Simplify, fix and improve similar images algorithm (#983)

* Random changer * CD * A1 * Almost? * Nein * Heif * Tests that needs to be fixed * Fixed test * Tests * Fixed image counting in reference folders * Lock * Catch possible more bugs in pdf * Find ever more bugs
qarmin · Jun 9, 2023 · 55b2744 · 55b2744
1 parent 04a91ae
commit 55b2744
Show file tree

Hide file tree

Showing 12 changed files with 736 additions and 657 deletions.
diff --git a/Cargo.lock b/Cargo.lock
diff --git a/Changelog.md b/Changelog.md
@@ -1,3 +1,15 @@
+## Version 5.2.0 - ?
+- Add finding similar audio files by content - [#970](https://github.com/qarmin/czkawka/pull/970)
+- Allow to find duplicates by name/size at once - [#956](https://github.com/qarmin/czkawka/pull/956)
+- Fixed bug when cache for music tags not worked - [#970](https://github.com/qarmin/czkawka/pull/970)
+- Allow to set number of threads from CLI - [#972](https://github.com/qarmin/czkawka/pull/972)
+- Fix problem with invalid item sorting in bad extensions mode - [#972](https://github.com/qarmin/czkawka/pull/972)
+- Big refactor/cleaning of code - [#956](https://github.com/qarmin/czkawka/pull/956)/[#970](https://github.com/qarmin/czkawka/pull/970)/[#972](https://github.com/qarmin/czkawka/pull/972)
+- Use builtin gtk webp loader for previews - [#923](https://github.com/qarmin/czkawka/pull/923)
+- Fixed docker build - [#947](https://github.com/qarmin/czkawka/pull/947)
+- Restore snap builds broken since GTk 4 port - [#965](https://github.com/qarmin/czkawka/pull/947)
+- Instruction how to build native ARM64 binaries on Mac - [#945](https://github.com/qarmin/czkawka/pull/945)/[#971](https://github.com/qarmin/czkawka/pull/971)
+
 ## Version 5.1.0 - 19.02.2023r
 - Added sort button - [#894](https://github.com/qarmin/czkawka/pull/894)
 - Allow to set number of thread used to scan - [#839](https://github.com/qarmin/czkawka/pull/839)

diff --git a/czkawka_cli/Cargo.toml b/czkawka_cli/Cargo.toml
@@ -10,10 +10,10 @@ homepage = "https://github.com/qarmin/czkawka"
 repository = "https://github.com/qarmin/czkawka"
 
 [dependencies]
-clap = { version = "4.2", features = ["derive"] }
+clap = { version = "4.3", features = ["derive"] }
 
 # For enum types
-image_hasher = "1.1"
+image_hasher = "1.2"
 
 [dependencies.czkawka_core]
 path = "../czkawka_core"

diff --git a/czkawka_core/Cargo.toml b/czkawka_core/Cargo.toml
@@ -19,14 +19,14 @@ crossbeam-channel = "0.5"
 directories-next = "2.0"
 
 # Needed by similar images
-image_hasher = "1.1"
+image_hasher = "1.2"
 bk-tree = "0.5"
 image = "0.24"
 hamming = "0.1"
 
 # Needed by same music
-bitflags = "2.2"
-lofty = "0.12"
+bitflags = "2.3"
+lofty = "0.14"
 
 # Futures - needed by async progress sender
 futures = "0.3.28"
@@ -41,11 +41,11 @@ rusty-chromaprint = "0.1"
 symphonia = { version = "0.5", features = ["all"] }
 
 # Hashes for duplicate files
-blake3 = "1.3"
+blake3 = "1.4"
 crc32fast = "1.3"
 xxhash-rust = { version = "0.8", features = ["xxh3"] }
 
-tempfile = "3.5"
+tempfile = "3.6"
 
 # Video Duplicates
 vid_dup_finder_lib = "0.1"
@@ -59,8 +59,8 @@ serde_json = "1.0"
 # Language
 i18n-embed = { version = "0.13", features = ["fluent-system", "desktop-requester"] }
 i18n-embed-fl = "0.6"
-rust-embed = "6.6"
-once_cell = "1.17"
+rust-embed = "6.7"
+once_cell = "1.18"
 
 # Raw image files
 rawloader = "0.37"
@@ -73,11 +73,12 @@ infer = "0.13"
 num_cpus = "1.15"
 
 # Heif/Heic
-libheif-rs = { version = "0.18.0", optional = true } # Do not upgrade now, since Ubuntu 22.04 not works with newer version
+libheif-rs = { version = "=0.18.0", optional = true } # Do not upgrade now, since Ubuntu 22.04 not works with newer version
+libheif-sys = { version = "=1.14.2", optional = true } # 1.14.3 brake compilation on Ubuntu 22.04
 anyhow = { version = "1.0" }
 
-state = "0.5"
+state = "0.6"
 
 [features]
 default = []
-heif = ["dep:libheif-rs"]
+heif = ["dep:libheif-rs", "dep:libheif-sys"]
diff --git a/czkawka_core/src/broken_files.rs b/czkawka_core/src/broken_files.rs
@@ -372,23 +372,27 @@ impl BrokenFiles {
 
         let mut file_entry_clone = file_entry.clone();
         let result = panic::catch_unwind(|| {
-            if let Err(e) = FileOptions::cached().parse_options(parser_options).open(&file_entry.path) {
-                if let PdfError::Io { .. } = e {
-                    return None;
-                }
-
-                let mut error_string = e.to_string();
-                // Workaround for strange error message https://github.com/qarmin/czkawka/issues/898
-                if error_string.starts_with("Try at") {
-                    if let Some(start_index) = error_string.find("/pdf-") {
-                        error_string = format!("Decoding error in pdf-rs library - {}", &error_string[start_index..]);
+            match FileOptions::cached().parse_options(parser_options).open(&file_entry.path) {
+                Ok(file) => {
+                    for idx in 0..file.num_pages() {
+                        if let Err(e) = file.get_page(idx) {
+                            let err = validate_pdf_error(&mut file_entry, e);
+                            if let PdfError::InvalidPassword = err {
+                                return None;
+                            } else {
+                                break;
+                            }
+                        }
                     }
                 }
-
-                file_entry.error_string = error_string;
-                let error = unpack_pdf_error(e);
-                if let PdfError::InvalidPassword = error {
-                    return None;
+                Err(e) => {
+                    if let PdfError::Io { .. } = e {
+                        return None;
+                    }
+                    let err = validate_pdf_error(&mut file_entry, e);
+                    if let PdfError::InvalidPassword = err {
+                        return None;
+                    }
                 }
             }
             Some(file_entry)
@@ -708,3 +712,16 @@ fn unpack_pdf_error(e: PdfError) -> PdfError {
         e
     }
 }
+
+fn validate_pdf_error(file_entry: &mut FileEntry, e: PdfError) -> PdfError {
+    let mut error_string = e.to_string();
+    // Workaround for strange error message https://github.com/qarmin/czkawka/issues/898
+    if error_string.starts_with("Try at") {
+        if let Some(start_index) = error_string.find("/pdf-") {
+            error_string = format!("Decoding error in pdf-rs library - {}", &error_string[start_index..]);
+        }
+    }
+
+    file_entry.error_string = error_string;
+    unpack_pdf_error(e)
+}
diff --git a/czkawka_core/src/common.rs b/czkawka_core/src/common.rs
@@ -24,7 +24,7 @@ use crate::common_directory::Directories;
 use crate::common_items::ExcludedItems;
 use crate::common_traits::ResultEntry;
 
-static NUMBER_OF_THREADS: state::Storage<usize> = state::Storage::new();
+static NUMBER_OF_THREADS: state::InitCell<usize> = state::InitCell::new();
 
 pub fn get_number_of_threads() -> usize {
     let data = NUMBER_OF_THREADS.get();

diff --git a/czkawka_core/src/common_dir_traversal.rs b/czkawka_core/src/common_dir_traversal.rs
@@ -473,7 +473,7 @@ where
                         }
                     }
                     if counter > 0 {
-                        // Do not increase counter one by one in threads, because usually it
+                        // Increase counter in batch, because usually it may be slow to add multiple times atomic value
                         atomic_counter.fetch_add(counter, Ordering::Relaxed);
                     }
                     (dir_result, warnings, fe_result, set_as_not_empty_folder_list, folder_entries_list)

diff --git a/czkawka_core/src/common_directory.rs b/czkawka_core/src/common_directory.rs
@@ -13,9 +13,9 @@ pub struct Directories {
     pub excluded_directories: Vec<PathBuf>,
     pub included_directories: Vec<PathBuf>,
     pub reference_directories: Vec<PathBuf>,
-    exclude_other_filesystems: Option<bool>,
+    pub exclude_other_filesystems: Option<bool>,
     #[cfg(target_family = "unix")]
-    included_dev_ids: Vec<u64>,
+    pub included_dev_ids: Vec<u64>,
 }
 
 impl Directories {