Preserve partial results when browser closes

Author: utkuvibing

diagnostics.py

@@ -27,6 +27,7 @@
     "timeline_empty": "No tweet articles were detected in the loaded timeline.",
     "timeline_stalled": "The timeline stopped advancing before the requested amount was collected.",
     "partial_target_not_met": "The scrape ended with fewer items than requested.",
+    "browser_window_closed": "The browser window closed or ChromeDriver lost the active web view during scraping.",
     "tweet_parse_failed": "A tweet article was detected but required fields could not be parsed.",
     "full_text_failed": "A long tweet was detected but full text extraction returned no content.",
     "article_extraction_failed": "An X Article was detected but article content extraction returned no content.",

main.py

@@ -582,6 +582,16 @@ def main():
         import traceback
         traceback.print_exc()
 
+        if scraper and hasattr(scraper, 'tweets_collected') and scraper.tweets_collected and not tweets:
+            tweets = scraper.tweets_collected
+
+        error_text = str(e).lower()
+        error_reason = (
+            "browser_window_closed"
+            if "no such window" in error_text or "web view not found" in error_text
+            else "unknown_error"
+        )
+
         # Hata durumunda da kaydetmeyi dene
         if tweets and config:
             print(f"\nHataya rağmen {len(tweets)} tweet kaydediliyor...")
@@ -609,10 +619,11 @@ def main():
                         "export_saving",
                         "warning",
                         "Error export saved after exception",
+                        reason=error_reason,
                         path=output_path,
                         total_tweets=len(tweets),
                     )
-                    save_cli_run_log(run_log, "failed")
+                    save_cli_run_log(run_log, "partial")
             except:
                 if run_log:
                     record_event(
@@ -630,7 +641,7 @@ def main():
                 "unknown_error",
                 "error",
                 f"Unhandled exception: {e}",
-                reason="unknown_error",
+                reason=error_reason,
             )
             save_cli_run_log(run_log, "failed")
 

scraper.py

@@ -19,7 +19,9 @@
 from selenium.common.exceptions import (
     TimeoutException,
     NoSuchElementException,
+    NoSuchWindowException,
     StaleElementReferenceException,
+    WebDriverException,
 )
 from webdriver_manager.chrome import ChromeDriverManager
 
@@ -983,6 +985,7 @@ def scrape_by_count(self, count: int) -> List[Tweet]:
         max_recovery_attempts = 3
         scan_cycles = 0
         max_scan_cycles = max(60, count * 8)
+        browser_lost = False
 
         try:
             while len(self.tweets_collected) < count:
@@ -1093,9 +1096,40 @@ def scrape_by_count(self, count: int) -> List[Tweet]:
         except KeyboardInterrupt:
             print(f"\n\nDurduruldu! {len(self.tweets_collected)} tweet toplandı.")
             raise  # Ana programa ilet
+        except NoSuchWindowException as e:
+            browser_lost = True
+            print(f"\nBrowser penceresi kapandı veya Chrome bağlantısı koptu. {len(self.tweets_collected)} tweet kısmi sonuç olarak kullanılacak.")
+            record_event(
+                self.run_log,
+                "browser",
+                "error" if not self.tweets_collected else "warning",
+                f"Browser window closed during count scrape: {e}",
+                reason="browser_window_closed",
+                collected=len(self.tweets_collected),
+                target=count,
+                scan_cycles=scan_cycles,
+            )
+        except WebDriverException as e:
+            if "no such window" not in str(e).lower() and "web view not found" not in str(e).lower():
+                raise
+            browser_lost = True
+            print(f"\nChrome webview kayboldu. {len(self.tweets_collected)} tweet kısmi sonuç olarak kullanılacak.")
+            record_event(
+                self.run_log,
+                "browser",
+                "error" if not self.tweets_collected else "warning",
+                f"Chrome webview was lost during count scrape: {e}",
+                reason="browser_window_closed",
+                collected=len(self.tweets_collected),
+                target=count,
+                scan_cycles=scan_cycles,
+            )
 
         # Scroll bitti, şimdi show more olan tweetlerin tam metnini al
-        self._process_show_more_tweets()
+        if browser_lost:
+            print("Browser kapandığı için show more/article tam metin alma adımı atlandı.")
+        else:
+            self._process_show_more_tweets()
 
         print(f"Toplam {len(self.tweets_collected)} tweet toplandı.")
         if not self.tweets_collected: