| import langdetect | |
| import json | |
| DATA_FILE = "data/thecrashes_data.json" | |
| def main(): | |
| texts = get_texts() | |
| for text in texts: | |
| if langdetect.detect(text) == "en": | |
| print("\n<-------------------------------") | |
| print(text) | |
| print("------------------------------>\n") | |
| def get_texts(): | |
| with open(DATA_FILE, encoding="utf-8") as f: | |
| data = json.load(f) | |
| texts = [] | |
| for event in data: | |
| for article in event["articles"]: | |
| texts.append(article["title"] + "\n\n" + article["summary"]) | |
| return texts | |
| if __name__ == '__main__': | |
| main() | |