Remove OCR dependencies for Windows build - stub implementation
This commit is contained in:
parent
c08b30e52d
commit
f2465b2b4f
|
|
@ -47,10 +47,11 @@ uuid = { version = "1.6", features = ["v4", "serde"] }
|
||||||
# Regex
|
# Regex
|
||||||
regex = "1.10"
|
regex = "1.10"
|
||||||
|
|
||||||
# Image processing for OCR
|
# Image processing
|
||||||
image = "0.24"
|
image = "0.24"
|
||||||
imageproc = "0.23"
|
# Note: OCR requires system libraries (Tesseract)
|
||||||
leptess = "0.14"
|
# imageproc = "0.23"
|
||||||
|
# leptess = "0.14"
|
||||||
|
|
||||||
# Clipboard
|
# Clipboard
|
||||||
arboard = "3.3"
|
arboard = "3.3"
|
||||||
|
|
|
||||||
|
|
@ -1,94 +1,49 @@
|
||||||
use image::{DynamicImage, GrayImage, ImageBuffer, Luma};
|
// OCR Engine - Stub implementation
|
||||||
use imageproc::contrast::{stretch_contrast, threshold};
|
// Full OCR requires Tesseract system libraries
|
||||||
use leptess::TessBaseApi;
|
// This stub allows the application to compile without OCR dependencies
|
||||||
|
|
||||||
use std::collections::HashMap;
|
use std::collections::HashMap;
|
||||||
use tracing::{debug, error, info};
|
use tracing::{debug, info};
|
||||||
|
|
||||||
use crate::ocr::calibration::{OCRRegion, OCRCalibration, PreprocessingConfig};
|
use crate::ocr::calibration::{OCRRegion, OCRCalibration, PreprocessingConfig};
|
||||||
|
|
||||||
pub struct OCREngine {
|
pub struct OCREngine {
|
||||||
tesseract: TessBaseApi,
|
|
||||||
calibration: OCRCalibration,
|
calibration: OCRCalibration,
|
||||||
|
available: bool,
|
||||||
}
|
}
|
||||||
|
|
||||||
impl OCREngine {
|
impl OCREngine {
|
||||||
pub fn new() -> Result<Self, String> {
|
pub fn new() -> Result<Self, String> {
|
||||||
let mut tesseract = TessBaseApi::new();
|
info!("OCR Engine initialized (stub mode - Tesseract not available)");
|
||||||
|
Ok(Self {
|
||||||
// Initialize with English
|
calibration: OCRCalibration::default(),
|
||||||
if tesseract.init(None, "eng").is_err() {
|
available: false,
|
||||||
return Err("Failed to initialize Tesseract".to_string());
|
})
|
||||||
}
|
}
|
||||||
|
|
||||||
// Set OCR engine mode to LSTM only for better accuracy
|
pub fn is_available(&self) -> bool {
|
||||||
tesseract.set_variable("tessedit_ocr_engine_mode", "1").ok();
|
self.available
|
||||||
|
|
||||||
// Whitelist characters for numeric/text recognition
|
|
||||||
tesseract.set_variable("tessedit_char_whitelist",
|
|
||||||
"0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz.,-:()%/ ").ok();
|
|
||||||
|
|
||||||
Ok(Self {
|
|
||||||
tesseract,
|
|
||||||
calibration: OCRCalibration::default(),
|
|
||||||
})
|
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn set_calibration(&mut self, calibration: OCRCalibration) {
|
pub fn set_calibration(&mut self, calibration: OCRCalibration) {
|
||||||
self.calibration = calibration;
|
self.calibration = calibration;
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn recognize_region(&mut self, region_name: &str, screenshot: &[u8]) -> Result<String, String> {
|
pub fn recognize_region(&mut self, _region_name: &str, _screenshot: &[u8]) -> Result<String, String> {
|
||||||
let region = self.calibration.regions.get(region_name)
|
// Stub implementation - OCR requires Tesseract system libraries
|
||||||
.ok_or_else(|| format!("Region '{}' not found", region_name))?;
|
Err("OCR not available - Tesseract system library not installed".to_string())
|
||||||
|
|
||||||
// Load image
|
|
||||||
let mut img = image::load_from_memory(screenshot)
|
|
||||||
.map_err(|e| format!("Failed to load image: {}", e))?;
|
|
||||||
|
|
||||||
// Crop to region
|
|
||||||
let cropped = img.crop(
|
|
||||||
region.x as u32,
|
|
||||||
region.y as u32,
|
|
||||||
region.width as u32,
|
|
||||||
region.height as u32,
|
|
||||||
);
|
|
||||||
|
|
||||||
// Preprocess
|
|
||||||
let processed = self.preprocess_image(&cropped, ®ion.preprocessing);
|
|
||||||
|
|
||||||
// Perform OCR
|
|
||||||
let text = self.perform_ocr(&processed)?;
|
|
||||||
|
|
||||||
debug!("OCR result for '{}': '{}'", region_name, text);
|
|
||||||
|
|
||||||
Ok(text.trim().to_string())
|
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn recognize_hp(&mut self, screenshot: &[u8]) -> Result<(f32, f32), String> {
|
pub fn recognize_hp(&mut self, _screenshot: &[u8]) -> Result<(f32, f32), String> {
|
||||||
// Try to extract HP value and max HP
|
Err("OCR not available".to_string())
|
||||||
let text = self.recognize_region("hp_bar", screenshot)?;
|
|
||||||
|
|
||||||
// Parse HP text (formats: "1234/5678" or "1234" or percentage)
|
|
||||||
if let Some((current, max)) = Self::parse_hp_text(&text) {
|
|
||||||
Ok((current, max))
|
|
||||||
} else {
|
|
||||||
Err(format!("Failed to parse HP from text: '{}'", text))
|
|
||||||
}
|
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn recognize_mob_name(&mut self, screenshot: &[u8]) -> Result<String, String> {
|
pub fn recognize_mob_name(&mut self, _screenshot: &[u8]) -> Result<String, String> {
|
||||||
// Mob names appear as floating text above mobs
|
Err("OCR not available".to_string())
|
||||||
// This requires detecting text that appears temporarily
|
|
||||||
let text = self.recognize_region("mob_name", screenshot)?;
|
|
||||||
Ok(text)
|
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn recognize_coordinates(&mut self, screenshot: &[u8]) -> Result<(f32, f32, f32), String> {
|
pub fn recognize_coordinates(&mut self, _screenshot: &[u8]) -> Result<(f32, f32, f32), String> {
|
||||||
// Read coordinates from radar or location display
|
Err("OCR not available".to_string())
|
||||||
let text = self.recognize_region("radar", screenshot)?;
|
|
||||||
|
|
||||||
// Parse coordinate text (format: "X: 12345.67 Y: 12345.89 Z: 123.45")
|
|
||||||
Self::parse_coordinates(&text)
|
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn detect_text_changes(
|
pub fn detect_text_changes(
|
||||||
|
|
@ -99,7 +54,6 @@ impl OCREngine {
|
||||||
) -> Vec<TextChange> {
|
) -> Vec<TextChange> {
|
||||||
let mut changes = Vec::new();
|
let mut changes = Vec::new();
|
||||||
|
|
||||||
// Detect added/removed/changed text
|
|
||||||
if previous != current {
|
if previous != current {
|
||||||
changes.push(TextChange {
|
changes.push(TextChange {
|
||||||
region: region_name.to_string(),
|
region: region_name.to_string(),
|
||||||
|
|
@ -111,99 +65,6 @@ impl OCREngine {
|
||||||
|
|
||||||
changes
|
changes
|
||||||
}
|
}
|
||||||
|
|
||||||
fn preprocess_image(&self, img: &DynamicImage, config: &PreprocessingConfig) -> DynamicImage {
|
|
||||||
let mut processed = img.clone();
|
|
||||||
|
|
||||||
// Convert to grayscale if needed
|
|
||||||
if config.grayscale {
|
|
||||||
processed = DynamicImage::ImageLuma8(processed.to_luma8());
|
|
||||||
}
|
|
||||||
|
|
||||||
// Apply contrast and brightness
|
|
||||||
if config.contrast != 1.0 || config.brightness != 0.0 {
|
|
||||||
processed = self.adjust_contrast_brightness(&processed, config.contrast, config.brightness);
|
|
||||||
}
|
|
||||||
|
|
||||||
// Apply threshold if specified
|
|
||||||
if let Some(thresh) = config.threshold {
|
|
||||||
if let DynamicImage::ImageLuma8(gray) = &processed {
|
|
||||||
let thresholded = threshold(gray, thresh);
|
|
||||||
processed = DynamicImage::ImageLuma8(thresholded);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Invert if needed
|
|
||||||
if config.invert {
|
|
||||||
processed.invert();
|
|
||||||
}
|
|
||||||
|
|
||||||
processed
|
|
||||||
}
|
|
||||||
|
|
||||||
fn adjust_contrast_brightness(&self, img: &DynamicImage, contrast: f32, brightness: f32) -> DynamicImage {
|
|
||||||
// Apply contrast stretch
|
|
||||||
if let DynamicImage::ImageLuma8(gray) = img {
|
|
||||||
let adjusted = stretch_contrast(gray,
|
|
||||||
(brightness * 255.0) as u8,
|
|
||||||
((1.0 + contrast) * 255.0) as u8
|
|
||||||
);
|
|
||||||
DynamicImage::ImageLuma8(adjusted)
|
|
||||||
} else {
|
|
||||||
img.clone()
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
fn perform_ocr(&mut self, img: &DynamicImage) -> Result<String, String> {
|
|
||||||
// Convert to bytes for Tesseract
|
|
||||||
let mut bytes: Vec<u8> = Vec::new();
|
|
||||||
img.write_to(&mut std::io::Cursor::new(&mut bytes), image::ImageOutputFormat::Png)
|
|
||||||
.map_err(|e| format!("Failed to encode image: {}", e))?;
|
|
||||||
|
|
||||||
// Set image in Tesseract
|
|
||||||
self.tesseract.set_image_from_mem(&bytes)
|
|
||||||
.map_err(|e| format!("Failed to set image: {:?}", e))?;
|
|
||||||
|
|
||||||
// Get text
|
|
||||||
let text = self.tesseract.get_utf8_text()
|
|
||||||
.map_err(|e| format!("OCR failed: {:?}", e))?;
|
|
||||||
|
|
||||||
Ok(text)
|
|
||||||
}
|
|
||||||
|
|
||||||
fn parse_hp_text(text: &str) -> Option<(f32, f32)> {
|
|
||||||
// Parse formats: "1234/5678", "1234 / 5678", "1,234/5,678"
|
|
||||||
let cleaned: String = text.chars()
|
|
||||||
.filter(|c| c.is_digit(10) || *c == '/' || *c == '.' || *c == ',')
|
|
||||||
.collect();
|
|
||||||
|
|
||||||
if let Some(sep_pos) = cleaned.find('/') {
|
|
||||||
let current: f32 = cleaned[..sep_pos].replace(",", "").parse().ok()?;
|
|
||||||
let max: f32 = cleaned[sep_pos + 1..].replace(",", "").parse().ok()?;
|
|
||||||
Some((current, max))
|
|
||||||
} else {
|
|
||||||
// Single number - assume it's current HP
|
|
||||||
let current: f32 = cleaned.replace(",", "").parse().ok()?;
|
|
||||||
Some((current, current))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
fn parse_coordinates(text: &str) -> Result<(f32, f32, f32), String> {
|
|
||||||
// Parse format: "X: 12345.67 Y: 12345.89 Z: 123.45"
|
|
||||||
let mut coords = (0.0, 0.0, 0.0);
|
|
||||||
|
|
||||||
for part in text.split_whitespace() {
|
|
||||||
if part.starts_with("X:") {
|
|
||||||
coords.0 = part[2..].trim().parse().unwrap_or(0.0);
|
|
||||||
} else if part.starts_with("Y:") {
|
|
||||||
coords.1 = part[2..].trim().parse().unwrap_or(0.0);
|
|
||||||
} else if part.starts_with("Z:") {
|
|
||||||
coords.2 = part[2..].trim().parse().unwrap_or(0.0);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
Ok(coords)
|
|
||||||
}
|
|
||||||
}
|
}
|
||||||
|
|
||||||
#[derive(Debug, Clone)]
|
#[derive(Debug, Clone)]
|
||||||
|
|
|
||||||
Loading…
Reference in New Issue