Avoid accidental library name matches when parsing ldd output.

* Use a single pattern that matches against potentially complete paths.
* Extract filename only afterwards on platforms where it is necessary.
* Match patterns against complete words in ldd output instead of searching
  for them inside the lines - this avoids unintentional matches without
  complexity of negative lookbehinds and negative lookaheads.

Fixes issue #208.

4 files changed, 144 insertions, 33 deletions

diff --git a/giscanner/shlibs.py b/giscanner/shlibs.py
index 525bdba2..32fc5e44 100644
--- a/giscanner/shlibs.py
+++ b/giscanner/shlibs.py
@@ -49,19 +49,17 @@ def _resolve_libtool(options, binary, libraries):
 #  libpangoft2-1.0.so.0 => /usr/lib/libpangoft2-1.0.so.0 (0x006c1000)
 #
 # We say that if something in the output looks like libpangoft2<blah>
-# then the *first* such in the output is the soname. We require <blah>
-# to start with [^A-Za-z0-9_-] to avoid problems with libpango vs libpangoft2
-#
-# The negative lookbehind at the start is to avoid problems if someone
-# is crazy enough to name a library liblib<foo> when lib<foo> exists.
-#
-# Match absolute paths on OS X to conform to how libraries are usually
-# referenced on OS X systems.
+# then the *first* such in the output is the soname.
 def _ldd_library_pattern(library_name):
-    pattern = "(?<![A-Za-z0-9_-])(lib*%s[^A-Za-z0-9_-][^\s\(\)]*)"
-    if platform.system() == 'Darwin':
-        pattern = "([^\s]*lib*%s[^A-Za-z0-9_-][^\s\(\)]*)"
-    return re.compile(pattern % re.escape(library_name))
+    return re.compile(r"""^
+    # Require trailing slash to avoid matching liblibfoo when looking for libfoo.
+    (.*[/])?
+    lib%s
+    # Prohibit library name characters to avoid matching libpangoft2 when looking for libpango.
+    [^/A-Za-z0-9_-]
+    # Anything but the path separator to avoid matching directories.
+    [^/]*
+    $""" % re.escape(library_name), re.VERBOSE)
 
 
 # This is a what we do for non-la files. We assume that we are on an
@@ -96,8 +94,7 @@ def _resolve_non_libtool(options, binary, libraries):
 
     if host_os() == 'nt':
         cc = CCompiler()
-        shlibs = cc.resolve_windows_libs(libraries, options)
-
+        return cc.resolve_windows_libs(libraries, options)
     else:
         args = []
         libtool = get_libtool_command(options)
@@ -109,31 +106,41 @@ def _resolve_non_libtool(options, binary, libraries):
             args.extend(['otool', '-L', binary.args[0]])
         else:
             args.extend(['ldd', binary.args[0]])
-        proc = subprocess.Popen(args, stdout=subprocess.PIPE)
-        patterns = {}
-        for library in libraries:
-            patterns[library] = _ldd_library_pattern(library)
-
-        shlibs = []
-        for line in proc.stdout:
-            line = line.decode('ascii')
-            # ldd on *BSD show the argument passed on the first line even if
-            # there is only one argument. We have to ignore it because it is
-            # possible for the name of the binary to match _ldd_library_pattern.
-            if line == binary.args[0] + ':\n':
-                continue
+        output = subprocess.check_output(args, universal_newlines=True, errors='replace')
+
+        # Use absolute paths on OS X to conform to how libraries are usually
+        # referenced on OS X systems, and file names everywhere else.
+        basename = platform.system() != 'Darwin'
+        return resolve_from_ldd_output(libraries, output, basename=basename)
+
+
+def resolve_from_ldd_output(libraries, output, basename=False):
+    patterns = {}
+    for library in libraries:
+        patterns[library] = _ldd_library_pattern(library)
+
+    shlibs = []
+    for line in output.splitlines():
+        # ldd on *BSD show the argument passed on the first line even if
+        # there is only one argument. We have to ignore it because it is
+        # possible for the name of the binary to match _ldd_library_pattern.
+        if line.endswith(':'):
+            continue
+        for word in line.split():
             for library, pattern in patterns.items():
-                m = pattern.search(line)
+                m = pattern.match(word)
                 if m:
                     del patterns[library]
-                    shlibs.append(m.group(1))
+                    shlibs.append(m.group())
                     break
 
-        if len(patterns) > 0:
-            raise SystemExit(
-                "ERROR: can't resolve libraries to shared libraries: " +
-                ", ".join(patterns.keys()))
+    if len(patterns) > 0:
+        raise SystemExit(
+            "ERROR: can't resolve libraries to shared libraries: " +
+            ", ".join(patterns.keys()))
 
+    if basename:
+        shlibs = list(map(os.path.basename, shlibs))
     return shlibs
 
 
diff --git a/tests/scanner/Makefile.am b/tests/scanner/Makefile.am
index 6b705f5d..279addfe 100644
--- a/tests/scanner/Makefile.am
+++ b/tests/scanner/Makefile.am
@@ -229,6 +229,7 @@ CHECKDOCS =
 endif
 
 PYTESTS = \
+	test_shlibs.py \
 	test_sourcescanner.py \
 	test_transformer.py
 
diff --git a/tests/scanner/meson.build b/tests/scanner/meson.build
index f8b794d2..3458c32e 100644
--- a/tests/scanner/meson.build
+++ b/tests/scanner/meson.build
@@ -3,6 +3,7 @@ scanner_test_env = environment()
 scanner_test_env.append('PYTHONPATH', join_paths(meson.current_build_dir(), '../../'))
 
 scanner_test_files = [
+  'test_shlibs.py',
   'test_sourcescanner.py',
   'test_transformer.py'
 ]
diff --git a/tests/scanner/test_shlibs.py b/tests/scanner/test_shlibs.py
new file mode 100644
index 00000000..c3e28892
--- /dev/null
+++ b/tests/scanner/test_shlibs.py
@@ -0,0 +1,102 @@
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
+
+import unittest
+
+from giscanner.shlibs import resolve_from_ldd_output
+
+
+class TestLddParser(unittest.TestCase):
+
+    def test_resolve_from_ldd_output(self):
+        output = '''\
+            libglib-2.0.so.0 => /usr/lib/x86_64-linux-gnu/libglib-2.0.so.0 (0x00007fbe12d68000)
+            libgtk-3.so.0 => /usr/lib/x86_64-linux-gnu/libgtk-3.so.0 (0x00007fbe12462000)
+            libgdk-3.so.0 => /usr/lib/x86_64-linux-gnu/libgdk-3.so.0 (0x00007fbe1216c000)
+            libpango-1.0.so.0 => /usr/lib/x86_64-linux-gnu/libpango-1.0.so.0 (0x00007fbe11d1a000)
+            libatk-1.0.so.0 => /usr/lib/x86_64-linux-gnu/libatk-1.0.so.0 (0x00007fbe11af4000)'''
+        libraries = ['glib-2.0', 'gtk-3', 'pango-1.0']
+
+        self.assertEqual(
+            ['libglib-2.0.so.0', 'libgtk-3.so.0', 'libpango-1.0.so.0'],
+            resolve_from_ldd_output(libraries, output))
+
+    def test_unresolved_library(self):
+        output = ''
+        libraries = ['foo']
+
+        with self.assertRaises(SystemExit, msg='can\'t resolve libraries to shared libraries: foo'):
+            resolve_from_ldd_output(libraries, output)
+
+    def test_prefixed_library_name(self):
+        output = '''\
+           /usr/lib/liblibX.so
+           /usr/lib/libX.so'''
+
+        self.assertEqual(
+            ['/usr/lib/libX.so'],
+            resolve_from_ldd_output(['X'], output))
+        self.assertEqual(
+            ['/usr/lib/liblibX.so'],
+            resolve_from_ldd_output(['libX'], output))
+
+    def test_suffixed_library_name(self):
+        output = '''\
+            libpangocairo.so.0 => /usr/lib/x86_64-linux-gnu/libpangocairo.so.0 (0x00)
+            libpangoft2.so.0 => /usr/lib/x86_64-linux-gnu/libpangoft2.so.0 (0x00)
+            libpango.so.0 => /usr/lib/x86_64-linux-gnu/libpango.so.0 (0x00)'''
+        libraries = ['pango']
+
+        self.assertEqual(
+            ['libpango.so.0'],
+            resolve_from_ldd_output(libraries, output))
+
+    def test_header_is_ignored(self):
+        output = '''/tmp-introspection/libfoo.so.999:
+            0000000000000000 0000000000000000 rlib  0    3   0      /usr/local/lib/libfoo.so.1'''
+        libraries = ['foo']
+
+        self.assertEqual(
+            ['/usr/local/lib/libfoo.so.1'],
+            resolve_from_ldd_output(libraries, output))
+
+    def test_executable_path_includes_library_name(self):
+        """
+        Regression test for https://gitlab.gnome.org/GNOME/gobject-introspection/issues/208
+        """
+        output = '''/usr/ports/pobj/libgepub-0.6.0/build-amd64/tmp-introspectnxmyodg1/Gepub-0.6:
+            Start            End              Type  Open Ref GrpRef Name
+            00001066c8400000 00001066c8605000 exe   2    0   0      /usr/ports/pobj/libgepub-0.6.0/build-amd64/tmp-introspectnxmyodg1/Gepub-0.6
+            000010690019c000 00001069003a8000 rlib  0    1   0      /usr/local/lib/libgepub-0.6.so.0.0'''
+        libraries = ['gepub-0.6']
+
+        self.assertEqual(
+            ['/usr/local/lib/libgepub-0.6.so.0.0'],
+            resolve_from_ldd_output(libraries, output))
+
+    def test_library_path_includes_library_name(self):
+        output = '''/usr/ports/pobj/gnome-music-3.28.1/build-amd64/tmp-introspectuz5xaun3/Gd-1.0:
+            Start            End              Type  Open Ref GrpRef Name
+            0000070e40f00000 0000070e41105000 exe   2    0   0      /usr/ports/pobj/gnome-music-3.28.1/build-amd64/tmp-introspectuz5xaun3/Gd-1.0
+            00000710f9b39000 00000710f9d51000 rlib  0    1   0      /usr/ports/pobj/gnome-music-3.28.1/build-amd64/subprojects/libgd/libgd/libgd.so'''
+        libraries = ['gd']
+
+        self.assertEqual(
+            ['/usr/ports/pobj/gnome-music-3.28.1/build-amd64/subprojects/libgd/libgd/libgd.so'],
+            resolve_from_ldd_output(libraries, output))
+
+    def test_basename(self):
+        output = '''/usr/lib/libfoo.so'''
+
+        self.assertEqual(
+            ['/usr/lib/libfoo.so'],
+            resolve_from_ldd_output(['foo'], output, basename=False))
+        self.assertEqual(
+            ['libfoo.so'],
+            resolve_from_ldd_output(['foo'], output, basename=True))
+
+
+if __name__ == '__main__':
+    unittest.main()