from ruia import Item
from ruia_ocr import (OcrField, BaiduOcrService, OcrSpider, OcrResponse, 
                     OcrRequest)

app_id = 'xxxx'
api_key ='xxxx' 
secret_key = 'xxxx'

img = ".\test.jpg"
img2 = ".\test2.png"
img3 = r'http://static.zongheng.com/upload/cover/07/f9/07f9c3d9899a7df6284c1d340d45ba8c.jpeg'
    
class OcrItem(Item):
    title = OcrField(re_select=r'.*')

class MySpider(OcrSpider):

    ocr_service = BaiduOcrService(app_id, api_key, secret_key)

    # ocr_region:str  x1,y1,x2,y2;...
    # default: '' means  the whole picture to ocr
    # 1,1,0.5,0.5 means: cut from image 1,1, image.width * 0.5, image.height * 0.5
    # 1,1,200,200 means: cut from image 1,1,200,200 
    # 1,1,0.5,0.5;0.5,0.5,0.99,0.99 means: 
    #   cut from image 1,1,image.width * 0.5, image.height * 0.5 get image1  
    #   cut from image image.width*0.5, image.height*0.5 ,image.width*0.99, image.height*0.99  get image2
    #   stitching image1 image2 by row get new image to ocr
    ocr_region = '1,1,0.5,0.99'

    start_urls = [img, img2, img3]

    concurrency = 1

    async def parse(self, response: OcrResponse):
        item = await OcrItem.get_item(html=await response.text())
        return item

TIPS:
    use OcrRequest(url=image_source, sevice=ocr_service)


        

if __name__ == '__main__':
    MySpider.start()