Merge pull request #23 from jbarrow/multiline

add multiline support for #11

Author: jbarrow

README.md

@@ -47,6 +47,7 @@ commonforms <input.pdf> <output.pdf>
 | `--image-size` | int | `1600` | Image size for inference |
 | `--confidence` | float | `0.3` | Confidence threshold for detection |
 | `--fast` | flag | `False` | If running on a CPU, you can trade off accuracy for speed and run in about half the time |
+| `--multiline` | flag | `False` | If you want the detected textboxes to allow multiline inputs |
 
 
 ## CommonForms API

commonforms/__main__.py

@@ -50,6 +50,11 @@ def main():
         action="store_true",
         help="If running on a CPU, you can use --fast to get a 50% speedup with a small accuracy penalty",
     )
+    parser.add_argument(
+        "--multiline",
+        action="store_true",
+        help="If you want the detected textboxes to allow multiline inputs.",
+    )
 
     args = parser.parse_args()
 
@@ -63,6 +68,7 @@ def main():
         image_size=args.image_size,
         confidence=args.confidence,
         fast=args.fast,
+        multiline=args.multiline,
     )
 
 

commonforms/inference.py

@@ -166,6 +166,7 @@ def prepare_form(
     image_size: int = 1600,
     confidence: float = 0.3,
     fast: bool = False,
+    multiline: bool = False,
 ):
     detector = FFDNetDetector(model_or_path, device=device, fast=fast)
 
@@ -187,7 +188,7 @@ def prepare_form(
             name = f"{widget.widget_type.lower()}_{widget.page}_{i}"
 
             if widget.widget_type == "TextBox":
-                writer.add_text_box(name, page_ix, widget.bounding_box)
+                writer.add_text_box(name, page_ix, widget.bounding_box, multiline=multiline)
             elif widget.widget_type == "ChoiceButton":
                 writer.add_checkbox(name, page_ix, widget.bounding_box)
             elif widget.widget_type == "Signature":

commonforms/multiline_tools.py

@@ -30,6 +30,19 @@ def test_inference_fast(tmp_path):
     doc.document.close()
 
 
+def test_mutlinline(tmp_path):
+    output_path = tmp_path / "output.pdf"
+    commonforms.prepare_form("./tests/resources/input.pdf", output_path, fast=True, multiline=True)
+
+    assert output_path.exists()
+
+    doc = formalpdf.open(output_path)
+    assert len(doc[0].widgets()) > 0
+
+    doc.document.close()
+
+
+
 def test_encrypted_failure(tmp_path):
     # Reminder to future Joe: password for encrypted PDF is "kanbanery"
     output_path = tmp_path / "output.pdf"