exception {
+        checkTreeResult(" # foo\npass");
+    }
+
     @Test
     public void annAssign01() throws Exception {
         checkTreeResult("a: int = 1");

graalpython/com.oracle.graal.python.test/testData/goldenFiles/BasicTests/leadingIndent2.tast

@@ -0,0 +1,8 @@
+ModuleRootNode Name: <module 'leadingIndent2'> SourceSection: [0,11]` # foo↵pass`
+  Signature: varArgs=False, varKeywordArgs=False, noArguments=True, positionalOnly=True, requiresKeywordArgs=False
+  FreeVars: None
+  NeedsCellFrame: False
+  FrameDescriptor:  Empty
+  Documentation: None
+  InnerRootNode SourceSection: [0,11]` # foo↵pass`
+    EmptyNode SourceSection: [6,11]`↵pass`

graalpython/com.oracle.graal.python/src/com/oracle/graal/python/parser/PythonErrorStrategy.java

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018, 2019, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2018, 2020, Oracle and/or its affiliates. All rights reserved.
  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
  *
  * The Universal Permissive License (UPL), Version 1.0

graalpython/com.oracle.graal.python/src/com/oracle/graal/python/parser/PythonParserImpl.java

@@ -30,6 +30,8 @@
 import java.io.DataInputStream;
 import java.io.DataOutputStream;
 import java.io.IOException;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
 
 import org.antlr.v4.runtime.CharStreams;
 import org.antlr.v4.runtime.CommonTokenStream;
@@ -76,6 +78,8 @@ public final class PythonParserImpl implements PythonParser, PythonCodeSerialize
     private long numberOfFiles = 0;
     private static final boolean IN_IMAGE_BUILD_TIME = ImageInfo.inImageBuildtimeCode();
 
+    private static final Pattern START_INDENT_REGEX = Pattern.compile("^([ \t]+)[^#\r\n\f]");
+
     public static final DescriptiveBailErrorListener ERROR_LISTENER = new DescriptiveBailErrorListener();
 
     public PythonParserImpl(Env env) {
@@ -274,6 +278,12 @@ private CacheItem parseWithANTLR(ParserMode mode, ParserErrorCallback errors, Py
                 throw errors.raiseInvalidSyntax(source, source.createUnavailableSection(), "encoding problem: %s", e.getEncodingName());
             }
         }
+        // We need to reject inputs starting with indent, but doing it in ANTLR is expensive, so we
+        // do it here manually
+        Matcher matcher = START_INDENT_REGEX.matcher(sourceText);
+        if (matcher.find()) {
+            throw errors.raiseInvalidSyntax(ErrorType.Indentation, source, source.createSection(0, matcher.end(1)), "unexpected indent");
+        }
         // ANTLR parsing
         Python3Parser parser = getPython3Parser(source, sourceText, errors);
         parser.setFactory(sstFactory);

graalpython/com.oracle.graal.python/src/com/oracle/graal/python/parser/antlr/Python3.g4

@@ -1774,8 +1774,8 @@ ASYNC : 'async';
 AWAIT : 'await';
 
 NEWLINE
- : (
-     {atStartOfInput()}?   SPACES |
+ : ( // For performance reasons, rejecting input starting with indent is handled by a preprocessing step in PythonParserImpl
+     // {atStartOfInput()}?   SPACES |
      ( '\r'? '\n' | '\r' | '\f' ) SPACES?
    )
    {