In what order does Python find syntax errors?

user_14375732 0
Viewed 323

Working on creating syntax debugging exercise for students. We have the following example.

def five():
    print('five')
return 5

def hello();
   print('hello')

However when running the file the syntax error is

def hello();
           ^
SyntaxError: invalid syntax

I've looked all over but can not figure out why the compiler [sic] doesn't complain about the return keyword outside of the function, but instead first finds the semicolon error under it.

In what order does Python check the file syntax? Is this part of the specification or is it implemenation defined?

syntax-error python-3.x syntax parsing python
1 Answers

There are (at least) two phases involved: first, the token stream is parsed to produce a parse tree according to the grammar rules. A return statement is part of the flow_stmt rule, which itself is not restricted to being used inside a def statement. It is not a parse error to have a bare return statement. Some selected, relevant rules from the grammar:

single_input: NEWLINE | simple_stmt | compound_stmt NEWLINE
file_input: (NEWLINE | stmt)* ENDMARKER
stmt: simple_stmt | compound_stmt

simple_stmt: small_stmt (';' small_stmt)* [';'] NEWLINE
small_stmt: (expr_stmt | del_stmt | pass_stmt | flow_stmt |
         import_stmt | global_stmt | nonlocal_stmt | assert_stmt)
flow_stmt: break_stmt | continue_stmt | return_stmt | raise_stmt | yield_stmt
return_stmt: 'return' [testlist_star_expr]

The next phase involves turning the parse tree into a syntax tree. At this point, finding a return statement outside of a def statement would produce a syntax error.

A ;, on the other hand, is not part of the definition of funcdef, so a ; in place of the expected : would immediately trigger an error while building the parse tree.

funcdef: 'def' NAME parameters ['->' test] ':' [TYPE_COMMENT] func_body_suite

While it may be possible to report the bare return statement earlier, it clearly does not need to happen, so I would say this is an implementation detail.

user_11268410
Related