From 742845a3298de8d4201870ed82cacf83e0a3cd99 Mon Sep 17 00:00:00 2001
From: Marty Oehme <contact@martyoeh.me>
Date: Thu, 5 Jun 2025 22:33:48 +0200
Subject: [PATCH] Extract images from summary in Original dataclasses

---
 prophet/app.py | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/prophet/app.py b/prophet/app.py
index fc0beda..657404e 100644
--- a/prophet/app.py
+++ b/prophet/app.py
@@ -2,6 +2,7 @@ import hashlib
 import json
 import os
 import pickle
+import re
 from dataclasses import dataclass, field
 from datetime import datetime
 from pathlib import Path
@@ -28,11 +29,29 @@ class Original:  # BadJoke: Sting
     summary: str
     link: str
     date: datetime
+    image_link: str | None = None
     id: str = field(init=False)
 
+    def _extract_img(self, s: str) -> tuple[str, str]:  # [img_link, rest of string]
+        img: str
+        m = re.match(r'<img src="(?P<img>.+?)"', s)
+        try:
+            img = m.group("img")
+        except (IndexError, NameError):
+            return ("", s)
+
+        if img:
+            rest = re.sub(r"<img src=.+?>", "", s)
+            return (img, rest)
+
     def __post_init__(self):
         self.id = hashlib.sha256(self.link.encode()).hexdigest()
 
+        extracted = self._extract_img(self.summary)
+        if extracted[0]:
+            self.image_link = extracted[0]
+            self.summary = extracted[1]
+
 
 @dataclass
 class Improvement:  # GoodJoke: Queen