Updated readme + sanitize subtitle file names before matching, ref #699

src/libs/core/impl/Path.cpp

@@ -86,4 +86,24 @@ namespace lms::core::pathUtils
 
         return longestCommonPath;
     }
+
+    std::string sanitizeFileStem(const std::string_view fileStem)
+    {
+        // Keep UTF8-encoded characters, but skip illegal ASCII characters
+        constexpr std::array<unsigned char, 9> illegalChars{ '/', '\\', ':', '*', '?', '"', '<', '>', '|' };
+        static_assert(std::all_of(std::begin(illegalChars), std::end(illegalChars), [](unsigned char c) { return c < 128; }), "Illegal characters must be ASCII");
+
+        std::string sanitized;
+        sanitized.reserve(fileStem.size());
+
+        for (const char c : fileStem)
+        {
+            if (std::any_of(std::begin(illegalChars), std::end(illegalChars), [c](char illegalChar) { return c == illegalChar; }))
+                continue;
+
+            sanitized.push_back(c);
+        }
+
+        return sanitized;
+    }
 } // namespace lms::core::pathUtils

src/libs/core/include/core/Path.hpp

@@ -21,6 +21,8 @@
 
 #include <filesystem>
 #include <span>
+#include <string>
+#include <string_view>
 
 #include <Wt/WDateTime.h>
 
@@ -54,4 +56,7 @@ namespace lms::core::pathUtils
 
         return longestCommonPath;
     }
+
+    // A method that sanitize a file stem, removing any illegal chars
+    std::string sanitizeFileStem(std::string_view fileStem);
 } // namespace lms::core::pathUtils

src/libs/core/test/Path.cpp

@@ -103,4 +103,34 @@ namespace lms::core::pathUtils::tests
             EXPECT_EQ(core::pathUtils::isPathInRootPath(test.path, test.rootPath), test.expectedResult) << "Failed: path = " << test.path << ", rootPath = " << test.rootPath;
         }
     }
+
+    TEST(Path, sanitizeFileStem)
+    {
+        struct TestCase
+        {
+            std::string input;
+            std::string_view expectedOutput;
+        };
+
+        TestCase tests[]{
+            { "", "" }, // empty input
+            { "valid_file_name", "valid_file_name" },
+            { "invalid:file*name?", "invalidfilename" },
+            { "another|invalid<name>", "anotherinvalidname" },
+            { "/leading/slash", "leadingslash" },
+            { "\\backslash\\file", "backslashfile" },
+            { "file_with_äöüß", "file_with_äöüß" },                // keep German umlauts
+            { "file_with_éèêë", "file_with_éèêë" },                // keep French accents
+            { "héllo 漢字", "héllo 漢字" },                        // keep UTF8 characters
+            { "file_with_üñîçødë", "file_with_üñîçødë" },          // keep special characters
+            { "file_with_!@#$%^&*()_+", "file_with_!@#$%^&()_+" }, // remove special characters
+            { "file_with_", "file_with_" },                        // handle double dots
+            { "file.with.extension", "file.with.extension" },      // keep extensions
+        };
+
+        for (const TestCase& test : tests)
+        {
+            EXPECT_EQ(core::pathUtils::sanitizeFileStem(test.input), test.expectedOutput) << "Failed: input = " << test.input;
+        }
+    }
 } // namespace lms::core::pathUtils::tests

src/libs/services/scanner/impl/steps/ScanStepAssociateArtistImages.cpp

@@ -73,7 +73,7 @@ namespace lms::scanner
                 {
                     db::Image::FindParameters params;
                     params.setDirectory(directory->getId());
-                    params.setFileStem(fileStem);
+                    params.setFileStem(fileStem); // no need to sanitize here, user is responsible for providing sanitized file stems in conf file
 
                     db::Image::find(session, params, [&](const db::Image::pointer foundImg) {
                         if (!image)
@@ -135,13 +135,11 @@ namespace lms::scanner
             {
                 // Expect layout like this:
                 // ReleaseArtist/Release/Tracks'
-                //              /artist.jpg
-                //              /someOtherUserConfiguredArtistFile.jpg
+                //              /someUserConfiguredArtistFile.jpg
                 //
                 // Or:
                 // ReleaseArtist/SomeGrouping/Release/Tracks'
-                //              /artist.jpg
-                //              /someOtherUserConfiguredArtistFile.jpg
+                //              /someUserConfiguredArtistFile.jpg
                 //
                 std::filesystem::path directoryToInspect{ core::pathUtils::getLongestCommonPath(std::cbegin(releasePaths), std::cend(releasePaths)) };
                 while (true)

src/libs/services/scanner/impl/steps/ScanStepAssociateMediumImages.cpp

@@ -28,6 +28,7 @@
 #include "core/IConfig.hpp"
 #include "core/IJob.hpp"
 #include "core/ILogger.hpp"
+#include "core/Path.hpp"
 #include "database/IDb.hpp"
 #include "database/Session.hpp"
 #include "database/Types.hpp"
@@ -105,8 +106,11 @@ namespace lms::scanner
                 if (image)
                     return;
 
-                if (const std::string_view mediumName{ medium->getName() }; !mediumName.empty())
-                    image = findImageInDirectory(session, directory, std::span{ &mediumName, 1 });
+                if (const std::string mediumName{ core::pathUtils::sanitizeFileStem(medium->getName()) }; !mediumName.empty())
+                {
+                    std::string_view mediumNameView{ mediumName };
+                    image = findImageInDirectory(session, directory, std::span{ &mediumNameView, 1 });
+                }
 
                 if (!image)
                     image = findImageInDirectory(session, directory, searchParams.mediumFileNames);