IDL parser rewrite in Python

Source/bindings/scripts/blink_idl_lexer.py

+# Copyright (C) 2013 Google Inc. All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions are
+# met:
+#
+#     * Redistributions of source code must retain the above copyright
+# notice, this list of conditions and the following disclaimer.
+#     * Redistributions in binary form must reproduce the above
+# copyright notice, this list of conditions and the following disclaimer
+# in the documentation and/or other materials provided with the
+# distribution.
+#     * Neither the name of Google Inc. nor the names of its
+# contributors may be used to endorse or promote products derived from
+# this software without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+"""Lexer for Blink IDL.
+
+The lexer uses the PLY (Python Lex-Yacc) library to build a tokenizer which
+understands the Blink dialect of Web IDL and produces a token stream suitable
+for the Blink IDL parser.
+
+Blink IDL is identical to Web IDL at the token level, but the base lexer

[haraken, 2013/07/16 14:17:51]

Let's add "FIXME:".

[Nils Barth (inactive), 2013/07/17 12:05:09]

Ok, consolidated FIXME at top.

+does not discard comments. We need to override (and not include comments in
+the token stream), as otherwise comments must be explicitly included in the
+phrase grammar of the parser.
+If the base lexer is changed to discard comments, we can simply used the base
+lexer and eliminate this separate lexer.
+
+Web IDL:
+    http://www.w3.org/TR/WebIDL/
+Web IDL Grammar:
+    http://www.w3.org/TR/WebIDL/#idl-grammar
+PLY:
+    http://www.dabeaz.com/ply/
+"""
+
+# Disable attribute validation, as lint can't import parent class to check
+# pylint: disable=E1101

[haraken, 2013/07/16 14:17:51]

Is this comment helpful?

[Nils Barth (inactive), 2013/07/17 12:05:09]

Yup, otherwise I get a pylint error, I think on presubmit.

+
+import os.path
+import sys
+
+# Base lexer is in Chromium src/tools/idl_parser
+module_path, module_name = os.path.split(__file__)
+tools_dir = os.path.join(module_path, os.pardir, os.pardir, os.pardir, os.pardir, os.pardir, 'tools')
+sys.path.append(tools_dir)

[haraken, 2013/07/16 14:17:51]

Instead of writing the relative path here, can you pass the path from a GYP
file? We want to centralize path handling in GYP files.

[Nils Barth (inactive), 2013/07/17 12:05:09]

(See long reply.)

[Nils Barth (inactive), 2013/07/17 12:09:41]

I don’t think there’s an easy way to specify the Python module search path from
GYP.
In Perl we can just use -I<dir> to specify include directories, but in
Python they don’t want you to do that.
To change the Python search path, you can either:
* modify sys.path inside the script, or
* change the environment variable PYTHONPATH:
http://docs.python.org/2/tutorial/modules.html#the-module-search-path

I don’t see how to set environment variables inside GYP, and no .gyp
file uses PYTHONPATH (0 hits):
https://code.google.com/p/chromium/codesearch#search/&q=PYTHONPATH%20file:.*%...
...so I believe this is not possible or discouraged.

This path handling is copied directly from the base Pepper parser
idl_lexer.py/idl_parser.py; I was aiming to be consistent:
https://code.google.com/p/chromium/codesearch#chromium/src/tools/idl_parser/i...
...and changing sys.path is very common in the Chromium codebase (782 hits):
https://code.google.com/p/chromium/codesearch#search/&q=sys.path%20lang:%5Epy...

Anyway, we still need to list the included files explicitly as ‘inputs’
(so GYP computes dependencies correctly), so if anything moves we’ll
detect it immediately on regyping; this shouldn’t bite us.


FWIW, if this were all one huge Python package, we could use:
from .....tools.idl_parser.idl_lexer import IDLLexer
...but we can’t use this across separate packages.

[haraken, 2013/07/21 14:31:50]

Given that this is a limitation of Python, I'm OK with the hard-coded paths.

+
+from idl_parser.idl_lexer import IDLLexer
+
+REMOVE_TOKENS = ['COMMENT']
+
+
+class BlinkIDLLexer(IDLLexer):
+    # ignore comments
+    # FIXME: Upstream; needed because base doesn't ignore comments

[haraken, 2013/07/16 14:17:51]

You can remove this comment, because you already said this above.

[Nils Barth (inactive), 2013/07/17 12:05:09]

Got it, consolidate FIXMEs at top.

+    def t_COMMENT(self, t):
+        r'(/\*(.|\n)*?\*/)|(//.*(\n[ \t]*//.*)*)'
+        self.AddLines(t.value.count('\n'))

[haraken, 2013/07/16 14:17:51]

I'm just curious: What is this doing?

[Nils Barth (inactive), 2013/07/17 12:05:09]

This tracks the current line number, which is used for error reporting.

+
+    # Analogs to _AddToken/_AddTokens in base lexer
+    # Needed to remove COMMENT token, since comments ignored
+    # FIXME: Upstream

[haraken, 2013/07/16 14:17:51]

I'd remove this comment.

[Nils Barth (inactive), 2013/07/17 12:05:09]

Got it, consolidated FIXME.

+    def _RemoveToken(self, token):
+        try:
+            self.tokens.remove(token)
+        except ValueError:
+            raise RuntimeError('Token "%s" not present, so cannot be removed ' % token)

[haraken, 2013/07/16 14:17:51]

Does this mean that we raise an exception when parsing an IDL file without any
comment?

[Nils Barth (inactive), 2013/07/17 12:05:09]

No no, this is used when initializing the lexer, not running it.
It means if you try to remove a token class that doesn’t exist, you’ll get an
error.
E.g., if you say  _RemoveToken(‘GREATEREQUAL’), you’ll get an error (b/c >=
isn’t a token in Web IDL).

+
+    def _RemoveTokens(self, tokens):
+        for token in tokens:
+            self._RemoveToken(token)
+
+    def __init__(self):
+        IDLLexer.__init__(self)
+        self._RemoveTokens(REMOVE_TOKENS)
+
+
+# If run by itself, attempt to build the lexer
+if __name__ == '__main__':
+    lexer = BlinkIDLLexer()