From 20a787adb4140002acb8e5ea21536069e184ffaf Mon Sep 17 00:00:00 2001
From: Radomir Dopieralski <openstack@sheep.art.pl>
Date: Tue, 25 Jul 2017 01:09:47 +0200
Subject: [PATCH] extmod/ure: Handle some escape sequences.

Fix MicroPython #3176

Handle escape sequences inside regular expressions. This adds
handling for \a, \b, \f, \n, \r, \v and \\.
---
 extmod/re1.5/compilecode.c | 46 +++++++++++++++++++++++++++++++++-----
 tests/extmod/ure1.py       | 13 +++++++++++
 2 files changed, 54 insertions(+), 5 deletions(-)

diff --git a/extmod/re1.5/compilecode.c b/extmod/re1.5/compilecode.c
index a685a508a0..01d3d14988 100644
--- a/extmod/re1.5/compilecode.c
+++ b/extmod/re1.5/compilecode.c
@@ -10,6 +10,29 @@
 #define EMIT(at, byte) (code ? (code[at] = byte) : (at))
 #define PC (prog->bytelen)
 
+
+static char unescape(char c) {
+    switch (c) {
+        case 'a':
+            return '\a';
+        case 'b':
+            return '\b';
+        case 'f':
+            return '\f';
+        case 'n':
+            return '\n';
+        case 'r':
+            return '\r';
+        case 'v':
+            return '\v';
+        case 'x':
+            return '\\';
+        default:
+            return c;
+    }
+}
+
+
 static const char *_compilecode(const char *re, ByteProg *prog, int sizecode)
 {
     char *code = sizecode ? NULL : prog->insts;
@@ -22,13 +45,16 @@ static const char *_compilecode(const char *re, ByteProg *prog, int sizecode)
         case '\\':
             re++;
             if (!*re) return NULL; // Trailing backslash
+            term = PC;
             if ((*re | 0x20) == 'd' || (*re | 0x20) == 's' || (*re | 0x20) == 'w') {
-                term = PC;
                 EMIT(PC++, NamedClass);
                 EMIT(PC++, *re);
-                prog->len++;
-                break;
+            } else {
+                EMIT(PC++, Char);
+                EMIT(PC++, unescape(*re));
             }
+            prog->len++;
+            break;
         default:
             term = PC;
             EMIT(PC++, Char);
@@ -54,11 +80,21 @@ static const char *_compilecode(const char *re, ByteProg *prog, int sizecode)
             prog->len++;
             for (cnt = 0; *re != ']'; re++, cnt++) {
                 if (!*re) return NULL;
-                EMIT(PC++, *re);
+                if (*re == '\\') {
+                    re += 1;
+                    EMIT(PC++, unescape(*re));
+                } else {
+                    EMIT(PC++, *re);
+                }
                 if (re[1] == '-' && re[2] != ']') {
                     re += 2;
                 }
-                EMIT(PC++, *re);
+                if (*re == '\\') {
+                    re += 1;
+                    EMIT(PC++, unescape(*re));
+                } else {
+                    EMIT(PC++, *re);
+                }
             }
             EMIT(term + 1, cnt);
             break;
diff --git a/tests/extmod/ure1.py b/tests/extmod/ure1.py
index 54471ed4f9..710720c8b6 100644
--- a/tests/extmod/ure1.py
+++ b/tests/extmod/ure1.py
@@ -28,6 +28,19 @@ try:
 except IndexError:
     print("IndexError")
 
+r = re.compile(r"\n")
+m = r.match("\n")
+print(m.group(0))
+m = r.match("\\")
+print(m)
+r = re.compile(r"[\n-\r]")
+m = r.match("\n")
+print(m.group(0))
+r = re.compile(r"[\]]")
+m = r.match("]")
+print(m.group(0))
+print("===")
+
 r = re.compile("[a-cu-z]")
 m = r.match("a")
 print(m.group(0))