From bd72b1d6d23d7ad759991d10bc468e6521d585ed Mon Sep 17 00:00:00 2001
From: Michael <michael@zllnr.com>
Date: Thu, 25 Jun 2026 16:10:13 +0200
Subject: [PATCH] paperless_to_openwebui.py aktualisiert

---
 paperless_to_openwebui.py | 134 +++++++++++++++++++-------------------
 1 file changed, 67 insertions(+), 67 deletions(-)

diff --git a/paperless_to_openwebui.py b/paperless_to_openwebui.py
index 83878c5..e0a517c 100644
--- a/paperless_to_openwebui.py
+++ b/paperless_to_openwebui.py
@@ -876,74 +876,74 @@ def wait_for_openwebui():
         log.info("Not ready yet, retrying in 10s...")
         time.sleep(10)
 
-    def ocr_fix_only_pass():
-        """
-        Geht alle Paperless-Dokumente durch und korrigiert nur den OCR-Text.
-        Kein Upload nach OpenWebUI – das erledigt der nächste reguläre Sync.
-        """
-        log.info("=== OCR FIX ONLY MODE ===")
-        log.info("Fetching all documents from Paperless...")
-    
-        docs = []
-        page = 1
-        while True:
-            url = f"{PAPERLESS_URL}/api/documents/?page_size=100&page={page}"
-            r = requests.get(url, headers=paperless_headers, timeout=120)
-            r.raise_for_status()
-            data = r.json()
-            results = data.get("results", [])
-            if not results:
-                break
-            docs.extend(results)
-            if not data.get("next"):
-                break
-            page += 1
-    
-        total = len(docs)
-        log.info("Found %s documents to check.", total)
-    
-        fixed = 0
-        skipped = 0
-        errors = 0
-    
-        for i, doc in enumerate(docs):
-            doc_id = doc["id"]
-            title = doc.get("title", f"doc_{doc_id}")
-            content = doc.get("content", "")
-    
-            if not content or len(content.strip()) < 50:
-                skipped += 1
-                continue
-    
-            if not needs_ocr_fix(content):
-                skipped += 1
-                if (i + 1) % 100 == 0:
-                    log.info("  Progress: %d/%d (fixed=%d, skipped=%d)",
-                             i + 1, total, fixed, skipped)
-                continue
-    
-            log.info("OCR fix needed: #%s '%s'", doc_id, title)
-            try:
-                fixed_text = fix_ocr_chunked(content)
-                if fixed_text and fixed_text != content:
-                    if update_paperless_content(doc_id, fixed_text):
-                        fixed += 1
-                        log.info("  ✓ Fixed #%s (delta: %+d chars)",
-                                 doc_id, len(fixed_text) - len(content))
-                    else:
-                        errors += 1
+def ocr_fix_only_pass():
+    """
+    Geht alle Paperless-Dokumente durch und korrigiert nur den OCR-Text.
+    Kein Upload nach OpenWebUI – das erledigt der nächste reguläre Sync.
+    """
+    log.info("=== OCR FIX ONLY MODE ===")
+    log.info("Fetching all documents from Paperless...")
+
+    docs = []
+    page = 1
+    while True:
+        url = f"{PAPERLESS_URL}/api/documents/?page_size=100&page={page}"
+        r = requests.get(url, headers=paperless_headers, timeout=120)
+        r.raise_for_status()
+        data = r.json()
+        results = data.get("results", [])
+        if not results:
+            break
+        docs.extend(results)
+        if not data.get("next"):
+            break
+        page += 1
+
+    total = len(docs)
+    log.info("Found %s documents to check.", total)
+
+    fixed = 0
+    skipped = 0
+    errors = 0
+
+    for i, doc in enumerate(docs):
+        doc_id = doc["id"]
+        title = doc.get("title", f"doc_{doc_id}")
+        content = doc.get("content", "")
+
+        if not content or len(content.strip()) < 50:
+            skipped += 1
+            continue
+
+        if not needs_ocr_fix(content):
+            skipped += 1
+            if (i + 1) % 100 == 0:
+                log.info("  Progress: %d/%d (fixed=%d, skipped=%d)",
+                         i + 1, total, fixed, skipped)
+            continue
+
+        log.info("OCR fix needed: #%s '%s'", doc_id, title)
+        try:
+            fixed_text = fix_ocr_chunked(content)
+            if fixed_text and fixed_text != content:
+                if update_paperless_content(doc_id, fixed_text):
+                    fixed += 1
+                    log.info("  ✓ Fixed #%s (delta: %+d chars)",
+                             doc_id, len(fixed_text) - len(content))
                 else:
-                    skipped += 1
-                    log.debug("  No changes for #%s", doc_id)
-            except Exception as e:
-                log.error("  ✗ Error fixing #%s: %s", doc_id, e)
-                errors += 1
-    
-        log.info("=== OCR FIX COMPLETE ===")
-        log.info("  Total:   %d", total)
-        log.info("  Fixed:   %d", fixed)
-        log.info("  Skipped: %d", skipped)
-        log.info("  Errors:  %d", errors)
+                    errors += 1
+            else:
+                skipped += 1
+                log.debug("  No changes for #%s", doc_id)
+        except Exception as e:
+            log.error("  ✗ Error fixing #%s: %s", doc_id, e)
+            errors += 1
+
+    log.info("=== OCR FIX COMPLETE ===")
+    log.info("  Total:   %d", total)
+    log.info("  Fixed:   %d", fixed)
+    log.info("  Skipped: %d", skipped)
+    log.info("  Errors:  %d", errors)
 
 # ============================================================
 # MAIN