[
https://issues.apache.org/jira/browse/AVRO-3562?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Kalle Niemitalo updated AVRO-3562:
----------------------------------
Description:
The C# avrogen tool generates C# comments from the "doc" properties of types
and fields. If the values of those properties contain newline characters
written as "\n" in the JSON avsc file, then avrogen copies those LF characters
into the comments. However, when avrogen is used in Windows, all the other
lines in the generated C# file end with CRLF. So the file has a mix of LF and
CRLF. This causes warnings from Visual Studio, and possibly from Git if
"core.safecrlf" has been set.
It would be better if avrogen normalized the newlines in the C# comments to
match the rest of the generated C# file.
This change should be made only to the generated comments. The
NamedSchema.Documentation and Field.Documentation properties, and the
Schema.ToString() method, should not normalize the newlines within
documentation. Likewise, the JSON string literals within the {{public static
Schema _SCHEMA = Avro.Schema.Parse(…);}} fields should have the \r and \n
characters as in the schema file.
was:
The C# avrogen tool generates C# comments from the "doc" properties of types
and fields. If the values of those comments contain newline characters written
as "\n" in the JSON avsc file, then avrogen copies those LF characters into the
comments. However, when avrogen is used in Windows, all the other lines in the
generated C# file end with CRLF. So the file has a mix of LF and CRLF. This
causes warnings from Visual Studio, and possibly from Git if "core.safecrlf"
has been set.
It would be better if avrogen normalized the newlines in the C# comments to
match the rest of the generated C# file.
This change should be made only to the generated comments. The
NamedSchema.Documentation and Field.Documentation properties, and the
Schema.ToString() method, should not normalize the newlines within
documentation. Likewise, the JSON string literals within the {{public static
Schema _SCHEMA = Avro.Schema.Parse(…);}} fields should have the \r and \n
characters as in the schema file.
> .NET/C# avrogen newline normalization in comments
> -------------------------------------------------
>
> Key: AVRO-3562
> URL: https://issues.apache.org/jira/browse/AVRO-3562
> Project: Apache Avro
> Issue Type: Improvement
> Components: csharp, tools
> Affects Versions: 1.11.0
> Reporter: Kalle Niemitalo
> Priority: Minor
>
> The C# avrogen tool generates C# comments from the "doc" properties of types
> and fields. If the values of those properties contain newline characters
> written as "\n" in the JSON avsc file, then avrogen copies those LF
> characters into the comments. However, when avrogen is used in Windows, all
> the other lines in the generated C# file end with CRLF. So the file has a mix
> of LF and CRLF. This causes warnings from Visual Studio, and possibly from
> Git if "core.safecrlf" has been set.
> It would be better if avrogen normalized the newlines in the C# comments to
> match the rest of the generated C# file.
> This change should be made only to the generated comments. The
> NamedSchema.Documentation and Field.Documentation properties, and the
> Schema.ToString() method, should not normalize the newlines within
> documentation. Likewise, the JSON string literals within the {{public static
> Schema _SCHEMA = Avro.Schema.Parse(…);}} fields should have the \r and \n
> characters as in the schema file.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
