I made an index "user-name" with a custom made analyzer called 'autocomplete':
client.indices.create({
index: 'user-name',
type: 'text',
settings: {
analysis: {
filter: {
autocomplete_filter: {
type: 'edge-ngram',
min_gram: 1,
max_gram: 20
}
},
analyzer: {
autocomplete: {
type: 'custom',
tokenizer: 'standard',
filter: [
'lowercase',
'autocomplete_filter'
]
}
}
}
}
}
Then I try to reference this custom made analyzer by trying to use it in a mapping:
client.indices.putMapping({
index: 'user-name',
type: 'text',
body: {
properties: {
name: {
type: 'string',
analyzer: 'autocomplete',
search_analyzer: 'standard'
}
}
}
})
but then I get this error: "reason": "analyzer [autocomplete] not found for field [name]". Why isn't my autocomplete analyzer being detected? Thanks.
You're almost there. You simply need to put the index settings inside the body parameter:
client.indices.create({
index: 'user-name',
type: 'text',
body: {
settings: {
analysis: {
filter: {
autocomplete_filter: {
type: 'edge-ngram',
min_gram: 1,
max_gram: 20
}
},
analyzer: {
autocomplete: {
type: 'custom',
tokenizer: 'standard',
filter: [
'lowercase',
'autocomplete_filter'
]
}
}
}
}
}
}
Related
I much prefer the look of buttons to a numbered list, but the default handling of prompt in waterfall dialog is to automatically change from buttons (default) to a numbered list after a certain length of content.
I'm currently implementing the prompt like this:
return await step.prompt(FOCUS_AREA_PROMPT, {
prompt: 'Got it. Can you confirm the focus area this is for?',
choices: ChoiceFactory.toChoices(FOCUS_AREAS)
});
I've tried adding a style attribute within this prompt, and also tried adding to the addDialogs line this.dialogs.add(new ChoicePrompt(FOCUS_AREA_PROMPT)); but nothing I have tried has modified the behavior of the options.
I've reviewed the ListStyle enum in the MS docs, but any method I've tried to add these in with has not made any difference. Is there any way to force buttons regardless of content length?
You can set up a choice prompt in the following manner to achieve the buttons look you are seeking. For reference, you can read more about forChannel here.
Alter to match your needs.
Hope of help!
[edit]
Updated below to represent the two ways a Choice Prompt can be assembled and how the value is output (via imBack). When using toChoices, the associated button value is returned in activity.text and in stepContext.result.value (as type Object). When using forChannel, the associated button value is returned in activity.text and in stepContext.result (as type String).
As discussed in the comments, the button title length has a character limit however this is channel specific. When testing in Web Chat, the limit is 20 characters. Adjusting the FOCUS_AREAS "AI & Machine Learning" value (21 chars) to "AI/Machine Learning" (19 chars) results in the choices displaying as buttons and not a list.
Option 1: using toChoices
async choiceStep ( stepContext ) {
const stepResult = stepContext.context.activity.text;
const FOCUS_AREAS = [ 'Chatbots', 'RPA', 'Blockchain', 'AR/VR', 'AI/Machine Learning' ]
if ( stepResult ) {
return await stepContext.prompt( CHOICE_PROMPT, {
prompt: 'Got it. Can you confirm the focus area this is for?',
choices: ChoiceFactory.toChoices( FOCUS_AREAS )
} );
}
}
activity:
{ type: 'message',
id: 'A50eelAPrFIHKv9XeCRm24-o|0000021',
timestamp: 2019-09-25T20:34:30.562Z,
serviceUrl: 'https://directline.botframework.com/',
channelId: 'directline',
from: [Object],
conversation: [Object],
recipient: [Object],
textFormat: 'plain',
locale: 'en-US',
text: 'Chatbots',
channelData: [Object] },
info:
{ index: 1,
options: {},
reason: 'endCalled',
result:
{ value: 'Chatbots', index: 0, score: 1, synonym: 'Chatbots' },
values: { instanceId: 'c10ed437-77eb-4502-cd24-e89d4c5e45cf' },
onNext: [AsyncFunction: onNext] }
Option 2: using forChannel
async choiceStep ( stepContext ) {
const stepResult = stepContext.context.activity.text;
if ( stepResult ) {
const message = ChoiceFactory.forChannel(
stepContext.context, [
{ value: 'Chatbots', action: { type: 'imBack', title: 'Chatbots', value: 'Chatbots' } },
{ value: 'RPA', action: { type: 'imBack', title: 'RPA', value: 'RPA' } },
{ value: 'Blockchain', action: { type: 'imBack', title: 'Blockchain', value: 'Blockchain' } },
{ value: 'AR/VR', action: { type: 'imBack', title: 'AR/VR', value: 'AR/VR' } },
{ value: 'AI/Machine Learning', action: { type: 'imBack', title: '', value: 'AI/Machine Learning' }, text: 'AI/Machine Learning' },
], `Which do you choose?`
);
await stepContext.context.sendActivity( message );
}
return { status: DialogTurnStatus.waiting };
}
activity:
{ type: 'message',
id: 'Cw5xvHTv6RCDWf3kkyS3Ir-o|0000205',
timestamp: 2019-09-25T20:21:30.320Z,
serviceUrl: 'https://directline.botframework.com/',
channelId: 'directline',
from: [Object],
conversation: [Object],
recipient: [Object],
textFormat: 'plain',
locale: 'en-US',
text: 'Chatbots',
channelData: [Object] },
info:
{ index: 1,
options: {},
reason: 'continueCalled',
result: 'Chatbots',
values: { instanceId: '4becefed-88d2-773e-6184-91456609a26a' },
onNext: [AsyncFunction: onNext] }
When I am doing custom mapping using kibana its working properly but when I am doing the same thing in my node program its showing mapper parsing exception.
Reason:Root mapping definition has unsupported parameters:tags(custom mapping name)
Because in kibana i am able to use include_type_name =true but in my node program it is not available.
var name = req.body.templatename;
var index_patterns = req.body.index_patterns;
console.log(index_patterns);
const opts: IndicesPutTemplateParams = {
name: name,
body: {
index_patterns: [index_patterns],
settings: {
analysis: {
filter: {
autocomplete_filter: {
type: "edge_ngram",
min_gram: 1,
max_gram: 20
}
},
analyzer: {
autocomplete: {
type: "custom",
tokenizer: "standard",
filter: [
"lowercase",
"autocomplete_filter"
]
}
}
}
},
mappings: {
tags: {
properties: {
name: {
type: "text",
analyzer: "autocomplete",
search_analyzer: "standard"
},
normalized: {
type: "text"
},
status: {
type: "text"
},
createdat: {
type: "date"
},
updatedat: {
type: "date"
}
}
}
}
}
}
try {
esClient.indices.putTemplate(opts).then((data: any) => {
return res.json({
data
});
console.log(data);
}).catch((err: any) => {
console.log(err);
res.status(500).json({
err
})
});
} catch (error) {
res.status(500).json({
error
})
}
}```
As Per documentation you need to give include_type_name as
client.indices.putTemplate({
name: string,
include_type_name: boolean, --->
order: number,
create: boolean,
timeout: string,
master_timeout: string,
flat_settings: boolean,
body: object -> mapping object
})
Or you can drop mapping name tags from mapping
mappings: {
tags: { ---> remove
Hi guys I have this code :
let test = await client.search({
index: 'test',
type: 'doc',
body: {
query: {
match: {
title: 'something',
}
}
}
});
this code is searching by 1 query which is title: 'something' , but I want to change it to search with multiple keys, for example:
let test = await client.search({
index: 'test',
type: 'doc',
body: {
query: {
match: {
title: 'something',
desc: 'some Qualifications'
}
}
}
});
but this code doesn't work and I can't find anything that will work like that, can anyone help?
You need to combine all the match queries using a bool/must query, like this:
let test = await client.search({
index: 'test',
type: 'doc',
body: {
query: {
bool: {
must: [
{
match: {
title: 'something',
}
},
{
match: {
desc: 'some Qualifications',
}
}
]
}
}
}
});
I have one document like that
var ConfigSchema = new Schema({
basicConfig: {
levelId: {
type: Number,
required: true
},
hostId: {
type: Number,
required: true
},
Name: {
type: String,
trim: true
},
Settings: {
Type1: {
// some types here...
}
Type2: {
// some types here...
}
}
},
enrolls: {
name: {type: String},
list: [String]
},
awards: {
enable: {type: Boolean},
Primary: {
amount: {type: Number},
type: {type: String}
}
}
Now I want to find configs with hostId matches 60, and selecting basicConfig field.
Config.findOne({ 'basicConfig.hostId': 60 })
.select('basicConfig').exec(function(err, configs) {
if (err) {
console.error(err);
return ;
}
console.log(configs);
});
However, all fields of this document will be returned. It seems that the select does NOT work? Why?
Output:
{ _id: 555c4144c0bff1541d0e4059,
enrolls: {},
awards: { primary: { PrimarySettings: {}, primaryAck: {} } },
basicConfig:
{ levelId: 24,
hostId: 60,
poolName: 'LC' } }
Also, those following codes have been test, it does not work.
BonusConfig.findOne({ 'basicConfig.hostId': 60 }, 'basicConfig', function(err, configs) {
if (err) {
console.error(err);
return ;
}
console.log(configs);
});
But, without the basicConfig field with select with the following codes, it work well.
BonusConfig.findOne({ 'basicConfig.hostId': 60 })
.select('-basicConfig').exec(function(err, configs) {
if (err) {
console.error(err);
return ;
}
console.log(configs);
});
What's wrong with my codes?
Mongoose version: 3.8.24
Mongodb version: 2.6.7
Edit 1
Here is the query log in mongoose debug mode.
Mongoose: configs.findOne({ 'basicConfig.hostId': 60 }) { fields: { basicConfig: 1 } }
Edit 2
After further investigation.
The output result of Config.findOne({ 'basicConfig.hostId': 60 }).select('basicConfig'):
{ _id: 555c4144c0bff1541d0e4059,
enrolls: {},
awards: { primary: { PrimarySettings: {}, primaryAck: {} } },
basicConfig:
{ levelId: 24,
hostId: 60,
poolName: 'LC' } }
Other fields are empty value except basicConfig. However, I want the result is
{ _id: 555c4144c0bff1541d0e4059,
basicConfig:
{ levelId: 24,
hostId: 60,
poolName: 'LC' } }
Your projection of fields that need to be returned is missing the additional parameter. Try this:
BonusConfig.findOne({ 'basicConfig.hostId': 60 }, {'basicConfig':1}, function(err, configs) {
if (err) {
console.error(err);
return ;
}
console.log(configs);
});
Quoting this
This behavior is by design, but admittedly it's not very well-designed. Mongoose is over-eager when it comes to creating sub-docs when loading from the database. Planning on changing that in v5.
I'm working on a Flowchart editor and I want the ui.inspector to edit labels on links.
I did the following:
function createInspector(cellView) {
if (!inspector || inspector.options.cellView !== cellView) {
if (inspector) {
inspector.remove();
}
inspector = new joint.ui.Inspector({
inputs: {
labels:
attrs: {
text:{
text: { type: 'textarea', group: 'Labels', label: 'Label', index: 2 },
}
}
},
},
},
groups: {
labels:[ { label: 'Labels', index: 1 },
}],
cellView: cellView
});
$('#inspector-holder-create').html(inspector.render().el);
}
}
paper.on('cell:pointerdown', function(cellView) {
createInspector(cellView);
});
However, when I edit a link it shows in the JSON output:
"labels": {
"0": {
"attrs": {
"text": {
"text": "Text I entered"
}
}
}
},
but doesn't actually render on the link in the stencil.
I think the problem is with the { "0": part the inspector adds. I want to remove that and replace with it [ ] so the output will be
labels: [
{ attrs: { text: { text: 'label' } } }
]
What should I do ??
It is possible to define Inspector inputs with paths.
'labels/0/attrs/text/text': {
type: 'text',
group: 'Text',
index: 1,
label: 'Label'
}
Or as a combination of attributes nesting and paths.
'labels/0/attrs': {
text: {
text: {
type: 'text',
group: 'Text',
index: 1,
label: 'Label'
},
fontSize: {
type: 'number',
group: 'Text',
index: 2,
label: 'Font Size'
}
}
}
This is valid for Rappid v2.4.0+.
inspector = new joint.ui.Inspector({
inputs: {
'labels': [
{attrs: {
text: {
text: {
type: 'text',
group: 'someGroup',
index: 1,
label: "Label"
}
}
}}
]
}});