I am having issue with form recognizer pre-build model JSON output. It is not giving me the keyValuePairs array results. I have watched alot of videos where they are getting this array in the JSON output but in my case it only returns "readResults" array rather that "keyValuePairs". I have taken a screenshot of keyValuePairs array from a video. Please have a look at:
And here is the output of my form recognizer API call results:
My URL sample is : https://myapp.com/formrecognizer/v2.1/prebuilt/invoice/analyze?includeTextDetails=true
Is there any kind of additional parameter needed to be passed in the URL to get this "keyValueParis" array ? Like for readResults I am passing the parameter includeTextDetails=true
Thanks
Your JSON output should have an additional section documrResults which includes the key value pairs. The readResult of the JSON output is all the text in the document.
I have to get number of input tabs or fields based on given xml or json file.
example : If xml has 3 tags, then form should contain 3 input fields.
Same with JSON.
You can use ng-repeat on service response array of objects. And also you can create new scope variable & assign unique property of response array object to it. Then you can have as many number of input fields based on service data & also dynamic ng-model on them to handle their model data to submit it.
This is small example which demonstrates your requirement: http://jsfiddle.net/DrQ77/
For xml response, better use xml to json library plugins to convert it to json data https://github.com/abdmob/x2js. Or prefer server side sending json data (with implementing xml to json functionality at server side).
I have used IBM Watson Retrieve and Rank Web Interface to create a collection of html articles. Via the web interface I was able to upload my html articles. The problem is when I query the collection the data for id and title are not usable. Here is the query I made in the browser:
https://MY-USER-NAME:MY-PASSWORD#gateway.watsonplatform.net/retrieve-and-rank/api/v1/solr_clusters/MY-CLUSTER/solr/MY-COLLECTION/select?q=what is the basic mechanism of the transonic aileron buzz&wt=json&fl=id,title
The response I get is:
{"responseHeader":{"status":0,"QTime":106,"params":{"q":"what is the basic mechanism of the transonic aileron buzz","fl":"id,title","wt":"json"}},"response":{"numFound":12,"start":0,"docs":[{"id":"6a06f47c-cb3f-4791-9914-c84772eb9415","title":"no-title"}.....
The bold section is the problem. When using the web interface is there a way to set the title and id when uploading documents? Or, better yet, is there another way I query my collection to get the file name of the document I uploaded and/or the text from the document?
When using the web interface is there a way to set the title and id when uploading documents?
No, sorry.
However, if you upload the documents yourself from outside of the web interface, you can specify the title and ID (and the documents will be shown in the web interface when you come back to it).
is there another way I query my collection to get the file name of the document I uploaded
Yes
In the query you posted above, the last parameters you have are the fields you want to retrieve
&fl=id,title
You're retrieving the ID and the title.
If you want the name of the file that the content came from, add fileName. For example:
https://MY-USER-NAME:MY-PASSWORD#gateway.watsonplatform.net/retrieve-and-rank/api/v1/solr_clusters/MY-CLUSTER/solr/MY-COLLECTION/select?q=what is the basic mechanism of the transonic aileron buzz&wt=json&fl=id,title,fileName
is there another way I query my collection to get text from the document
Yes.
Similar to above, you just need to update the list of fields that you retrieve. The contents of the doc is put in a field called body.
So to get the ID, title, and the body, you could use:
https://MY-USER-NAME:MY-PASSWORD#gateway.watsonplatform.net/retrieve-and-rank/api/v1/solr_clusters/MY-CLUSTER/solr/MY-COLLECTION/select?q=what is the basic mechanism of the transonic aileron buzz&wt=json&fl=id,title,body
That gets you a plain text version of the contents. If you want the HTML, use contentHtml instead.
https://MY-USER-NAME:MY-PASSWORD#gateway.watsonplatform.net/retrieve-and-rank/api/v1/solr_clusters/MY-CLUSTER/solr/MY-COLLECTION/select?q=what is the basic mechanism of the transonic aileron buzz&wt=json&fl=id,title,contentHtml
At the moment I have an input field with the following directive:
ng-model="event.location"
I also have the input field attached to a Google Maps auto complete script. By way of example, if my user types "apa", the Google auto complete list will come up and there is the option of clicking "Apartments". However, the String "Apartments" does not bind to my variable and the "location" variable only contains the String "apa". How can I get AngularJS to bind to the full location once the user selects an option from Google Maps?
Update: It's just a one liner for Google Maps
new google.maps.places.Autocomplete(document.getElementById('location-text'));
Although I'm sure a lot more is happening behind the scenes
I'm attempting to upload a photo to Facebook via the Graph API on a test user account for my app. With just the url, link, name parameters present, the upload works fine returning a valid photo id.
However, if I use the additional tags parameter, I end up with the following error returned:
{
"error": {
"message": "(#100) param tags must be an array.",
"type": "OAuthException",
"code": 100
}
}
I've tried to provide the value for tags in almost every possible way I can think of, as I know the Graph API isn't straightforward (even the parameter url which is used to upload a photo from a URL isn't listed under the photo Graph API method);
A single user id
tags=100003919104407
Multiple CS'd user ids
tags=100003919104407,100003919104408,100003919104409
Array with ids not as integers
tags=[100003919104407, 100003919104404,100003919104405]
Array with ids as strings
tags=["100003919104407", "100003919104404","100003919104405"]
Array containing objects, as per the Facebook Graph API documentation
tags=[{"id":"100003919104407"},{"id":"100003919104404"},{"id":"100003919104405"}]
If someone could tell me the right format/another parameter through which to pass user ids in order to have them tagged in a photo, I'd be very grateful.
Try this
It should be in this format
[{"to":"100003919104407","x":0,"y":0},
{"to":"100003919104408","x":10,"y":10},
{"to":"100003919104409","x":20,"y":20}]
or
[{"tag_uid":"100003919104407","x":0,"y":0},
{"tag_uid":"100003919104408","x":10,"y":10},
{"tag_uid":"100003919104409","x":20,"y":20}]
Where did you get your information from? Have you check the tags section?
Create
You can create a tag on the photo by issuing an HTTP POST request to
the tags connection, PHOTO_ID/tags.
Note: This feature is intended to help users tag their friends in real
photos. You should not use this feature to encourage users to tag
their friends if their friends are not actually in that photo, or to
tag friends in composite photos. If your app is found to be
encouraging this behavior, your usage of this feature may be disabled.
You can specify which user to tag using two methods: in the URL path
as PHOTO_ID/tags/USER_ID, or in a URL parameter as
PHOTO_ID/tags?to=USER_ID. To add several tags at once, you can specify
a tags property which contains an array of tags like so
PHOTO_ID/tags?tags=[{"id":"1234"}, {"id":"12345"}]. Currently, you
cannot tag a Page in a photo using this API.
There's also an example.
the X and Y value of array need to be %, not in pixels
Example 1 (wrong)
img:(640,480) xy:(320,240)
[{tag_uid:user_id,x:320,y:240}] <- wrong, because 320 > 100%
Example 2 (right)
img:(640,480) xy:(320,240) px:(320,240) < %
[{tag_uid:user_id,x:50,y:50}] <- right, because 50 < 100%