【Document AI】確定申告のために為替差益の計算に必要なデータを抽出する

1
# [START documentai_process_document]
2
from typing import Optional
3

4
from google.api_core.client_options import ClientOptions
5
from google.cloud import documentai  # type: ignore
6

7
# TODO(developer): Uncomment these variables before running the sample.
8
# project_id = "YOUR_PROJECT_ID"
9
# location = "YOUR_PROCESSOR_LOCATION" # Format is "us" or "eu"
10
# processor_id = "YOUR_PROCESSOR_ID" # Create processor before running sample
11
# file_path = "/path/to/local/pdf"
12
# mime_type = "application/pdf" # Refer to <https://cloud.google.com/document-ai/docs/file-types> for supported file types
13
# field_mask = "text,entities,pages.pageNumber"  # Optional. The fields to return in the Document object.
14
# processor_version_id = "YOUR_PROCESSOR_VERSION_ID" # Optional. Processor version to use
15

16
def process_document_sample(
17
    project_id: str,
18
    location: str,
19
    processor_id: str,
20
    file_path: str,
21
    mime_type: str,
22
    field_mask: Optional[str] = None,
23
    processor_version_id: Optional[str] = None,
24
) -> None:
25
    # You must set the `api_endpoint` if you use a location other than "us".
26
    opts = ClientOptions(api_endpoint=f"{location}-documentai.googleapis.com")
27

28
    client = documentai.DocumentProcessorServiceClient(client_options=opts)
29

30
    if processor_version_id:
31
        # The full resource name of the processor version, e.g.:
32
        # `projects/{project_id}/locations/{location}/processors/{processor_id}/processorVersions/{processor_version_id}`
33
        name = client.processor_version_path(
34
            project_id, location, processor_id, processor_version_id
35
        )
36
    else:
37
        # The full resource name of the processor, e.g.:
38
        # `projects/{project_id}/locations/{location}/processors/{processor_id}`
39
        name = client.processor_path(project_id, location, processor_id)
40

41
    # Read the file into memory
42
    with open(file_path, "rb") as image:
43
        image_content = image.read()
44

45
    # Load binary data
46
    raw_document = documentai.RawDocument(content=image_content, mime_type=mime_type)
47

48
    # For more information: <https://cloud.google.com/document-ai/docs/reference/rest/v1/ProcessOptions>
49
    # Optional: Additional configurations for processing.
50
    process_options = documentai.ProcessOptions(
51
        # Process only specific pages
52
        individual_page_selector=documentai.ProcessOptions.IndividualPageSelector(
53
            pages=[1]
54
        )
55
    )
56

57
    # Configure the process request
58
    request = documentai.ProcessRequest(
59
        name=name,
60
        raw_document=raw_document,
61
        field_mask=field_mask,
62
        process_options=process_options,
63
    )
64

65
    result = client.process_document(request=request)
66

67
    # For a full list of `Document` object attributes, reference this page:
68
    # <https://cloud.google.com/document-ai/docs/reference/rest/v1/Document>
69
    document = result.document
70

71
    # Read the text recognition output from the processor
72
    print("The document contains the following text:")
73
    print(document.text)
74

75
# [END documentai_process_document]
76

77
process_document_sample(
78
    project_id="000000000000",
79
    location="us",
80
    processor_id="xxxxxxxxxxxxxx ",
81
    file_path = "/content/drive/MyDrive/TemporarySave/ImagesForAgentWorking/スクリーンショット 2025-03-14 072159.png",
82
    mime_type="image/png",
83
)

1
The document contains the following text:
2
Document Al / プロセッサ / プロセッサ: xxxxxxxxxxxxxx / 始める
3
←
4
!i!
5
=フィルタ
6
テキストを入力して結果を
7
絞り込む
8
+ 新しいフィールドを作成
9
dvidend
10
3,946
11
rate_less
12
149.4900
13
rate_more
14
150.2600
15
ラベル付きとしてマーク
16
I>
17
52
18
Q
19
(国内源泉徴収税の明細)
20
配当金等支払日
21
国内支払日
22
現地基準日
23
銘柄コード
24
銘
25
柄名
26
2024/02/15
27
2024/02/16
28
2024/02/12
29
304-AAPL
30
アップル
31
分配通貨
32
外国源泉税率(%)
33
1単位あたり金額
34
決済方法
35
円貨決済用レート
36
口座区分
37
勘定設定年
38
備考
39
米国ドル
40
10.0
41
0.24
42
外貨決済
43
数量
44
配当金等金額
45
外国源泉
46
徴収税額
47
外国手数料
48
外国精算金額
49
国内源泉
50
徴収税額
51
国内手数料
52
消費税
53
受取金額
54
外貨
55
23.76
56
4.78
57
0.00
58
0.00
59
18.98
60
110
61
26.40
62
2.64
63
0.00
64
円貨
65
ラベル 「rate_more」を作成しました
66
X
67
申告レート基準日
68
為替レート基準日
69
申告レート
70
為替レート
71
配当金等金額(円)
72
外国源泉
73
徴収税額(円)
74
国内課税所得額(円)
75
所得税
76
地方税
77
国内源泉
78
徴収税額
79
2024/02/15
80
149.4900
81
外貨
82
3.61
83
1.17
84
4.78
85
2024/02/16
86
150.2600
87
3,946
88
394
89
3,552
90
円貨
91
543
92
177