Update app.py

app.py

@@ -8,22 +8,37 @@ GLMOCR_BASE = os.path.dirname(glmocr.__file__)
 config_path = os.path.join(GLMOCR_BASE, "config.yaml")
 formatter_path = os.path.join(GLMOCR_BASE, "postprocess", "result_formatter.py")
 
-# ── STEP 1: Fix config ────────────────────────────────────────
+# Fallback band when API reports no gap (so all PDFs get header/footer)
+FALLBACK_HEADER_FRAC = 0.05
+FALLBACK_FOOTER_FRAC = 0.05
+
+# ── STEP 1: Remove header/footer from abandon lists only ──────
 with open(config_path, "r") as f:
     config = yaml.safe_load(f)
 
 config["pipeline"]["maas"]["enabled"] = True
 config["pipeline"]["maas"]["api_key"] = "4570c28bdea5493c9efae9dae68edc66.sGbA9DLlcX1GlvqV"
-config["pipeline"]["result_formatter"]["abandon"] = [
-    "number", "footnote", "aside_text",
-    "reference", "footer_image", "header_image",
-]
-config["pipeline"]["enable_layout"] = True
+
+to_remove = {"header", "footer"}
+
+if "layout" in config["pipeline"] and "label_task_mapping" in config["pipeline"]["layout"]:
+    abandon = config["pipeline"]["layout"]["label_task_mapping"].get("abandon")
+    if isinstance(abandon, list):
+        config["pipeline"]["layout"]["label_task_mapping"]["abandon"] = [
+            x for x in abandon if x not in to_remove
+        ]
+
+if "result_formatter" in config["pipeline"]:
+    abandon = config["pipeline"]["result_formatter"].get("abandon")
+    if isinstance(abandon, list):
+        config["pipeline"]["result_formatter"]["abandon"] = [
+            x for x in abandon if x not in to_remove
+        ]
 
 with open(config_path, "w") as f:
     yaml.dump(config, f, default_flow_style=False, sort_keys=False)
 
-print("✅ config.yaml fixed")
+print("✅ config.yaml updated (header/footer removed from abandon only)")
 
 # ── STEP 2: Fix result_formatter.py ──────────────────────────
 with open(formatter_path, "r") as f:
@@ -47,10 +62,7 @@ print("✅ result_formatter.py fixed")
 
 # ── STEP 3: Helpers ───────────────────────────────────────────
 def get_header_footer_zones(regions, norm_height=1000):
-    """
-    From API regions' bbox_2d (normalized 0 to norm_height), return
-    (header_end_frac, footer_start_frac). No hardcoded percentages.
-    """
+    """From bbox_2d (0–norm_height) return (header_end_frac, footer_start_frac)."""
     if not regions:
         return None, None
     y_tops = []
@@ -62,9 +74,7 @@ def get_header_footer_zones(regions, norm_height=1000):
             y_bottoms.append(bbox[3])
     if not y_tops:
         return None, None
-    header_end = min(y_tops) / norm_height
-    footer_start = max(y_bottoms) / norm_height
-    return header_end, footer_start
+    return min(y_tops) / norm_height, max(y_bottoms) / norm_height
 
 
 def extract_zone_text(pdf_path, page_num, y_start_frac, y_end_frac):
@@ -131,17 +141,19 @@ def run_ocr(uploaded_file):
 
             if is_pdf:
                 header_end, footer_start = get_header_footer_zones(regions, 1000)
+                he = max(header_end, FALLBACK_HEADER_FRAC) if header_end is not None else FALLBACK_HEADER_FRAC
+                fs = min(footer_start, 1.0 - FALLBACK_FOOTER_FRAC) if footer_start is not None else (1.0 - FALLBACK_FOOTER_FRAC)
 
-                if header_end is not None and header_end > 0:
-                    hdr = extract_zone_text(path, page_num, 0, header_end)
+                if he > 0:
+                    hdr = extract_zone_text(path, page_num, 0, he)
                     if hdr and hdr.strip():
                         parts.append(hdr.strip())
 
                 if page_md:
                     parts.append(page_md)
 
-                if footer_start is not None and footer_start < 1.0:
-                    ftr = extract_zone_text(path, page_num, footer_start, 1.0)
+                if fs < 1.0:
+                    ftr = extract_zone_text(path, page_num, fs, 1.0)
                     if ftr and ftr.strip():
                         parts.append(ftr.strip())
             else:
@@ -163,7 +175,7 @@ def run_ocr(uploaded_file):
 with gr.Blocks(title="GLM-OCR") as demo:
     gr.Markdown("""
     # 🔍 GLM-OCR
-    Upload a PDF or image. Headers and footers from the PDF text layer (all PDFs).
+    Upload a PDF or image. Headers and footers from PDF text layer (all PDFs).
     """)
 
     file_input = gr.File(