file-utilities: rework common prefix computation

Currently, the process for getting the common prefix of a list of file
names is a tad too greedy:
    1. Find the common prefix of all the strings.
    2. Strip the extension from the prefix.
    3. Strip trailing punctuation.

Step 2 may strip dots if there’s trailing whitespace and step 3 may
strip useful punctuation (e.g. parentheses). This commit reworks the
process as such:
    1. Strip the extension from all the file names.
    2. Find the common prefix of all the strings.
    3. Trim trailing whitespace.

Fixes #174.

5 files changed, 46 insertions, 176 deletions

diff --git a/eel/eel-string.c b/eel/eel-string.c
index 85c0841a4..0a41a0bf9 100644
--- a/eel/eel-string.c
+++ b/eel/eel-string.c
@@ -224,42 +224,6 @@ eel_str_replace_substring (const char *string,
     return result;
 }
 
-char *
-eel_str_rtrim_punctuation (char *str)
-{
-    int num_punctuation_chars;
-    int str_len;
-    int num_chars_left;
-    char *current_char_pos;
-    gunichar current_char;
-
-    num_punctuation_chars = 0;
-    str_len = g_utf8_strlen (str, -1);
-    current_char_pos = g_utf8_offset_to_pointer (str, str_len);
-
-    while (num_punctuation_chars <= str_len)
-    {
-        current_char_pos = g_utf8_prev_char (current_char_pos);
-        current_char = g_utf8_get_char (current_char_pos);
-
-        if (!g_unichar_ispunct (current_char) && !g_unichar_isspace (current_char))
-        {
-            break;
-        }
-
-        ++num_punctuation_chars;
-    }
-
-    if (num_punctuation_chars == 0)
-    {
-        return g_strdup (str);
-    }
-
-    num_chars_left = str_len - num_punctuation_chars;
-
-    return g_utf8_substring (str, 0, num_chars_left);
-}
-
 /**
  * get_common_prefix_length:
  * @str_a: first string
diff --git a/eel/eel-string.h b/eel/eel-string.h
index 98cca7e9f..5df4b6b2a 100644
--- a/eel/eel-string.h
+++ b/eel/eel-string.h
@@ -57,14 +57,6 @@ char *   eel_str_strip_substring_and_after (const char    *str,
 char *   eel_str_replace_substring         (const char    *str,
 					    const char    *substring,
 					    const char    *replacement);
-/**
- * eel_str_rtrim_punctuation:
- * @str: string
- *
- * Returns: a copy of str with trailing punctuation characters removed
- */
-char *   eel_str_rtrim_punctuation         (char *str);
-
 
 /**
  * eel_str_get_common_prefix:
diff --git a/src/nautilus-file-utilities.c b/src/nautilus-file-utilities.c
index 78d693eca..828b829aa 100644
--- a/src/nautilus-file-utilities.c
+++ b/src/nautilus-file-utilities.c
@@ -1315,26 +1315,67 @@ nautilus_get_common_filename_prefix (GList *file_list,
     return result;
 }
 
+static char *
+trim_whitespace (const gchar *string)
+{
+    glong space_count;
+    glong length;
+    gchar *offset;
+
+    space_count = 0;
+    length = g_utf8_strlen (string, -1);
+    offset = g_utf8_offset_to_pointer (string, length);
+
+    while (space_count <= length)
+    {
+        gunichar character;
+
+        offset = g_utf8_prev_char (offset);
+        character = g_utf8_get_char (offset);
+
+        if (!g_unichar_isspace (character))
+        {
+            break;
+        }
+
+        space_count++;
+    }
+
+    if (space_count == 0)
+    {
+        return g_strdup (string);
+    }
+
+    return g_utf8_substring (string, 0, length - space_count);
+}
+
 char *
 nautilus_get_common_filename_prefix_from_filenames (GList *filenames,
                                                     int    min_required_len)
 {
+    GList *stripped_filenames = NULL;
     char *common_prefix;
     char *truncated;
     int common_prefix_len;
 
-    common_prefix = eel_str_get_common_prefix (filenames, min_required_len);
+    for (GList *i = filenames; i != NULL; i = i->next)
+    {
+        gchar *stripped_filename;
+
+        stripped_filename = eel_filename_strip_extension (i->data);
 
+        stripped_filenames = g_list_prepend (stripped_filenames, stripped_filename);
+    }
+
+    common_prefix = eel_str_get_common_prefix (stripped_filenames, min_required_len);
     if (common_prefix == NULL)
     {
         return NULL;
     }
 
-    truncated = eel_filename_strip_extension (common_prefix);
-    g_free (common_prefix);
-    common_prefix = truncated;
+    g_list_free_full (stripped_filenames, g_free);
 
-    truncated = eel_str_rtrim_punctuation (common_prefix);
+    truncated = trim_whitespace (common_prefix);
     g_free (common_prefix);
 
     common_prefix_len = g_utf8_strlen (truncated, -1);
diff --git a/test/meson.build b/test/meson.build
index 0c7d830d5..f6ad5b007 100644
--- a/test/meson.build
+++ b/test/meson.build
@@ -16,10 +16,6 @@ test_file_utilities_get_common_filename_prefix = executable ('test-file-utilitie
                                                              'test-file-utilities-get-common-filename-prefix.c',
                                                              dependencies: libnautilus_dep)
 
-test_eel_string_rtrim_punctuation = executable ('test-eel-string-rtrim-punctuation',
-                                                'test-eel-string-rtrim-punctuation.c',
-                                                dependencies: libnautilus_dep)
-
 test_eel_string_get_common_prefix = executable ('test-eel-string-get-common-prefix',
                                                 'test-eel-string-get-common-prefix.c',
                                                 dependencies: libnautilus_dep)
@@ -27,5 +23,4 @@ test_eel_string_get_common_prefix = executable ('test-eel-string-get-common-pref
 test ('test-nautilus-search-engine', test_nautilus_search_engine)
 test ('test-nautilus-directory-async', test_nautilus_directory_async)
 test ('test-file-utilities-get-common-filename-prefix', test_file_utilities_get_common_filename_prefix)
-test ('test-eel-string-rtrim-punctuation', test_eel_string_rtrim_punctuation)
 test ('test-eel-string-get-common-prefix', test_eel_string_get_common_prefix)
diff --git a/test/test-eel-string-rtrim-punctuation.c b/test/test-eel-string-rtrim-punctuation.c
deleted file mode 100644
index 526ac9fb4..000000000
--- a/test/test-eel-string-rtrim-punctuation.c
+++ /dev/null
@@ -1,122 +0,0 @@
-#include <glib.h>
-
-#include "eel/eel-string.h"
-
-
-static void
-test_single_punctuation_character_removed ()
-{
-    char *actual;
-    actual = eel_str_rtrim_punctuation ("Yossarian-");
-    g_assert_cmpstr ("Yossarian", ==, actual);
-    g_free (actual);
-}
-
-static void
-test_tailing_space_is_removed ()
-{
-    char *actual;
-    actual = eel_str_rtrim_punctuation ("Yossarian ");
-    g_assert_cmpstr ("Yossarian", ==, actual);
-    g_free (actual);
-}
-
-static void
-test_multiple_punctuation_characters_removed ()
-{
-    char *actual;
-    actual = eel_str_rtrim_punctuation ("Yossarian-$$!£");
-    g_assert_cmpstr ("Yossarian", ==, actual);
-    g_free (actual);
-}
-
-static void
-test_multiple_punctuation_characters_removed_try_all_punctuation_characters ()
-{
-    char *actual;
-    actual = eel_str_rtrim_punctuation ("Yossarian-`¬!\"£$%^&*()_+-= {}[]:@~;'#<>?,./\\");
-    g_assert_cmpstr ("Yossarian", ==, actual);
-    g_free (actual);
-}
-
-static void
-test_punctuation_characters_removed_when_punctuation_in_middle_of_string ()
-{
-    char *actual;
-    actual = eel_str_rtrim_punctuation ("Yoss,,arian-$$!£");
-    g_assert_cmpstr ("Yoss,,arian", ==, actual);
-    g_free (actual);
-}
-
-static void
-test_punctuation_characters_removed_when_prefix_is_single_character ()
-{
-    char *actual;
-    actual = eel_str_rtrim_punctuation ("Y-$$  !£");
-    g_assert_cmpstr ("Y", ==, actual);
-    g_free (actual);
-}
-
-static void
-test_punctuation_characters_removed_when_unicode_characters_are_used ()
-{
-    char *actual;
-    actual = eel_str_rtrim_punctuation ("Y✺ǨǨǨ-$$  !£");
-    g_assert_cmpstr ("Y✺ǨǨǨ", ==, actual);
-    g_free (actual);
-}
-
-static void
-test_when_no_trailing_punctuation ()
-{
-    char *actual;
-    actual = eel_str_rtrim_punctuation ("Yoss,,arian");
-    g_assert_cmpstr ("Yoss,,arian", ==, actual);
-    g_free (actual);
-}
-
-static void
-test_when_single_character_and_no_trailing_punctuation ()
-{
-    char *actual;
-    actual = eel_str_rtrim_punctuation ("t");
-    g_assert_cmpstr ("t", ==, actual);
-    g_free (actual);
-}
-
-static void
-setup_test_suite ()
-{
-    g_test_add_func ("/rtrim-punctuation/1.0",
-                     test_single_punctuation_character_removed);
-    g_test_add_func ("/rtrim-punctuation/1.1",
-                     test_tailing_space_is_removed);
-    g_test_add_func ("/rtrim-punctuation/1.2",
-                     test_multiple_punctuation_characters_removed);
-    g_test_add_func ("/rtrim-punctuation/1.3",
-                     test_multiple_punctuation_characters_removed_try_all_punctuation_characters);
-    g_test_add_func ("/rtrim-punctuation/1.4",
-                     test_punctuation_characters_removed_when_punctuation_in_middle_of_string);
-    g_test_add_func ("/rtrim-punctuation/1.5",
-                     test_punctuation_characters_removed_when_prefix_is_single_character);
-    g_test_add_func ("/rtrim-punctuation/1.6",
-                     test_punctuation_characters_removed_when_unicode_characters_are_used);
-
-    g_test_add_func ("/rtrim-punctuation/2.0",
-                     test_when_no_trailing_punctuation);
-    g_test_add_func ("/rtrim-punctuation/2.1",
-                     test_when_single_character_and_no_trailing_punctuation);
-}
-
-int
-main (int   argc,
-      char *argv[])
-{
-    g_test_init (&argc, &argv, NULL);
-    g_test_bug_base ("http://bugzilla.gnome.org/show_bug.cgi?id=747907");
-    g_test_set_nonfatal_assertions ();
-
-    setup_test_suite ();
-
-    return g_test_run ();
-}