Textract-API not returning TABLE specific data when using PHP client

0

I'm using the following code:

private function analyze(string $fileName): Result
{
    return $this->textractClient->analyzeDocument(
        [
            'Document' => [
                'S3Object' => [
                    'Bucket' => $this->bucketName,
                    'Name' => $fileName,
                ],
            ],
            'FeatureTypes' => ['TABLES', 'FORMS']
        ]
    );
}

I point to a JPEG image in the bucket and everything seems to work. However, the image contains a table with information which is processed correctly when I use the Textract web interface (by uploading the original PDF where the mentioned JPEG image was extracted from) but in the PHP result, there are no block types "TABLE" or "CELL"; they're all of the type "LINE".

Am I doing something wrong?

Any help would be highly appreciated.

Michiel
質問済み 2年前386ビュー
1回答
0

Hi, just want to confirm is your PDF file a multi-page document or just one-page document? If it is multi-page document, then it might be that the result is paginated so you cannot search for TABLE result. Thanks.

AWS
回答済み 2年前
  • The PDF is indeed a multi-page document but the JPEG I'm using with the AnalyzeDocument call is just one page as it's a single image. And this page has a table on it, which gets extracted fine when I use the web interface with the original PDF but gets extracted as LINE blocks when I analyze the JPEG.

ログインしていません。 ログイン 回答を投稿する。

優れた回答とは、質問に明確に答え、建設的なフィードバックを提供し、質問者の専門分野におけるスキルの向上を促すものです。

質問に答えるためのガイドライン

関連するコンテンツ