From 3fa8efc34b089ef6a4b89012eda4bba714d1f776 Mon Sep 17 00:00:00 2001
From: Florian Bruhin <git@the-compiler.org>
Date: Mon, 3 Nov 2014 21:27:07 +0100
Subject: [PATCH] Initial shlex fork

---
 qutebrowser/commands/runners.py      |   4 +-
 qutebrowser/test/utils/test_split.py |  76 ++++++
 qutebrowser/test/utils/test_utils.py |  50 ----
 qutebrowser/utils/split.py           | 331 +++++++++++++++++++++++++++
 qutebrowser/utils/utils.py           |  49 ----
 5 files changed, 409 insertions(+), 101 deletions(-)
 create mode 100644 qutebrowser/test/utils/test_split.py
 create mode 100644 qutebrowser/utils/split.py

diff --git a/qutebrowser/commands/runners.py b/qutebrowser/commands/runners.py
index c28c0b96f..f053a710b 100644
--- a/qutebrowser/commands/runners.py
+++ b/qutebrowser/commands/runners.py
@@ -24,7 +24,7 @@ from PyQt5.QtWebKitWidgets import QWebPage
 
 from qutebrowser.config import config
 from qutebrowser.commands import cmdexc, cmdutils
-from qutebrowser.utils import message, log, utils, objreg
+from qutebrowser.utils import message, log, utils, objreg, split
 
 
 def replace_variables(win_id, arglist):
@@ -238,7 +238,7 @@ class CommandRunner(QObject):
         if argstr is None:
             self._args = []
         elif self._cmd.split:
-            self._args = utils.safe_shlex_split(argstr)
+            self._args = split.split(argstr)
         else:
             # If split=False, we still want to split the flags, but not
             # everything after that.
diff --git a/qutebrowser/test/utils/test_split.py b/qutebrowser/test/utils/test_split.py
new file mode 100644
index 000000000..139af6c86
--- /dev/null
+++ b/qutebrowser/test/utils/test_split.py
@@ -0,0 +1,76 @@
+# vim: ft=python fileencoding=utf-8 sts=4 sw=4 et:
+
+# Copyright 2014 Florian Bruhin (The Compiler) <mail@qutebrowser.org>
+#
+# This file is part of qutebrowser.
+#
+# qutebrowser is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# qutebrowser is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with qutebrowser.  If not, see <http://www.gnu.org/licenses/>.
+
+"""Tests for qutebrowser.utils.split."""
+
+import unittest
+
+from qutebrowser.utils import split
+
+
+class SplitTests(unittest.TestCase):
+
+    """Test split."""
+
+    def test_normal(self):
+        """Test split with a simple string."""
+        items = split.split('one two')
+        self.assertEqual(items, ['one', 'two'])
+
+    def test_quoted(self):
+        """Test split with a normally quoted string."""
+        items = split.split('one "two three" four')
+        self.assertEqual(items, ['one', 'two three', 'four'])
+
+    def test_single_quoted(self):
+        """Test split with a single quoted string."""
+        items = split.split("one 'two three' four")
+        self.assertEqual(items, ['one', 'two three', 'four'])
+
+    def test_escaped(self):
+        """Test split with a normal escaped string."""
+        items = split.split(r'one "two\" three" four')
+        self.assertEqual(items, ['one', 'two" three', 'four'])
+
+    def test_escaped_single(self):
+        """Test split with a single escaped string."""
+        items = split.split(r"one 'two'\'' three' four")
+        self.assertEqual(items, ['one', "two' three", 'four'])
+
+    def test_unbalanced_quotes(self):
+        """Test split with unbalanded quotes."""
+        items = split.split(r'one "two three')
+        self.assertEqual(items, ['one', 'two three'])
+
+    def test_unbalanced_single_quotes(self):
+        """Test split with unbalanded single quotes."""
+        items = split.split(r"one 'two three")
+        self.assertEqual(items, ['one', "two three"])
+
+    def test_unfinished_escape(self):
+        """Test split with an unfinished escape."""
+        items = split.split('one\\')
+        self.assertEqual(items, ['one\\'])
+
+    def test_both(self):
+        """Test split with an unfinished escape and quotes.."""
+        items = split.split('one "two\\')
+        self.assertEqual(items, ['one', 'two\\'])
+
+
diff --git a/qutebrowser/test/utils/test_utils.py b/qutebrowser/test/utils/test_utils.py
index 470cd449a..6b8ad1553 100644
--- a/qutebrowser/test/utils/test_utils.py
+++ b/qutebrowser/test/utils/test_utils.py
@@ -110,56 +110,6 @@ class DottedGetattrTests(unittest.TestCase):
             _ = utils.dotted_getattr(self, 'test.foo.baz')
 
 
-class SafeShlexSplitTests(unittest.TestCase):
-
-    """Test safe_shlex_split."""
-
-    def test_normal(self):
-        """Test safe_shlex_split with a simple string."""
-        items = utils.safe_shlex_split('one two')
-        self.assertEqual(items, ['one', 'two'])
-
-    def test_quoted(self):
-        """Test safe_shlex_split with a normally quoted string."""
-        items = utils.safe_shlex_split('one "two three" four')
-        self.assertEqual(items, ['one', 'two three', 'four'])
-
-    def test_single_quoted(self):
-        """Test safe_shlex_split with a single quoted string."""
-        items = utils.safe_shlex_split("one 'two three' four")
-        self.assertEqual(items, ['one', 'two three', 'four'])
-
-    def test_escaped(self):
-        """Test safe_shlex_split with a normal escaped string."""
-        items = utils.safe_shlex_split(r'one "two\" three" four')
-        self.assertEqual(items, ['one', 'two" three', 'four'])
-
-    def test_escaped_single(self):
-        """Test safe_shlex_split with a single escaped string."""
-        items = utils.safe_shlex_split(r"one 'two'\'' three' four")
-        self.assertEqual(items, ['one', "two' three", 'four'])
-
-    def test_unbalanced_quotes(self):
-        """Test safe_shlex_split with unbalanded quotes."""
-        items = utils.safe_shlex_split(r'one "two three')
-        self.assertEqual(items, ['one', 'two three'])
-
-    def test_unbalanced_single_quotes(self):
-        """Test safe_shlex_split with unbalanded single quotes."""
-        items = utils.safe_shlex_split(r"one 'two three")
-        self.assertEqual(items, ['one', "two three"])
-
-    def test_unfinished_escape(self):
-        """Test safe_shlex_split with an unfinished escape."""
-        items = utils.safe_shlex_split('one\\')
-        self.assertEqual(items, ['one\\'])
-
-    def test_both(self):
-        """Test safe_shlex_split with an unfinished escape and quotes.."""
-        items = utils.safe_shlex_split('one "two\\')
-        self.assertEqual(items, ['one', 'two\\'])
-
-
 class InterpolateColorTests(unittest.TestCase):
 
     """Tests for interpolate_color.
diff --git a/qutebrowser/utils/split.py b/qutebrowser/utils/split.py
new file mode 100644
index 000000000..b201aeef2
--- /dev/null
+++ b/qutebrowser/utils/split.py
@@ -0,0 +1,331 @@
+# vim: ft=python fileencoding=utf-8 sts=4 sw=4 et:
+
+# Copyright 2014 Florian Bruhin (The Compiler) <mail@qutebrowser.org>
+#
+# This file is part of qutebrowser.
+#
+# qutebrowser is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# qutebrowser is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with qutebrowser.  If not, see <http://www.gnu.org/licenses/>.
+
+"""Our own fork of shlex.split with some added and removed features."""
+
+import os
+import re
+import sys
+from collections import deque
+
+from io import StringIO
+
+
+class ShellLexer:
+    "A lexical analyzer class for simple shell-like syntaxes."
+    def __init__(self, instream=None, infile=None, posix=False):
+        if isinstance(instream, str):
+            instream = StringIO(instream)
+        if instream is not None:
+            self.instream = instream
+            self.infile = infile
+        else:
+            self.instream = sys.stdin
+            self.infile = None
+        self.posix = posix
+        if posix:
+            self.eof = None
+        else:
+            self.eof = ''
+        self.commenters = '#'
+        self.wordchars = ('abcdfeghijklmnopqrstuvwxyz'
+                          'ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_')
+        if self.posix:
+            self.wordchars += ('ßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿ'
+                               'ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞ')
+        self.whitespace = ' \t\r\n'
+        self.whitespace_split = False
+        self.quotes = '\'"'
+        self.escape = '\\'
+        self.escapedquotes = '"'
+        self.state = ' '
+        self.pushback = deque()
+        self.lineno = 1
+        self.debug = 0
+        self.token = ''
+        self.filestack = deque()
+        self.source = None
+        if self.debug:
+            print('shlex: reading from %s, line %d' \
+                  % (self.instream, self.lineno))
+
+    def push_token(self, tok):
+        "Push a token onto the stack popped by the get_token method"
+        if self.debug >= 1:
+            print("shlex: pushing token " + repr(tok))
+        self.pushback.appendleft(tok)
+
+    def push_source(self, newstream, newfile=None):
+        "Push an input source onto the lexer's input source stack."
+        if isinstance(newstream, str):
+            newstream = StringIO(newstream)
+        self.filestack.appendleft((self.infile, self.instream, self.lineno))
+        self.infile = newfile
+        self.instream = newstream
+        self.lineno = 1
+        if self.debug:
+            if newfile is not None:
+                print('shlex: pushing to file %s' % (self.infile,))
+            else:
+                print('shlex: pushing to stream %s' % (self.instream,))
+
+    def pop_source(self):
+        "Pop the input source stack."
+        self.instream.close()
+        (self.infile, self.instream, self.lineno) = self.filestack.popleft()
+        if self.debug:
+            print('shlex: popping to %s, line %d' \
+                  % (self.instream, self.lineno))
+        self.state = ' '
+
+    def get_token(self):
+        "Get a token from the input stream (or from stack if it's nonempty)"
+        if self.pushback:
+            tok = self.pushback.popleft()
+            if self.debug >= 1:
+                print("shlex: popping token " + repr(tok))
+            return tok
+        # No pushback.  Get a token.
+        raw = self.read_token()
+        # Handle inclusions
+        if self.source is not None:
+            while raw == self.source:
+                spec = self.sourcehook(self.read_token())
+                if spec:
+                    (newfile, newstream) = spec
+                    self.push_source(newstream, newfile)
+                raw = self.get_token()
+        # Maybe we got EOF instead?
+        while raw == self.eof:
+            if not self.filestack:
+                return self.eof
+            else:
+                self.pop_source()
+                raw = self.get_token()
+        # Neither inclusion nor EOF
+        if self.debug >= 1:
+            if raw != self.eof:
+                print("shlex: token=" + repr(raw))
+            else:
+                print("shlex: token=EOF")
+        return raw
+
+    def read_token(self):
+        quoted = False
+        escapedstate = ' '
+        while True:
+            nextchar = self.instream.read(1)
+            if nextchar == '\n':
+                self.lineno = self.lineno + 1
+            if self.debug >= 3:
+                print("shlex: in state", repr(self.state), \
+                      "I see character:", repr(nextchar))
+            if self.state is None:
+                self.token = ''        # past end of file
+                break
+            elif self.state == ' ':
+                if not nextchar:
+                    self.state = None  # end of file
+                    break
+                elif nextchar in self.whitespace:
+                    if self.debug >= 2:
+                        print("shlex: I see whitespace in whitespace state")
+                    if self.token or (self.posix and quoted):
+                        break   # emit current token
+                    else:
+                        continue
+                elif nextchar in self.commenters:
+                    self.instream.readline()
+                    self.lineno = self.lineno + 1
+                elif self.posix and nextchar in self.escape:
+                    escapedstate = 'a'
+                    self.state = nextchar
+                elif nextchar in self.wordchars:
+                    self.token = nextchar
+                    self.state = 'a'
+                elif nextchar in self.quotes:
+                    if not self.posix:
+                        self.token = nextchar
+                    self.state = nextchar
+                elif self.whitespace_split:
+                    self.token = nextchar
+                    self.state = 'a'
+                else:
+                    self.token = nextchar
+                    if self.token or (self.posix and quoted):
+                        break   # emit current token
+                    else:
+                        continue
+            elif self.state in self.quotes:
+                quoted = True
+                if not nextchar:      # end of file
+                    if self.debug >= 2:
+                        print("shlex: I see EOF in quotes state")
+                    # XXX what error should be raised here?
+                    raise ValueError("No closing quotation")
+                if nextchar == self.state:
+                    if not self.posix:
+                        self.token = self.token + nextchar
+                        self.state = ' '
+                        break
+                    else:
+                        self.state = 'a'
+                elif self.posix and nextchar in self.escape and \
+                     self.state in self.escapedquotes:
+                    escapedstate = self.state
+                    self.state = nextchar
+                else:
+                    self.token = self.token + nextchar
+            elif self.state in self.escape:
+                if not nextchar:      # end of file
+                    if self.debug >= 2:
+                        print("shlex: I see EOF in escape state")
+                    # XXX what error should be raised here?
+                    raise ValueError("No escaped character")
+                # In posix shells, only the quote itself or the escape
+                # character may be escaped within quotes.
+                if escapedstate in self.quotes and \
+                   nextchar != self.state and nextchar != escapedstate:
+                    self.token = self.token + self.state
+                self.token = self.token + nextchar
+                self.state = escapedstate
+            elif self.state == 'a':
+                if not nextchar:
+                    self.state = None   # end of file
+                    break
+                elif nextchar in self.whitespace:
+                    if self.debug >= 2:
+                        print("shlex: I see whitespace in word state")
+                    self.state = ' '
+                    if self.token or (self.posix and quoted):
+                        break   # emit current token
+                    else:
+                        continue
+                elif nextchar in self.commenters:
+                    self.instream.readline()
+                    self.lineno = self.lineno + 1
+                    if self.posix:
+                        self.state = ' '
+                        if self.token or (self.posix and quoted):
+                            break   # emit current token
+                        else:
+                            continue
+                elif self.posix and nextchar in self.quotes:
+                    self.state = nextchar
+                elif self.posix and nextchar in self.escape:
+                    escapedstate = 'a'
+                    self.state = nextchar
+                elif nextchar in self.wordchars or nextchar in self.quotes \
+                    or self.whitespace_split:
+                    self.token = self.token + nextchar
+                else:
+                    self.pushback.appendleft(nextchar)
+                    if self.debug >= 2:
+                        print("shlex: I see punctuation in word state")
+                    self.state = ' '
+                    if self.token:
+                        break   # emit current token
+                    else:
+                        continue
+        result = self.token
+        self.token = ''
+        if self.posix and not quoted and result == '':
+            result = None
+        if self.debug > 1:
+            if result:
+                print("shlex: raw token=" + repr(result))
+            else:
+                print("shlex: raw token=EOF")
+        return result
+
+    def sourcehook(self, newfile):
+        "Hook called on a filename to be sourced."
+        if newfile[0] == '"':
+            newfile = newfile[1:-1]
+        # This implements cpp-like semantics for relative-path inclusion.
+        if isinstance(self.infile, str) and not os.path.isabs(newfile):
+            newfile = os.path.join(os.path.dirname(self.infile), newfile)
+        return (newfile, open(newfile, "r"))
+
+    def error_leader(self, infile=None, lineno=None):
+        "Emit a C-compiler-like, Emacs-friendly error-message leader."
+        if infile is None:
+            infile = self.infile
+        if lineno is None:
+            lineno = self.lineno
+        return "\"%s\", line %d: " % (infile, lineno)
+
+    def __iter__(self):
+        return self
+
+    def __next__(self):
+        token = self.get_token()
+        if token == self.eof:
+            raise StopIteration
+        return token
+
+
+def _get_lexer(s):
+    """Get an shlex lexer for split."""
+    if s is None:
+        raise TypeError("Refusing to create a lexer with s=None!")
+    lexer = ShellLexer(s, posix=True)
+    lexer.whitespace_split = True
+    lexer.commenters = ''
+    return lexer
+
+
+def split(s):
+    r"""Split a string via shlex safely (don't bail out on unbalanced quotes).
+
+    We split while the user is typing (for completion), and as
+    soon as ", ' or \ is typed, the string is invalid for shlex,
+    because it encounters EOF while in quote/escape state.
+
+    Here we fix this error temporarily so shlex doesn't blow up,
+    and then retry splitting again.
+
+    Since shlex raises ValueError in both cases we unfortunately
+    have to parse the exception string...
+
+    We try 3 times so multiple errors can be fixed.
+    """
+    orig_s = s
+    for i in range(3):
+        lexer = _get_lexer(s)
+        try:
+            tokens = list(lexer)
+        except ValueError as e:
+            if str(e) not in ("No closing quotation", "No escaped character"):
+                raise
+            # eggs "bacon ham -> eggs "bacon ham"
+            # eggs\ -> eggs\\
+            if lexer.state not in lexer.escape + lexer.quotes:
+                raise AssertionError(
+                    "Lexer state is >{}< while parsing >{}< (attempted fixup: "
+                    ">{}<)".format(lexer.state, orig_s, s))
+            s += lexer.state
+        else:
+            return tokens
+    # We should never arrive here.
+    raise AssertionError(
+        "Gave up splitting >{}< after {} tries. Attempted fixup: >{}<.".format(
+            orig_s, i, s))  # pylint: disable=undefined-loop-variable
+
+
diff --git a/qutebrowser/utils/utils.py b/qutebrowser/utils/utils.py
index 3a96bc09d..1dd13ee77 100644
--- a/qutebrowser/utils/utils.py
+++ b/qutebrowser/utils/utils.py
@@ -22,7 +22,6 @@
 import io
 import sys
 import enum
-import shlex
 import inspect
 import os.path
 import urllib.request
@@ -99,54 +98,6 @@ def dotted_getattr(obj, path):
     return functools.reduce(getattr, path.split('.'), obj)
 
 
-def _get_lexer(s):
-    """Get an shlex lexer for safe_shlex_split."""
-    if s is None:
-        raise TypeError("Refusing to create a lexer with s=None!")
-    lexer = shlex.shlex(s, posix=True)
-    lexer.whitespace_split = True
-    lexer.commenters = ''
-    return lexer
-
-
-def safe_shlex_split(s):
-    r"""Split a string via shlex safely (don't bail out on unbalanced quotes).
-
-    We split while the user is typing (for completion), and as
-    soon as ", ' or \ is typed, the string is invalid for shlex,
-    because it encounters EOF while in quote/escape state.
-
-    Here we fix this error temporarily so shlex doesn't blow up,
-    and then retry splitting again.
-
-    Since shlex raises ValueError in both cases we unfortunately
-    have to parse the exception string...
-
-    We try 3 times so multiple errors can be fixed.
-    """
-    orig_s = s
-    for i in range(3):
-        lexer = _get_lexer(s)
-        try:
-            tokens = list(lexer)
-        except ValueError as e:
-            if str(e) not in ("No closing quotation", "No escaped character"):
-                raise
-            # eggs "bacon ham -> eggs "bacon ham"
-            # eggs\ -> eggs\\
-            if lexer.state not in lexer.escape + lexer.quotes:
-                raise AssertionError(
-                    "Lexer state is >{}< while parsing >{}< (attempted fixup: "
-                    ">{}<)".format(lexer.state, orig_s, s))
-            s += lexer.state
-        else:
-            return tokens
-    # We should never arrive here.
-    raise AssertionError(
-        "Gave up splitting >{}< after {} tries. Attempted fixup: >{}<.".format(
-            orig_s, i, s))  # pylint: disable=undefined-loop-variable
-
-
 def pastebin(name, title, text, parent=None):
     """Paste the text into a pastebin and return the URL.