From f5633927ff56c057679347258931e4aa67a3e0bb Mon Sep 17 00:00:00 2001
From: Michel Oosterhof <micheloosterhof@users.noreply.github.com>
Date: Sun, 16 Sep 2018 16:46:49 +0400
Subject: [PATCH] Fix #883, add ',' to wordchars (#884)

* Fix #883, add ',' to wordchars
---
 src/cowrie/shell/honeypot.py | 17 +++++++++--------
 src/cowrie/shell/shlex.py    |  3 ---
 src/cowrie/test/test_echo.py |  7 +++++++
 3 files changed, 16 insertions(+), 11 deletions(-)

diff --git a/src/cowrie/shell/honeypot.py b/src/cowrie/shell/honeypot.py
index c76238f1..3de09e60 100644
--- a/src/cowrie/shell/honeypot.py
+++ b/src/cowrie/shell/honeypot.py
@@ -32,15 +32,21 @@ class HoneyPotShell(object):
 
     def lineReceived(self, line):
         log.msg(eventid='cowrie.command.input', input=line, format='CMD: %(input)s')
-        self.lexer = shlex.shlex(instream=line, punctuation_chars=True)
+        self.lexer = shlex.shlex(instream=line, punctuation_chars=True, posix=True)
         # Add these special characters that are not in the default lexer
-        self.lexer.wordchars += '@%{}=$:+^'
+        self.lexer.wordchars += '@%{}=$:+^,'
         tokens = []
         while True:
             try:
                 tok = self.lexer.get_token()
                 # log.msg("tok: %s" % (repr(tok)))
 
+                if tok == self.lexer.eof:
+                    if tokens:
+                        self.cmdpending.append((tokens))
+                        tokens = []
+                    break
+
                 # Ignore parentheses
                 tok_len = len(tok)
                 tok = tok.strip('(')
@@ -48,13 +54,8 @@ class HoneyPotShell(object):
                 if len(tok) != tok_len and tok == '':
                     continue
 
-                if tok == self.lexer.eof:
-                    if tokens:
-                        self.cmdpending.append((tokens))
-                        tokens = []
-                    break
                 # For now, treat && and || same as ;, just execute without checking return code
-                elif tok == '&&' or tok == '||':
+                if tok == '&&' or tok == '||':
                     if tokens:
                         self.cmdpending.append((tokens))
                         tokens = []
diff --git a/src/cowrie/shell/shlex.py b/src/cowrie/shell/shlex.py
index 06b25edf..45053a6b 100644
--- a/src/cowrie/shell/shlex.py
+++ b/src/cowrie/shell/shlex.py
@@ -43,9 +43,6 @@ class shlex:
         self.commenters = '#'
         self.wordchars = ('abcdfeghijklmnopqrstuvwxyz'
                           'ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_')
-        if self.posix:
-            self.wordchars += ('ßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿ'
-                               'ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞ')
         self.whitespace = ' \t\r\n'
         self.whitespace_split = False
         self.quotes = '\'"'
diff --git a/src/cowrie/test/test_echo.py b/src/cowrie/test/test_echo.py
index 31fd34b0..2919048c 100644
--- a/src/cowrie/test/test_echo.py
+++ b/src/cowrie/test/test_echo.py
@@ -115,5 +115,12 @@ class ShellEchoCommandTests(unittest.TestCase):
         self.proto.lineReceived(b'echo test > test012; grep test test012')
         self.assertEquals(self.tr.value(), b'test\n' + PROMPT)
 
+    def test_echo_command_013(self):
+        """
+        echo "ls""ls"
+        """
+        self.proto.lineReceived(b'echo "ls""ls"')
+        self.assertEquals(self.tr.value(), b'lsls\n' + PROMPT)
+
     def tearDown(self):
         self.proto.connectionLost("tearDown From Unit Test")