From 8650e166888997416fbef75883cd0b4570a0a6ef Mon Sep 17 00:00:00 2001
From: Aleš Raszka <araszka@redhat.com>
Date: Aug 26 2015 12:46:59 +0000
Subject: [PATCH 1/2] Decode .spec file with UTF-8


Resolves: bz#1242997
Signed-off-by: Aleš Raszka <araszka@redhat.com>

---

diff --git a/src/pyrpkg/__init__.py b/src/pyrpkg/__init__.py
index 2d3d3db..68b466f 100644
--- a/src/pyrpkg/__init__.py
+++ b/src/pyrpkg/__init__.py
@@ -2301,7 +2301,16 @@ class Commands(object):
         # Create a list for unused patches
         unused = []
         # Get the content of spec into memory for fast searching
-        spec = open(self.spec, 'r').read()
+        try:
+            spec = open(self.spec, 'r').read().decode('UTF-8')
+        except UnicodeDecodeError as error:
+            # when can't decode file, ignore chars and show warning
+            spec = open(self.spec, 'r').read().decode(encoding='UTF-8', errors='ignore')
+            line, offset = self._byte_offset_to_line_number(spec, error.start)
+            self.log.warning("\'%s\' codec can't decode byte in position %d:%d : %s"
+                                % (error.encoding, line, offset, error.reason))
+        except:
+            raise
         # Replace %{name} with the package name
         spec = spec.replace("%{name}", self.module_name)
         # Replace %{version} with the package version
@@ -2317,6 +2326,17 @@ class Commands(object):
                 unused.append(file)
         return unused
 
+    def _byte_offset_to_line_number(self, text, offset):
+        offset_inc = 0
+        line_num = 1
+        for line in text.split('\n'):
+            if offset_inc + len(line) + 1 > offset:
+                break
+            else:
+                offset_inc += len(line) + 1
+                line_num += 1
+        return [line_num, offset - offset_inc + 1]
+
     def verify_files(self, builddir=None):
         """Run rpmbuild -bl on a module to verify the %files section
 

From c9ce9d5485ad4503ae41d49fb332cd5e2a3dfa77 Mon Sep 17 00:00:00 2001
From: Ales Raszka <araszka@redhat.com>
Date: Sep 16 2015 14:59:54 +0000
Subject: [PATCH 2/2] Add test and docstring to byte offset convertor


Signed-off-by: Ales Raszka <araszka@redhat.com>

---

diff --git a/src/pyrpkg/__init__.py b/src/pyrpkg/__init__.py
index 68b466f..23b2c16 100644
--- a/src/pyrpkg/__init__.py
+++ b/src/pyrpkg/__init__.py
@@ -2309,8 +2309,6 @@ class Commands(object):
             line, offset = self._byte_offset_to_line_number(spec, error.start)
             self.log.warning("\'%s\' codec can't decode byte in position %d:%d : %s"
                                 % (error.encoding, line, offset, error.reason))
-        except:
-            raise
         # Replace %{name} with the package name
         spec = spec.replace("%{name}", self.module_name)
         # Replace %{version} with the package version
@@ -2327,6 +2325,11 @@ class Commands(object):
         return unused
 
     def _byte_offset_to_line_number(self, text, offset):
+        """
+        Convert byte offset (given by e.g. DecodeError) to human readable
+        format (line number and char possition)
+        Return a list with line number and char offset
+        """
         offset_inc = 0
         line_num = 1
         for line in text.split('\n'):
diff --git a/test/commands/test_patch.py b/test/commands/test_patch.py
new file mode 100644
index 0000000..18921f3
--- /dev/null
+++ b/test/commands/test_patch.py
@@ -0,0 +1,53 @@
+# -*- coding: utf-8 -*-
+from . import CommandTestCase
+
+
+class CommandPatchTestCase(CommandTestCase):
+    def setUp(self):
+        super(CommandPatchTestCase, self).setUp()
+        self.text_ascii = "Lorem ipsum dolor sit amet, consectetur elit.\n" \
+            "Sed vel enim nec tortor posuere sodales sit amet mauris.\n" \
+            "Duis ipsum dui, consectetur pretium a, vestibulum.\n"\
+            "Nunc vel consectetur libero. Aenean , metus quis posuere\n" \
+            "vulputate, purus metus fringilla, sit amet interdum tellus\n"
+        self.text_utf8 = "ěšč\n" \
+                         "ščř\n" \
+                         "ýáí"
+
+    def test_byte_offset_first_line(self):
+        import pyrpkg
+        cmd = pyrpkg.Commands(self.path, self.lookaside, self.lookasidehash,
+                              self.lookaside_cgi, self.gitbaseurl,
+                              self.anongiturl, self.branchre, self.kojiconfig,
+                              self.build_client, self.user, self.dist,
+                              self.target, self.quiet)
+        line, offset = cmd._byte_offset_to_line_number(self.text_ascii, 10)
+        #10 byte offset mean line 1 and character 11
+        self.assertEqual(line, 1)
+        self.assertEqual(offset, 11)
+
+    def test_byte_offset_next_line(self):
+        import pyrpkg
+        cmd = pyrpkg.Commands(self.path, self.lookaside, self.lookasidehash,
+                              self.lookaside_cgi, self.gitbaseurl,
+                              self.anongiturl, self.branchre, self.kojiconfig,
+                              self.build_client, self.user, self.dist,
+                              self.target, self.quiet)
+
+        line, offset = cmd._byte_offset_to_line_number(self.text_ascii, 46)
+        # 46 byte offset is first character on second line
+        self.assertEqual(line, 2)
+        self.assertEqual(offset, 1)
+
+    def test_byte_offset_utf8(self):
+        import pyrpkg
+        cmd = pyrpkg.Commands(self.path, self.lookaside, self.lookasidehash,
+                              self.lookaside_cgi, self.gitbaseurl,
+                              self.anongiturl, self.branchre, self.kojiconfig,
+                              self.build_client, self.user, self.dist,
+                              self.target, self.quiet)
+        text = self.text_utf8.decode(encoding='UTF-8', errors='ignore')
+        line, offset = cmd._byte_offset_to_line_number(text, 9)
+        #9 byte offset mean line 3 and second character
+        self.assertEqual(line, 3)
+        self.assertEqual(offset, 2)