kreuzberg 4.0.0.rc2 → 4.0.1

data/vendor/kreuzberg-tesseract/src/result_iterator.rs

@@ -0,0 +1,286 @@
+use crate::api::TessDeleteText;
+use crate::enums::TessPageIteratorLevel;
+use crate::error::{Result, TesseractError};
+use std::ffi::CStr;
+use std::os::raw::{c_char, c_float, c_int, c_void};
+use std::sync::{Arc, Mutex};
+
+pub struct ResultIterator {
+    pub handle: Arc<Mutex<*mut c_void>>,
+}
+
+unsafe impl Send for ResultIterator {}
+unsafe impl Sync for ResultIterator {}
+
+impl ResultIterator {
+    /// Creates a new instance of the ResultIterator.
+    ///
+    /// # Arguments
+    ///
+    /// * `handle` - Pointer to the ResultIterator.
+    ///
+    /// # Returns
+    ///
+    /// Returns the new instance of the ResultIterator.
+    pub fn new(handle: *mut c_void) -> Self {
+        ResultIterator {
+            handle: Arc::new(Mutex::new(handle)),
+        }
+    }
+
+    /// Gets the UTF-8 text of the current iterator.
+    ///
+    /// # Arguments
+    ///
+    /// * `level` - Level of the text.
+    ///
+    /// # Returns
+    ///
+    /// Returns the UTF-8 text as a `String` if successful, otherwise returns an error.
+    pub fn get_utf8_text(&self, level: TessPageIteratorLevel) -> Result<String> {
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+        let text_ptr = unsafe { TessResultIteratorGetUTF8Text(*handle, level as c_int) };
+        if text_ptr.is_null() {
+            return Err(TesseractError::NullPointerError);
+        }
+        let c_str = unsafe { CStr::from_ptr(text_ptr) };
+        let result = c_str.to_str()?.to_owned();
+        unsafe { TessDeleteText(text_ptr as *mut c_char) };
+        Ok(result)
+    }
+
+    /// Gets the confidence of the current iterator.
+    ///
+    /// # Arguments
+    ///
+    /// * `level` - Level of the confidence.
+    ///
+    /// # Returns
+    ///
+    /// Returns the confidence as a `f32`.
+    pub fn confidence(&self, level: TessPageIteratorLevel) -> Result<f32> {
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+        Ok(unsafe { TessResultIteratorConfidence(*handle, level as c_int) })
+    }
+
+    /// Gets the recognition language of the current iterator.
+    ///
+    /// # Returns
+    ///
+    /// Returns the recognition language as a `String` if successful, otherwise returns an error.
+    pub fn word_recognition_language(&self) -> Result<String> {
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+        let lang_ptr = unsafe { TessResultIteratorWordRecognitionLanguage(*handle) };
+        if lang_ptr.is_null() {
+            return Err(TesseractError::NullPointerError);
+        }
+        let c_str = unsafe { CStr::from_ptr(lang_ptr) };
+        Ok(c_str.to_str()?.to_owned())
+    }
+
+    /// Gets the font attributes of the current iterator.
+    ///
+    /// # Returns
+    ///
+    /// Returns the font attributes as a tuple if successful, otherwise returns an error.
+    pub fn word_font_attributes(&self) -> Result<(bool, bool, bool, bool, bool, bool, i32, i32)> {
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+        let mut is_bold = 0;
+        let mut is_italic = 0;
+        let mut is_underlined = 0;
+        let mut is_monospace = 0;
+        let mut is_serif = 0;
+        let mut is_smallcaps = 0;
+        let mut pointsize = 0;
+        let mut font_id = 0;
+
+        let result = unsafe {
+            TessResultIteratorWordFontAttributes(
+                *handle,
+                &mut is_bold,
+                &mut is_italic,
+                &mut is_underlined,
+                &mut is_monospace,
+                &mut is_serif,
+                &mut is_smallcaps,
+                &mut pointsize,
+                &mut font_id,
+            )
+        };
+
+        if result == 0 {
+            Err(TesseractError::InvalidParameterError)
+        } else {
+            Ok((
+                is_bold != 0,
+                is_italic != 0,
+                is_underlined != 0,
+                is_monospace != 0,
+                is_serif != 0,
+                is_smallcaps != 0,
+                pointsize,
+                font_id,
+            ))
+        }
+    }
+
+    /// Checks if the current iterator is from the dictionary.
+    ///
+    /// # Returns
+    ///
+    /// Returns `true` if the current iterator is from the dictionary, otherwise returns `false`.
+    pub fn word_is_from_dictionary(&self) -> Result<bool> {
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+        Ok(unsafe { TessResultIteratorWordIsFromDictionary(*handle) != 0 })
+    }
+
+    /// Checks if the current iterator is numeric.
+    ///
+    /// # Returns
+    ///
+    /// Returns `true` if the current iterator is numeric, otherwise returns `false`.
+    pub fn word_is_numeric(&self) -> Result<bool> {
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+        Ok(unsafe { TessResultIteratorWordIsNumeric(*handle) != 0 })
+    }
+
+    /// Checks if the current iterator is superscript.
+    ///
+    /// # Returns
+    ///
+    /// Returns `true` if the current iterator is superscript, otherwise returns `false`.
+    pub fn symbol_is_superscript(&self) -> Result<bool> {
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+        Ok(unsafe { TessResultIteratorSymbolIsSuperscript(*handle) != 0 })
+    }
+
+    /// Checks if the current iterator is subscript.
+    ///
+    /// # Returns
+    ///
+    /// Returns `true` if the current iterator is subscript, otherwise returns `false`.
+    pub fn symbol_is_subscript(&self) -> Result<bool> {
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+        Ok(unsafe { TessResultIteratorSymbolIsSubscript(*handle) != 0 })
+    }
+
+    /// Checks if the current iterator is dropcap.
+    ///
+    /// # Returns
+    ///
+    /// Returns `true` if the current iterator is dropcap, otherwise returns `false`.
+    pub fn symbol_is_dropcap(&self) -> Result<bool> {
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+        Ok(unsafe { TessResultIteratorSymbolIsDropcap(*handle) != 0 })
+    }
+
+    /// Moves to the next iterator.
+    ///
+    /// # Arguments
+    ///
+    /// * `level` - Level of the next iterator.
+    ///
+    /// # Returns
+    ///
+    /// Returns `true` if the next iterator exists, otherwise returns `false`.
+    pub fn next(&self, level: TessPageIteratorLevel) -> Result<bool> {
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+        Ok(unsafe { TessResultIteratorNext(*handle, level as c_int) != 0 })
+    }
+
+    /// Gets the current word from the iterator with its bounding box and confidence.
+    ///
+    /// # Returns
+    ///
+    /// Returns a tuple of (text, left, top, right, bottom, confidence) if successful
+    pub fn get_word_with_bounds(&self) -> Result<(String, i32, i32, i32, i32, f32)> {
+        let text = self.get_utf8_text(TessPageIteratorLevel::RIL_WORD)?;
+        let (left, top, right, bottom) = self.get_bounding_box(TessPageIteratorLevel::RIL_WORD)?;
+        let confidence = self.confidence(TessPageIteratorLevel::RIL_WORD)?;
+
+        Ok((text, left, top, right, bottom, confidence))
+    }
+
+    /// Advances the iterator to the next word.
+    ///
+    /// # Returns
+    ///
+    /// Returns true if successful, false if there are no more words
+    pub fn next_word(&self) -> Result<bool> {
+        self.next(TessPageIteratorLevel::RIL_WORD)
+    }
+
+    /// Gets the word information for the current position in the iterator.
+    /// Should be called before next() to ensure valid data.
+    ///
+    /// # Returns
+    /// Returns a tuple of (text, left, top, right, bottom, confidence) if successful
+    pub fn get_current_word(&self) -> Result<(String, i32, i32, i32, i32, f32)> {
+        let text = self.get_utf8_text(TessPageIteratorLevel::RIL_WORD)?;
+        let (left, top, right, bottom) = self.get_bounding_box(TessPageIteratorLevel::RIL_WORD)?;
+        let confidence = self.confidence(TessPageIteratorLevel::RIL_WORD)?;
+
+        Ok((text, left, top, right, bottom, confidence))
+    }
+
+    /// Gets the bounding box for the current element.
+    pub fn get_bounding_box(&self, level: TessPageIteratorLevel) -> Result<(i32, i32, i32, i32)> {
+        let mut left = 0;
+        let mut top = 0;
+        let mut right = 0;
+        let mut bottom = 0;
+
+        let handle = self.handle.lock().map_err(|_| TesseractError::MutexLockError)?;
+
+        let result = unsafe {
+            TessPageIteratorBoundingBox(*handle, level as c_int, &mut left, &mut top, &mut right, &mut bottom)
+        };
+
+        if result == 0 {
+            Err(TesseractError::InvalidParameterError)
+        } else {
+            Ok((left, top, right, bottom))
+        }
+    }
+}
+
+impl Drop for ResultIterator {
+    fn drop(&mut self) {
+        if let Ok(handle) = self.handle.lock() {
+            unsafe { TessResultIteratorDelete(*handle) };
+        }
+    }
+}
+
+#[cfg(feature = "build-tesseract")]
+unsafe extern "C" {
+    pub fn TessResultIteratorDelete(handle: *mut c_void);
+    pub fn TessResultIteratorGetUTF8Text(handle: *mut c_void, level: c_int) -> *mut c_char;
+    pub fn TessResultIteratorConfidence(handle: *mut c_void, level: c_int) -> c_float;
+    pub fn TessResultIteratorWordRecognitionLanguage(handle: *mut c_void) -> *const c_char;
+    pub fn TessResultIteratorWordFontAttributes(
+        handle: *mut c_void,
+        is_bold: *mut c_int,
+        is_italic: *mut c_int,
+        is_underlined: *mut c_int,
+        is_monospace: *mut c_int,
+        is_serif: *mut c_int,
+        is_smallcaps: *mut c_int,
+        pointsize: *mut c_int,
+        font_id: *mut c_int,
+    ) -> c_int;
+    pub fn TessResultIteratorWordIsFromDictionary(handle: *mut c_void) -> c_int;
+    pub fn TessResultIteratorWordIsNumeric(handle: *mut c_void) -> c_int;
+    pub fn TessResultIteratorSymbolIsSuperscript(handle: *mut c_void) -> c_int;
+    pub fn TessResultIteratorSymbolIsSubscript(handle: *mut c_void) -> c_int;
+    pub fn TessResultIteratorSymbolIsDropcap(handle: *mut c_void) -> c_int;
+    pub fn TessResultIteratorNext(handle: *mut c_void, level: c_int) -> c_int;
+    pub fn TessPageIteratorBoundingBox(
+        handle: *mut c_void,
+        level: c_int,
+        left: *mut c_int,
+        top: *mut c_int,
+        right: *mut c_int,
+        bottom: *mut c_int,
+    ) -> c_int;
+}

data/vendor/kreuzberg-tesseract/src/result_renderer.rs

@@ -0,0 +1,183 @@
+use crate::TesseractAPI;
+use crate::error::{Result, TesseractError};
+use std::ffi::{CStr, CString};
+use std::os::raw::{c_char, c_int, c_void};
+use std::sync::Arc;
+use std::sync::Mutex;
+
+pub struct TessResultRenderer {
+    handle: Arc<Mutex<*mut c_void>>,
+}
+
+unsafe impl Send for TessResultRenderer {}
+unsafe impl Sync for TessResultRenderer {}
+
+impl TessResultRenderer {
+    /// Creates a new instance of the TessResultRenderer.
+    ///
+    /// # Arguments
+    ///
+    /// * `outputbase` - Output base path.
+    ///
+    /// # Returns
+    ///
+    /// Returns the new instance of the TessResultRenderer.
+    pub fn new_text_renderer(outputbase: &str) -> Result<Self> {
+        let outputbase = CString::new(outputbase).unwrap();
+        let handle = unsafe { TessTextRendererCreate(outputbase.as_ptr()) };
+        if handle.is_null() {
+            Err(TesseractError::NullPointerError)
+        } else {
+            Ok(TessResultRenderer {
+                handle: Arc::new(Mutex::new(handle)),
+            })
+        }
+    }
+
+    /// Creates a new instance of the TessResultRenderer for HOCR.
+    ///
+    /// # Arguments
+    ///
+    /// * `outputbase` - Output base path.
+    ///
+    /// # Returns
+    ///
+    /// Returns the new instance of the TessResultRenderer.
+    pub fn new_hocr_renderer(outputbase: &str) -> Result<Self> {
+        let outputbase = CString::new(outputbase).unwrap();
+        let handle = unsafe { TessHOcrRendererCreate(outputbase.as_ptr()) };
+        if handle.is_null() {
+            Err(TesseractError::NullPointerError)
+        } else {
+            Ok(TessResultRenderer {
+                handle: Arc::new(Mutex::new(handle)),
+            })
+        }
+    }
+
+    /// Creates a new instance of the TessResultRenderer for PDF.
+    ///
+    /// # Arguments
+    ///
+    /// * `outputbase` - Output base path.
+    /// * `datadir` - Data directory path.
+    /// * `textonly` - Whether to include text only.
+    ///
+    /// # Returns
+    ///
+    /// Returns the new instance of the TessResultRenderer.
+    pub fn new_pdf_renderer(outputbase: &str, datadir: &str, textonly: bool) -> Result<Self> {
+        let outputbase = CString::new(outputbase).unwrap();
+        let datadir = CString::new(datadir).unwrap();
+        let handle = unsafe { TessPDFRendererCreate(outputbase.as_ptr(), datadir.as_ptr(), textonly as c_int) };
+        if handle.is_null() {
+            Err(TesseractError::NullPointerError)
+        } else {
+            Ok(TessResultRenderer {
+                handle: Arc::new(Mutex::new(handle)),
+            })
+        }
+    }
+
+    /// Begins a new document.
+    ///
+    /// # Arguments
+    ///
+    /// * `title` - Title of the document.
+    ///
+    /// # Returns
+    ///
+    /// Returns `true` if the document was created successfully, otherwise returns `false`.
+    pub fn begin_document(&self, title: &str) -> bool {
+        let title = CString::new(title).unwrap();
+        let handle = self.handle.lock().unwrap();
+        unsafe { TessResultRendererBeginDocument(*handle, title.as_ptr()) != 0 }
+    }
+
+    /// Adds an image to the document.
+    ///
+    /// # Arguments
+    ///
+    /// * `api` - The TesseractAPI instance.
+    ///
+    /// # Returns
+    ///
+    /// Returns `true` if the image was added successfully, otherwise returns `false`.
+    pub fn add_image(&self, api: &TesseractAPI) -> bool {
+        let api_handle = api.handle.lock().unwrap();
+        let handle = self.handle.lock().unwrap();
+        unsafe { TessResultRendererAddImage(*handle, *api_handle) != 0 }
+    }
+
+    /// Ends the document.
+    ///
+    /// # Returns
+    ///
+    /// Returns `true` if the document was ended successfully, otherwise returns `false`.
+    pub fn end_document(&self) -> bool {
+        let handle = self.handle.lock().unwrap();
+        unsafe { TessResultRendererEndDocument(*handle) != 0 }
+    }
+
+    /// Gets the extension of the document.
+    ///
+    /// # Returns
+    ///
+    /// Returns the extension as a `String` if successful, otherwise returns an error.
+    pub fn get_extension(&self) -> Result<String> {
+        let handle = self.handle.lock().unwrap();
+        let ext_ptr = unsafe { TessResultRendererExtention(*handle) };
+        if ext_ptr.is_null() {
+            Err(TesseractError::NullPointerError)
+        } else {
+            let c_str = unsafe { CStr::from_ptr(ext_ptr) };
+            Ok(c_str.to_str()?.to_owned())
+        }
+    }
+
+    /// Gets the title of the document.
+    ///
+    /// # Returns
+    ///
+    /// Returns the title as a `String` if successful, otherwise returns an error.
+    pub fn get_title(&self) -> Result<String> {
+        let handle = self.handle.lock().unwrap();
+        let title_ptr = unsafe { TessResultRendererTitle(*handle) };
+        if title_ptr.is_null() {
+            Err(TesseractError::NullPointerError)
+        } else {
+            let c_str = unsafe { CStr::from_ptr(title_ptr) };
+            Ok(c_str.to_str()?.to_owned())
+        }
+    }
+
+    /// Gets the number of images in the document.
+    ///
+    /// # Returns
+    ///
+    /// Returns the number of images as an `i32`.
+    pub fn get_image_num(&self) -> i32 {
+        let handle = self.handle.lock().unwrap();
+        unsafe { TessResultRendererImageNum(*handle) }
+    }
+}
+
+impl Drop for TessResultRenderer {
+    fn drop(&mut self) {
+        let handle = self.handle.lock().unwrap();
+        unsafe { TessDeleteResultRenderer(*handle) };
+    }
+}
+
+unsafe extern "C" {
+    pub fn TessTextRendererCreate(outputbase: *const c_char) -> *mut c_void;
+    pub fn TessHOcrRendererCreate(outputbase: *const c_char) -> *mut c_void;
+    pub fn TessPDFRendererCreate(outputbase: *const c_char, datadir: *const c_char, textonly: c_int) -> *mut c_void;
+    pub fn TessDeleteResultRenderer(renderer: *mut c_void);
+    pub fn TessResultRendererBeginDocument(renderer: *mut c_void, title: *const c_char) -> c_int;
+    pub fn TessResultRendererAddImage(renderer: *mut c_void, api: *mut c_void) -> c_int;
+    pub fn TessResultRendererEndDocument(renderer: *mut c_void) -> c_int;
+    pub fn TessResultRendererExtention(renderer: *mut c_void) -> *const c_char;
+    pub fn TessResultRendererTitle(renderer: *mut c_void) -> *const c_char;
+    pub fn TessResultRendererImageNum(renderer: *mut c_void) -> c_int;
+}

data/vendor/kreuzberg-tesseract/tests/integration_test.rs

@@ -0,0 +1,211 @@
+use kreuzberg_tesseract::TesseractAPI;
+use std::path::{Path, PathBuf};
+
+fn get_default_tessdata_dir() -> PathBuf {
+    if cfg!(target_os = "macos") {
+        let home_dir = std::env::var("HOME").expect("HOME environment variable not set");
+        PathBuf::from(home_dir)
+            .join("Library")
+            .join("Application Support")
+            .join("tesseract-rs")
+            .join("tessdata")
+    } else if cfg!(target_os = "linux") {
+        let system_paths = [
+            PathBuf::from("/usr/share/tesseract-ocr/5/tessdata"),
+            PathBuf::from("/usr/share/tesseract-ocr/tessdata"),
+        ];
+        for path in &system_paths {
+            if path.exists() {
+                return path.clone();
+            }
+        }
+        let home_dir = std::env::var("HOME").expect("HOME environment variable not set");
+        PathBuf::from(home_dir).join(".tesseract-rs").join("tessdata")
+    } else if cfg!(target_os = "windows") {
+        PathBuf::from(std::env::var("APPDATA").expect("APPDATA environment variable not set"))
+            .join("tesseract-rs")
+            .join("tessdata")
+    } else {
+        panic!("Unsupported operating system");
+    }
+}
+
+fn get_tessdata_dir() -> PathBuf {
+    match std::env::var("TESSDATA_PREFIX") {
+        Ok(dir) => {
+            let prefix_path = PathBuf::from(dir);
+            let tessdata_path = if prefix_path.ends_with("tessdata") {
+                prefix_path
+            } else {
+                prefix_path.join("tessdata")
+            };
+            println!("Using TESSDATA_PREFIX directory: {:?}", tessdata_path);
+            tessdata_path
+        }
+        Err(_) => {
+            let default_dir = get_default_tessdata_dir();
+            println!("TESSDATA_PREFIX not set, using default directory: {:?}", default_dir);
+            default_dir
+        }
+    }
+}
+
+fn ensure_eng_traineddata_exists(tessdata_dir: &Path) {
+    let eng_traineddata = tessdata_dir.join("eng.traineddata");
+    assert!(
+        eng_traineddata.exists(),
+        "eng.traineddata not found in {}. Set TESSDATA_PREFIX or install English tessdata.",
+        tessdata_dir.display()
+    );
+}
+
+fn repo_root() -> PathBuf {
+    PathBuf::from(env!("CARGO_MANIFEST_DIR")).join("..").join("..")
+}
+
+fn load_test_image(relative: &str) -> Result<(Vec<u8>, u32, u32), Box<dyn std::error::Error>> {
+    let mut path = repo_root();
+    path.push("test_documents");
+    path.push(relative);
+
+    let img = image::open(&path)
+        .map_err(|e| format!("Failed to open test image {}: {}", path.display(), e))?
+        .to_rgb8();
+    let (width, height) = img.dimensions();
+    Ok((img.into_raw(), width, height))
+}
+
+#[test]
+fn test_ocr_on_hello_world_image() {
+    let tessdata_dir = get_tessdata_dir();
+    ensure_eng_traineddata_exists(&tessdata_dir);
+
+    let api = TesseractAPI::new();
+    api.init(tessdata_dir.to_str().unwrap(), "eng")
+        .expect("Failed to initialize Tesseract");
+
+    let (image_data, width, height) =
+        load_test_image("images/test_hello_world.png").expect("Failed to load test image");
+    api.set_image(&image_data, width as i32, height as i32, 3, 3 * width as i32)
+        .expect("Failed to set image");
+
+    let text = api.get_utf8_text().expect("Failed to perform OCR");
+    assert!(
+        text.to_lowercase().contains("hello"),
+        "Text does not contain expected word. Found: {}",
+        text
+    );
+}
+
+#[test]
+fn test_ocr_on_table_image() {
+    let tessdata_dir = get_tessdata_dir();
+    ensure_eng_traineddata_exists(&tessdata_dir);
+
+    let api = TesseractAPI::new();
+    api.init(tessdata_dir.to_str().unwrap(), "eng")
+        .expect("Failed to initialize Tesseract");
+    api.set_variable("tessedit_pageseg_mode", "1")
+        .expect("Failed to set PSM");
+
+    let (image_data, width, height) = load_test_image("tables/simple_table.png").expect("Failed to load test image");
+    api.set_image(&image_data, width as i32, height as i32, 3, 3 * width as i32)
+        .expect("Failed to set image");
+
+    let text = api.get_utf8_text().expect("Failed to perform OCR");
+    let lowercase = text.to_lowercase();
+    assert!(
+        lowercase.contains("product") && lowercase.contains("price"),
+        "Table text missing expected words. Found: {}",
+        text
+    );
+}
+
+#[test]
+fn test_invalid_language_code() {
+    let tessdata_dir = get_tessdata_dir();
+    ensure_eng_traineddata_exists(&tessdata_dir);
+
+    let api = TesseractAPI::new();
+
+    let result = api.init(tessdata_dir.to_str().unwrap(), "invalid_lang");
+    assert!(result.is_err());
+}
+
+#[test]
+fn test_empty_image_data() {
+    let tessdata_dir = get_tessdata_dir();
+    ensure_eng_traineddata_exists(&tessdata_dir);
+
+    let api = TesseractAPI::new();
+    api.init(tessdata_dir.to_str().unwrap(), "eng")
+        .expect("Failed to initialize Tesseract");
+
+    let empty_data: Vec<u8> = Vec::new();
+    let res = api.set_image(&empty_data, 100, 100, 3, 300);
+    assert!(res.is_err());
+}
+
+#[test]
+fn test_invalid_image_parameters() {
+    let tessdata_dir = get_tessdata_dir();
+    ensure_eng_traineddata_exists(&tessdata_dir);
+
+    let api = TesseractAPI::new();
+    api.init(tessdata_dir.to_str().unwrap(), "eng")
+        .expect("Failed to initialize Tesseract");
+
+    let (image_data, width, height) =
+        load_test_image("images/test_hello_world.png").expect("Failed to load test image");
+
+    let res = api.set_image(&image_data, -1, height as i32, 3, 3 * width as i32);
+    assert!(res.is_err());
+
+    let res = api.set_image(&image_data, width as i32, 0, 3, 3 * width as i32);
+    assert!(res.is_err());
+
+    let res = api.set_image(&image_data, width as i32, height as i32, 0, 3 * width as i32);
+    assert!(res.is_err());
+
+    let res = api.set_image(&image_data, width as i32, height as i32, 3, width as i32);
+    assert!(res.is_err());
+}
+
+#[test]
+fn test_variable_setting() {
+    let tessdata_dir = get_tessdata_dir();
+    ensure_eng_traineddata_exists(&tessdata_dir);
+
+    let api = TesseractAPI::new();
+    api.init(tessdata_dir.to_str().unwrap(), "eng")
+        .expect("Failed to initialize Tesseract");
+
+    let res = api.set_variable("invalid_variable_name", "1");
+    assert!(res.is_err());
+
+    let res = api.set_variable("tessedit_char_whitelist", "");
+    assert!(res.is_ok());
+
+    assert!(api.set_variable("tessedit_pageseg_mode", "1").is_ok());
+    assert!(api.set_variable("tessedit_ocr_engine_mode", "1").is_ok());
+}
+
+#[test]
+fn test_multiple_operations() {
+    let tessdata_dir = get_tessdata_dir();
+    ensure_eng_traineddata_exists(&tessdata_dir);
+
+    let api = TesseractAPI::new();
+    api.init(tessdata_dir.to_str().unwrap(), "eng")
+        .expect("Failed to initialize Tesseract");
+
+    let (image_data, width, height) =
+        load_test_image("images/test_hello_world.png").expect("Failed to load test image");
+
+    for _ in 0..3 {
+        let res = api.set_image(&image_data, width as i32, height as i32, 3, 3 * width as i32);
+        assert!(res.is_ok());
+        let text = api.get_utf8_text().expect("Failed to perform OCR");
+        assert!(!text.is_empty());
+    }
+}