1import json
2import sys
3 
4from pypdf import PdfReader
5 
6 
7 
8 
9def get_full_annotation_field_id(annotation):
10    components = []
11    while annotation:
12        field_name = annotation.get('/T')
13        if field_name:
14            components.append(field_name)
15        annotation = annotation.get('/Parent')
16    return ".".join(reversed(components)) if components else None
17 
18 
19def make_field_dict(field, field_id):
20    field_dict = {"field_id": field_id}
21    ft = field.get('/FT')
22    if ft == "/Tx":
23        field_dict["type"] = "text"
24    elif ft == "/Btn":
25        field_dict["type"] = "checkbox"  
26        states = field.get("/_States_", [])
27        if len(states) == 2:
28            if "/Off" in states:
29                field_dict["checked_value"] = states[0] if states[0] != "/Off" else states[1]
30                field_dict["unchecked_value"] = "/Off"
31            else:
32                print(f"Unexpected state values for checkbox `${field_id}`. Its checked and unchecked values may not be correct; if you're trying to check it, visually verify the results.")
33                field_dict["checked_value"] = states[0]
34                field_dict["unchecked_value"] = states[1]
35    elif ft == "/Ch":
36        field_dict["type"] = "choice"
37        states = field.get("/_States_", [])
38        field_dict["choice_options"] = [{
39            "value": state[0],
40            "text": state[1],
41        } for state in states]
42    else:
43        field_dict["type"] = f"unknown ({ft})"
44    return field_dict
45 
46 
47def get_field_info(reader: PdfReader):
48    fields = reader.get_fields()
49 
50    field_info_by_id = {}
51    possible_radio_names = set()
52 
53    for field_id, field in fields.items():
54        if field.get("/Kids"):
55            if field.get("/FT") == "/Btn":
56                possible_radio_names.add(field_id)
57            continue
58        field_info_by_id[field_id] = make_field_dict(field, field_id)
59 
60 
61    radio_fields_by_id = {}
62 
63    for page_index, page in enumerate(reader.pages):
64        annotations = page.get('/Annots', [])
65        for ann in annotations:
66            field_id = get_full_annotation_field_id(ann)
67            if field_id in field_info_by_id:
68                field_info_by_id[field_id]["page"] = page_index + 1
69                field_info_by_id[field_id]["rect"] = ann.get('/Rect')
70            elif field_id in possible_radio_names:
71                try:
72                    on_values = [v for v in ann["/AP"]["/N"] if v != "/Off"]
73                except KeyError:
74                    continue
75                if len(on_values) == 1:
76                    rect = ann.get("/Rect")
77                    if field_id not in radio_fields_by_id:
78                        radio_fields_by_id[field_id] = {
79                            "field_id": field_id,
80                            "type": "radio_group",
81                            "page": page_index + 1,
82                            "radio_options": [],
83                        }
84                    radio_fields_by_id[field_id]["radio_options"].append({
85                        "value": on_values[0],
86                        "rect": rect,
87                    })
88 
89    fields_with_location = []
90    for field_info in field_info_by_id.values():
91        if "page" in field_info:
92            fields_with_location.append(field_info)
93        else:
94            print(f"Unable to determine location for field id: {field_info.get('field_id')}, ignoring")
95 
96    def sort_key(f):
97        if "radio_options" in f:
98            rect = f["radio_options"][0]["rect"] or [0, 0, 0, 0]
99        else:
100            rect = f.get("rect") or [0, 0, 0, 0]
101        adjusted_position = [-rect[1], rect[0]]
102        return [f.get("page"), adjusted_position]
103    
104    sorted_fields = fields_with_location + list(radio_fields_by_id.values())
105    sorted_fields.sort(key=sort_key)
106 
107    return sorted_fields
108 
109 
110def write_field_info(pdf_path: str, json_output_path: str):
111    reader = PdfReader(pdf_path)
112    field_info = get_field_info(reader)
113    with open(json_output_path, "w") as f:
114        json.dump(field_info, f, indent=2)
115    print(f"Wrote {len(field_info)} fields to {json_output_path}")
116 
117 
118if __name__ == "__main__":
119    if len(sys.argv) != 3:
120        print("Usage: extract_form_field_info.py [input pdf] [output json]")
121        sys.exit(1)
122    write_field_info(sys.argv[1], sys.argv[2])
123