1import json
2import sys
3 
4from pypdf import PdfReader, PdfWriter
5 
6from extract_form_field_info import get_field_info
7 
8 
9 
10 
11def fill_pdf_fields(input_pdf_path: str, fields_json_path: str, output_pdf_path: str):
12    with open(fields_json_path) as f:
13        fields = json.load(f)
14    fields_by_page = {}
15    for field in fields:
16        if "value" in field:
17            field_id = field["field_id"]
18            page = field["page"]
19            if page not in fields_by_page:
20                fields_by_page[page] = {}
21            fields_by_page[page][field_id] = field["value"]
22    
23    reader = PdfReader(input_pdf_path)
24 
25    has_error = False
26    field_info = get_field_info(reader)
27    fields_by_ids = {f["field_id"]: f for f in field_info}
28    for field in fields:
29        existing_field = fields_by_ids.get(field["field_id"])
30        if not existing_field:
31            has_error = True
32            print(f"ERROR: `{field['field_id']}` is not a valid field ID")
33        elif field["page"] != existing_field["page"]:
34            has_error = True
35            print(f"ERROR: Incorrect page number for `{field['field_id']}` (got {field['page']}, expected {existing_field['page']})")
36        else:
37            if "value" in field:
38                err = validation_error_for_field_value(existing_field, field["value"])
39                if err:
40                    print(err)
41                    has_error = True
42    if has_error:
43        sys.exit(1)
44 
45    writer = PdfWriter(clone_from=reader)
46    for page, field_values in fields_by_page.items():
47        writer.update_page_form_field_values(writer.pages[page - 1], field_values, auto_regenerate=False)
48 
49    writer.set_need_appearances_writer(True)
50    
51    with open(output_pdf_path, "wb") as f:
52        writer.write(f)
53 
54 
55def validation_error_for_field_value(field_info, field_value):
56    field_type = field_info["type"]
57    field_id = field_info["field_id"]
58    if field_type == "checkbox":
59        checked_val = field_info["checked_value"]
60        unchecked_val = field_info["unchecked_value"]
61        if field_value != checked_val and field_value != unchecked_val:
62            return f'ERROR: Invalid value "{field_value}" for checkbox field "{field_id}". The checked value is "{checked_val}" and the unchecked value is "{unchecked_val}"'
63    elif field_type == "radio_group":
64        option_values = [opt["value"] for opt in field_info["radio_options"]]
65        if field_value not in option_values:
66            return f'ERROR: Invalid value "{field_value}" for radio group field "{field_id}". Valid values are: {option_values}' 
67    elif field_type == "choice":
68        choice_values = [opt["value"] for opt in field_info["choice_options"]]
69        if field_value not in choice_values:
70            return f'ERROR: Invalid value "{field_value}" for choice field "{field_id}". Valid values are: {choice_values}'
71    return None
72 
73 
74def monkeypatch_pydpf_method():
75    from pypdf.generic import DictionaryObject
76    from pypdf.constants import FieldDictionaryAttributes
77 
78    original_get_inherited = DictionaryObject.get_inherited
79 
80    def patched_get_inherited(self, key: str, default = None):
81        result = original_get_inherited(self, key, default)
82        if key == FieldDictionaryAttributes.Opt:
83            if isinstance(result, list) and all(isinstance(v, list) and len(v) == 2 for v in result):
84                result = [r[0] for r in result]
85        return result
86 
87    DictionaryObject.get_inherited = patched_get_inherited
88 
89 
90if __name__ == "__main__":
91    if len(sys.argv) != 4:
92        print("Usage: fill_fillable_fields.py [input pdf] [field_values.json] [output pdf]")
93        sys.exit(1)
94    monkeypatch_pydpf_method()
95    input_pdf = sys.argv[1]
96    fields_json = sys.argv[2]
97    output_pdf = sys.argv[3]
98    fill_pdf_fields(input_pdf, fields_json, output_pdf)
99