From 6e153b2c017be294a23e78469367346d0f9250e2 Mon Sep 17 00:00:00 2001
From: rjt-gupta <rajat.gupta99924@gmail.com>
Date: Sun, 24 Feb 2019 01:45:45 +0530
Subject: [PATCH] filter unicode fix

---
 mitmproxy/flowfilter.py           | 23 +++++++++++++++++++++++
 test/mitmproxy/test_flowfilter.py | 23 +++++++++++++++++++++++
 2 files changed, 46 insertions(+)

diff --git a/mitmproxy/flowfilter.py b/mitmproxy/flowfilter.py
index 7f8df96f5..0d8f10622 100644
--- a/mitmproxy/flowfilter.py
+++ b/mitmproxy/flowfilter.py
@@ -475,7 +475,30 @@ def _make():
         parts.append(f)
 
     simplerex = "".join(c for c in pp.printables if c not in "()~'\"")
+    alphdevanagari = pp.pyparsing_unicode.Devanagari.alphas
+    alphcyrillic = pp.pyparsing_unicode.Cyrillic.alphas
+    alphgreek = pp.pyparsing_unicode.Greek.alphas
+    alphchinese = pp.pyparsing_unicode.Chinese.alphas
+    alpharabic = pp.pyparsing_unicode.Arabic.alphas
+    alphhebrew = pp.pyparsing_unicode.Hebrew.alphas
+    alphjapanese = pp.pyparsing_unicode.Japanese.alphas
+    alphkorean = pp.pyparsing_unicode.Korean.alphas
+    alphlatin1 = pp.pyparsing_unicode.Latin1.alphas
+    alphlatinA = pp.pyparsing_unicode.LatinA.alphas
+    alphlatinB = pp.pyparsing_unicode.LatinB.alphas
+
     rex = pp.Word(simplerex) |\
+        pp.Word(alphcyrillic) |\
+        pp.Word(alphgreek) |\
+        pp.Word(alphchinese) |\
+        pp.Word(alpharabic) |\
+        pp.Word(alphdevanagari) |\
+        pp.Word(alphhebrew) |\
+        pp.Word(alphjapanese) |\
+        pp.Word(alphkorean) |\
+        pp.Word(alphlatin1) |\
+        pp.Word(alphlatinA) |\
+        pp.Word(alphlatinB) |\
         pp.QuotedString("\"", escChar='\\') |\
         pp.QuotedString("'", escChar='\\')
     for klass in filter_rex:
diff --git a/test/mitmproxy/test_flowfilter.py b/test/mitmproxy/test_flowfilter.py
index 4eb37d813..d53cec7d7 100644
--- a/test/mitmproxy/test_flowfilter.py
+++ b/test/mitmproxy/test_flowfilter.py
@@ -28,6 +28,9 @@ class TestParsing:
         self._dump(p)
         assert len(p.lst) == 2
 
+    def test_non_ascii(self):
+        assert flowfilter.parse("~s шгн")
+
     def test_naked_url(self):
         a = flowfilter.parse("foobar ~h rex")
         assert a.lst[0].expr == "foobar"
@@ -173,10 +176,30 @@ class TestMatchingHTTPFlow:
         assert not self.q("~bq message", q)
         assert not self.q("~bq message", s)
 
+        s.response.text = 'яч'           # Cyrillic
+        assert self.q("~bs яч", s)
+        s.response.text = '测试'          # Chinese
+        assert self.q('~bs 测试', s)
+        s.response.text = 'ॐ'            # Hindi
+        assert self.q('~bs ॐ', s)
+        s.response.text = 'لله'           # Arabic
+        assert self.q('~bs لله', s)
+        s.response.text = 'θεός'          # Greek
+        assert self.q('~bs θεός', s)
+        s.response.text = 'לוהים'          # Hebrew
+        assert self.q('~bs לוהים', s)
+        s.response.text = '神'            # Japanese
+        assert self.q('~bs 神', s)
+        s.response.text = '하나님'         # Korean
+        assert self.q('~bs 하나님', s)
+        s.response.text = 'Äÿ'            # Latin
+        assert self.q('~bs Äÿ', s)
+
         assert not self.q("~bs nomatch", s)
         assert not self.q("~bs content", q)
         assert not self.q("~bs content", s)
         assert not self.q("~bs message", q)
+        s.response.text = 'message'
         assert self.q("~bs message", s)
 
     def test_body(self):