[
https://issues.apache.org/jira/browse/NIFI-4185?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16394487#comment-16394487
]
Johannes Peter edited comment on NIFI-4185 at 3/11/18 12:38 PM:
----------------------------------------------------------------
[~alopresto]:
Started implementing an XML Record Reader. Shall I create a separate ticket
for this?
Similar to the JSON readers, the XML reader will expect either a single record
(e. g. <root><field1>content<field1/><field2> ... <root/>) or an array of
records (e. g. <root><record><field1>content<field1/><field2> ...
<record/><record> ... <root/>)
The reader will be aligned with common transformators. "Normal" fields (e. g.
String, Integer) can be described by simple key-value pairs:
XML definition
<root>
<record>
<field1>content<field1/>
<field2>123<field2/>
<record/>
<root/>
Schema definition
{code}
{ "name": "testschema", "namespace": "nifi", "type": "record",
"fields": [ \{ "name": "field1", "type": "string" }, \{ "name": "field2",
"type": "int" } ] }
{code}
Parsing of attributes or nested fields require the definition of nested records
and a field name for the content (optional, a prefix for attributes can be
defined):
Property: CONTENT_FIELD=content_field
Property: ATTRIBUTE_PREFIX=attr.
XML definition
<root>
<record>
<field1 attribute="attr123">some text<field1/>
<field2 attribute="attr123">
<nested1>some nested text</nested1>
<nested2>some other nested text</nested2>
<field2/>
<record/>
<root/>
Schema definition
{
"name": "testschema",
"namespace": "nifi",
"type": "record",
"fields": [
{
"name": "field1",
"type":
{ "name": "NestedRecord", "type": "record", "fields"
: [ \\{"name": "attr.attribute", "type": "string"}
,
\{"name": "content_field", "type": "string"}
]
}
},
{
"name": "field2",
"type":
{ "name": "NestedRecord", "type": "record", "fields" : [
\\{"name": "attr.attribute", "type": "string"}
,
\{"name": "nested1", "type": "string"},
\{"name": "nested2", "type": "string"}
]
}
}
]
}
What do you say?
was (Author: jope):
[~alopresto]:
Started implementing an XML Record Reader. Shall I create a separate ticket
for this?
Similar to the JSON readers, the XML reader will expect either a single record
(e. g. <root><field1>content<field1/><field2> ... <root/>) or an array of
records (e. g. <root><record><field1>content<field1/><field2> ...
<record/><record> ... <root/>)
The reader will be aligned with common transformators. "Normal" fields (e. g.
String, Integer) can be described by simple key-value pairs:
XML definition
<root>
<record>
<field1>content<field1/>
<field2>123<field2/>
<record/>
<root/>
Schema definition
```json
{ "name": "testschema", "namespace": "nifi", "type": "record",
"fields": [ \{ "name": "field1", "type": "string" }, \{ "name": "field2",
"type": "int" } ] }
```
Parsing of attributes or nested fields require the definition of nested records
and a field name for the content (optional, a prefix for attributes can be
defined):
Property: CONTENT_FIELD=content_field
Property: ATTRIBUTE_PREFIX=attr.
XML definition
<root>
<record>
<field1 attribute="attr123">some text<field1/>
<field2 attribute="attr123">
<nested1>some nested text</nested1>
<nested2>some other nested text</nested2>
<field2/>
<record/>
<root/>
Schema definition
{
"name": "testschema",
"namespace": "nifi",
"type": "record",
"fields": [
{
"name": "field1",
"type":
{ "name": "NestedRecord", "type": "record", "fields"
: [ \\{"name": "attr.attribute", "type": "string"}
,
\{"name": "content_field", "type": "string"}
]
}
},
{
"name": "field2",
"type":
{ "name": "NestedRecord", "type": "record", "fields" : [
\\{"name": "attr.attribute", "type": "string"}
,
\{"name": "nested1", "type": "string"},
\{"name": "nested2", "type": "string"}
]
}
}
]
}
What do you say?
> Add XML record reader & writer services
> ---------------------------------------
>
> Key: NIFI-4185
> URL: https://issues.apache.org/jira/browse/NIFI-4185
> Project: Apache NiFi
> Issue Type: New Feature
> Components: Extensions
> Affects Versions: 1.3.0
> Reporter: Andy LoPresto
> Priority: Major
> Labels: json, records, xml
>
> With the addition of the {{RecordReader}} and {{RecordSetWriter}} paradigm,
> XML conversion has not yet been targeted. This will replace the previous
> ticket for XML to JSON conversion.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
