发票解析器文档列出了该处理器试图在文档中找到的字段。
我只对发票处理器中的某些字段感兴趣，是否有一种方法可以获取字段列表并请求处理器到找到一组选定的字段的数据点？
例如，我只对Invoice_Date和Invoice_ID感兴趣。因此，处理器只为我找到这两个领域，并且不会浪费时间找到其他领域。
我找到了一个检索处理器的详细信息但没有获得现场列表的端点。
https://cloud.google。 com/document-ai/docs/reference/reth/v1beta3/projects.locations.locates.locessers/get
此过程端点没有任何指定我感兴趣的字段。
https://cloud.google。 com/document-ai/docs/reference/rest/v1beta3/projects.locations.locates.processors/process
我的研究表明此功能不存在，希望有人告诉我我错了。
Invoice Parser documentation lists the fields that this processor attempts to find in a document.
https://cloud.google.com/document-ai/docs/processors-list?hl=en_US#processor_invoice-processor
I am only interested in some of the fields in the Invoice processor, is there a way to get the list of fields and request the processor to find the data points for a selected set of fields?
E.g. I am only interested in invoice_date and invoice_id. So, the processor only finds these two fields for me and does not waste its time finding others.
I found an endpoint that retrieves the processor's details but it does not get the field list.
https://cloud.google.com/document-ai/docs/reference/rest/v1beta3/projects.locations.processors/get
This process endpoint does not have anything to specify the fields that I am interested in.
https://cloud.google.com/document-ai/docs/reference/rest/v1beta3/projects.locations.processors/process
My research indicates that this feature does not exist, hoping someone tells me that I am wrong.
发布评论
评论（1）
目前，不可能限制专用处理器提取的特定实体。处理器将始终处理整个文档，并提取可以找到的所有内容。
但是，最近将一个新参数添加到
processREST API中的方法可以提供field> fieldmask限制将在文档 对象。对于％20more％20个信息。-，field_mask， - ％3A％20spifies％20％20 fields“ rel =“ nofollow noreferrer”>用于制作处理请求的方法指南当前，这仅适用于REST API （不是客户库）和在线处理。这并不是您要寻找的，但是它可以大大限制
文档响应如果您不需要content> contentCurrently, it's not possible to limit the specific entities that are extracted by a specialized processor. The processor will always processes the entire document and extract everything it can find.
However, a new parameter has recently been added to the
processmethod in the REST API that lets you provide afieldMaskto limit the fields that will be returned in theDocumentobject. This has also been added to the How-to Guide for making a processing request. Currently, this only works for the REST API (not client libraries) and online processing.This isn't exactly what you're looking for, but it can significantly limit the size of the
Documentresponse if you don't need fields likecontent