feat: add file size tracking and formatting

- store file_size_bytes on versions
- add human_size() utility
- display normalized sizes in UI

enables size vs fidelity analysis

app/logic/document_outputs.py

@@ -469,6 +469,7 @@ def create_ocr_corrected_pdf_version(db: Session, document: Document, output_pat
     compress_pdf_with_ghostscript(out_path)
     _write_pdf_metadata(out_path, document, next_version_number, "ocr_corrected")
 
+    file_size = out_path.stat().st_size
     file_hash = sha256_for_file(out_path)
     try:
         mirror_path = _mirror_to_secondary_owner(document, out_path)
@@ -496,6 +497,7 @@ def create_ocr_corrected_pdf_version(db: Session, document: Document, output_pat
         version_type="ocr_corrected",
         file_path=str(out_path),
         sha256=file_hash,
+        file_size_bytes=file_size,
         created_by="save_ocr_corrected_pdf",
         notes="C1 output: rebuilt searchable PDF using reviewed text aligned to OCR line boxes.",
     )
@@ -539,6 +541,7 @@ def create_field_enriched_pdf_version(db: Session, document: Document, output_pa
         shutil.copy2(current_file, out_path)
 
     _write_pdf_metadata(out_path, document, next_version_number, "field_enriched")
+    file_size = out_path.stat().st_size
     file_hash = sha256_for_file(out_path)
 
     try:
@@ -556,6 +559,7 @@ def create_field_enriched_pdf_version(db: Session, document: Document, output_pa
         version_type="field_enriched",
         file_path=str(out_path),
         sha256=file_hash,
+        file_size_bytes=file_size,
         created_by="save_field_enriched_pdf",
         notes="Scaffold output: copied current file; extracted fields not yet embedded into PDF.",
     )

app/models/document_version.py

@@ -1,5 +1,5 @@
 from datetime import datetime
-from sqlalchemy import String, DateTime, ForeignKey, Text, Integer
+from sqlalchemy import String, DateTime, ForeignKey, Text, Integer, BigInteger
 from sqlalchemy.orm import Mapped, mapped_column, relationship
 
 from app.db.base import Base
@@ -20,6 +20,7 @@ class DocumentVersion(Base):
 
     file_path: Mapped[str] = mapped_column(Text, nullable=False)
     sha256: Mapped[str | None] = mapped_column(String(64), nullable=True)
+    file_size_bytes: Mapped[int | None] = mapped_column(BigInteger, nullable=True)
 
     created_by: Mapped[str | None] = mapped_column(String(100), nullable=True)
     notes: Mapped[str | None] = mapped_column(Text, nullable=True)

app/routes/documents.py

@@ -2,6 +2,7 @@ from copy import deepcopy
 from datetime import datetime
 from decimal import Decimal, InvalidOperation
 import re
+import traceback
 import os
 import hashlib
 import json
@@ -32,6 +33,7 @@ from app.models.document import Document
 from app.models.document_additional_field import DocumentAdditionalField
 from app.models.document_preset import DocumentPreset
 from app.models.text_version import TextVersion
+from app.utils.filesize import human_size
 
 router = APIRouter(prefix="/documents", tags=["documents"])
 
@@ -174,6 +176,7 @@ def _document_export_payload(document) -> dict:
 
 BASE_DIR = Path(__file__).resolve().parent.parent
 templates = Jinja2Templates(directory=str(BASE_DIR / "templates"))
+templates.env.globals["human_size"] = human_size
 
 QUALITY_FLAG_OPTIONS = [
     "bad_embedded_text",
@@ -764,7 +767,9 @@ def save_pdf(document_id: str, output_path: str = Form(""), db: Session = Depend
             create_field_enriched_pdf_version(db, document, output_path=output_path_obj)
         else:
             create_ocr_corrected_pdf_version(db, document, output_path=output_path_obj)
-    except Exception:
+    except Exception as e:
+        print("save_pdf failed:", repr(e), flush=True)
+        traceback.print_exc()
         return RedirectResponse(
             url=f"/documents/{document.document_id}?error=save_pdf_failed",
             status_code=303,

app/templates/documents/detail.html

@@ -325,6 +325,7 @@
                                             <th>Version</th>
                                             <th>Type</th>
                                             <th>Path</th>
+                                            <th>Size</th>
                                             <th>Created</th>
                                             <th>Notes</th>
                                         </tr>
@@ -342,6 +343,7 @@
         {% endif %}
     </div>
 </td>
+                                            <td>{{ human_size(version.file_size_bytes) }}</td>
 <td>{{ version.created_at }}</td>
                                             <td>{{ version.notes or "" }}</td>
                                         </tr>

app/utils/filesize.py

@@ -0,0 +1,14 @@
+def human_size(num_bytes: int | None) -> str:
+    if not num_bytes:
+        return ""
+
+    units = ["B", "KB", "MB", "GB", "TB", "PB"]
+    size = float(num_bytes)
+
+    for unit in units:
+        if size < 1024 or unit == units[-1]:
+            s = f"{size:.3f}".rstrip("0").rstrip(".")
+            if "." not in s:
+                s += ".0"
+            return f"{s} {unit}"
+        size /= 1024