Does AWS Glue support fixed-byte-length source data?

Hello. Can AWS Glue read source data file like below?

  20220101E00011000AAABBBCCC
  20220101E00021000あいCCC

The second record contains Japanese characters, and is the same byte length with the first record. The numbers of characters vary by record.

Thank you.

rePost-User-2300507
2 年前
If the numbers of characters are the same by record, I can use Grok pattern like "(?<col0>.{6})(?<col1>.{2})..." when making a data catalog using crawler. But in this case, numbers differ, and bytes are the same. Can anyone tell AWS Glue does support/not support data set like this?

主題

分析資料庫

標籤

AWS Glue 提取轉換與載入資料

語言

English

rePost-User-2300507

已提問 2 年前檢視次數 413 次

1 個回答

最新
最多得票
最多評論

這些答案是否有幫助？支持正確答案，以幫助社區從您的知識中受益。

已接受的答案

Unfortunately parsing data by Bytes right now is not supported in glue. I observed your data is unstructured and the only way to parse that data is Grok SerDe or Regex SerDe but they parse data by identifying patterns so they are not feasible. I will recommend you to preprocess your data and then load them in glue.

unstructured data -> preprocessing using some Custom built parser function(csv) -> S3 -> crawl and createDatabase in glue.

支援工程師

Shubh

已回答 2 年前

rePost-User-2300507
2 年前
Thank you for your reply.

Does AWS Glue support fixed-byte-length source data?

相關內容