def on_get(self, req, resp):
ocr_url = req.get_param('ocr_url', required=True)
try:
insights = extract_ocr_insights(ocr_url,
[InsightType.nutrient.name])
except requests.exceptions.RequestException:
resp.media = {
'error': "download_error",
'error_description':
"an error occurred during OCR JSON download",
}
return
except OCRParsingException as e:
logger.error(e)
resp.media = {
'error': "invalid_ocr",
'error_description': "an error occurred during OCR parsing",
}
return
if not insights:
resp.media = {
'nutrients': {},
}
else:
resp.media = {'nutrients': insights['nutrient'][0]['nutrients']}
def on_get(self, req, resp):
ocr_url = req.get_param("ocr_url", required=True)
try:
insights = extract_ocr_insights(ocr_url, DEFAULT_INSIGHT_TYPES)
except requests.exceptions.RequestException:
resp.media = {
"error": "download_error",
"error_description":
"an error occurred during OCR JSON download",
}
return
except OCRParsingException as e:
logger.error(e)
resp.media = {
"error": "invalid_ocr",
"error_description": "an error occurred during OCR parsing",
}
return
resp.media = {
"insights": insights,
}
def on_get(self, req: falcon.Request, resp: falcon.Response):
ocr_url = req.get_param("ocr_url", required=True)
if not ocr_url.endswith(".json"):
raise falcon.HTTPBadRequest("a JSON file is expected")
try:
insights = extract_ocr_insights(ocr_url, [InsightType.nutrient])
except requests.exceptions.RequestException:
resp.media = {
"error": "download_error",
"error_description":
"an error occurred during OCR JSON download",
}
return
except OCRParsingException as e:
logger.error(e)
resp.media = {
"error": "invalid_ocr",
"error_description": "an error occurred during OCR parsing",
}
return
if not insights:
resp.media = {
"nutrients": {},
}
else:
nutrient_insights = insights[InsightType.nutrient]
resp.media = nutrient_insights.to_dict()
def predict_insight(ocr_url: str) -> None:
import json
from robotoff.insights.extraction import DEFAULT_INSIGHT_TYPES, extract_ocr_insights
from robotoff.utils import get_logger
get_logger()
results = extract_ocr_insights(ocr_url, DEFAULT_INSIGHT_TYPES)
print(json.dumps(results, indent=4))