contentstack / datasync-filesystem-sdk Goto Github PK

According to the documentation

it should be possible to pass 1, to active the field again. But it has no effect, because the check only checks if the property is present, it doesn't check the value

this.projections = Object.keys(this.contentStore.projections);
This gets all projections regardless if 0 or 1

result.docs.forEach((doc) => { this.projections.forEach((key) => { if (doc.hasOwnProperty(key)) { delete doc[key]; } }); });
This only checks the "key" regardless if the value is 0 or 1.

If I use .includeReferences() on a query, then all assets in the entry returned by that query lack relevant fields such as _internal_url. For example if I had a file field named "image" all I would see for that field in the returned JSON is "image":"".

Describe the bug

• .tag() method accepts an empty array

To Reproduce

Query a content type
Create a query that uses to filter by tags
In the query, enter an empty array

Notice: The result is empty. In contentstack-javascript the result is all the entries. Thus, the behaviour is different in both

Query

Stack.contentType('blog')
  .entries()
  .tags([])
  .find()

Expected behavior
The query should be invalidated OR throw errors

For future references: If you wish to query for an empty 'Array' field, use $size operator inside query

// example
Stack.contentType('blog')
  .entries()
  .query({tags: {$size: 0}})

Regards,
Ramanathan K

We get the following exception, when we query a content type including references, and one (or more) references are not yet published.
TypeError: Cannot read properties of undefined (reading 'hasOwnProperty') at Stack.<anonymous> (/home/node/app/node_modules/datasync-filesystem-sdk/dist/stack.js:1208:49) at Generator.next (<anonymous>) at fulfilled (/home/node/app/node_modules/datasync-filesystem-sdk/dist/stack.js:11:58)

We are using
blogQuery.include([
"buttons",
"content",
"content.deal_items",
"content.deal_items.image16_9.channel",
"content.deal_items.image16_9.competition",
"content.deal_items.default_comparison",
"content.deal_items.secondary_button.reference",
"content.deal_items.filter_tags",
"content.rail.manually.content_link.reference",
"content.primary_button.reference",
"content.secondary_button.reference"
])

And the content type instance looks like this
`{
"title": "Deal Selection - Deal Cards",
"headline": "",
"subheadline": "",
"text": "",
"config": {
"variant": "dealCard",
"mobile": "stacked"
},
"deal_items": [
{
"uid": "blt3ea1b0c7d505e185",
"_content_type_uid": "deal_item"
},
{
"uid": "blt7cbdd8f42d317990",
"_content_type_uid": "deal_item"
},
{
"uid": "blt4019d2d57c5d735d",
"_content_type_uid": "deal_item"
},
{
"uid": "blt21af4abc87ccba18",
"_content_type_uid": "deal_item"
},
{
"uid": "blt7e20357fb7549664",
"_content_type_uid": "deal_item"
},
{
"uid": "bltacb034455a73cb53",
"_content_type_uid": "deal_item"
},
{
"uid": "blt587c8e2ad2466b21",
"_content_type_uid": "deal_item"
},
{
"uid": "blt08d2a4624e84e497",
"_content_type_uid": "deal_item"
}
],
"tags": [],
"locale": "de-de",
"_in_progress": false,
"_embedded_items": {
"text": []
},
"publish_details": [

],
"_rules": []
}`

One of the deal_item was not published and then we got this exception. It was solved by publishing the missing item, but the SDK should not throw an error in that case

Scenario

1. Create a new entry of a content type that includes an image entry
2. Publish the new entry WITHOUT references, so the image itself is not published.

ContentStack API will return "null" for that image when querying that content type
"image": null
DataSync API returns the uid of the referenced image
"image": "blt5ec250ecf650b59b"

This breaks some logic because "null" or "undefined" checks fail because the value is for example "blt5ec250ecf650b59b" instead of "null".

Same happens when you "unpublish" the image without publishing the content type where its referenced.

If a content type with a reference, which itself includes another reference, is published, but a nested reference is not. This causes a runtime exception

TypeError: Cannot read properties of undefined (reading 'hasOwnProperty')
at Stack. (/home/node/app/node_modules/datasync-filesystem-sdk/dist/stack.js:1556:61)
at Generator.next ()
at fulfilled (/home/node/app/node_modules/datasync-filesystem-sdk/dist/stack.js:11:58)"

Because the object that is checked here oldObjectPointerList[i].path[e]

is null if not all references are published. This happens if you use the blogQuery.includeReferences() which will include all nested references

There is a problem in the internal getReferencePath function

We have several content types where we have a "form" reference field, which itself has "fields" reference field.
When trying to fetch the fields entries its not working because the check is not working correctly for nested fields.
Simplified examples of the two content types
`{
"_content_type_uid": "pin_reset_page",
"locale": "de-de",
"unknown_customer": {
"form": [
{
"uid": "blt208e828db2bc1155",
"_content_type_uid": "form"
}
]
},
"updated_at": "2022-11-25T12:12:39.528Z",
"url": "/unified-mpp-pin-reset-page"
}

{
"_content_type_uid": "form",
"locale": "de-de",
"additional_text": "",
"fields": [{ "uid": "bltd0a76355d956c079", "_content_type_uid": "input_field" }],
"headline": "Jugendschutz-PIN vergessen?",
"tags": []

}
{
"_content_type_uid": "input_field",
"rules": [{ "uid": "blt58b06a7fe2b613b5", "_content_type_uid": "validation_rule" }],
"tags": [],
}
`

So when we call
blogQuery.includes(["unknown_customer.form", "unknown_customer.form.fields", " "unknown_customer.form.fields.rules"]);

We have the following values in the getReferencePath
currentInclude = [ 'unknown_customer', 'unknown_customer.form', 'unknown_customer.form.fields', 'unknown_customer.form.fields.rules' ] entryReferences = { 'unknown_customer.form': [ 'form' ] }
when we come to

This is a side effect of
https://github.com/contentstack/datasync-filesystem-sdk/pull/13/files#diff-1ebd2ec0fa0f7b0aac61f8683a96f14f9ba2f36cc068918c16b6c3aaf8418e0a

the part where the includePaths are checked

Just tested it with the old version, and that works.
But of course there was the other issue #10 that then breaks again

So we need a more sophisticated implementation here

I guess the problem is deeply nested references

@vkalta @netrajpatel @abhinav-from-contentstack

Stack.contentType("account_defaultpage").entries().include(["back_link.reference","content.content_link.reference"]).find()
will not include the reference for content.content_link.reference.
If I remove "back_link.reference", it works.

I could narrow it down to getReferencePath in stack.ts
Because the if path is taken instead of the else.
if (path.length !== includePath.length) { subPath = subStr; pendingPath.push(includePath.slice(path.length + 1)); } else { subPath = includePath; }
Added related _contents content to reproduce this.
_contents.zip

When it comes to check "content_link.reference" it matches with "content" and therefore "link.references" is pushed instead of "content_link.reference" because there is also a reference called "content"

Describe the bug

• .entry(uid) doesn't work
• .asset(uid) doesn't work

To Reproduce
Query a content type/asset
In the query, enter the entry's uid/asset's uid

Notice: The result is empty.

Query

Stack.contentType('blog')
  .entry({uid})

// OR

Stack.asset({uid})

Expected behavior
The query should return an object, if it exists

Workarounds
Use raw .query() for list-entries OR .entry()/.asset() without on singleton-content type

// example 1: On multiple content type
Stack.contentType('blog')
  .entries()
  .query({uid: {uid}})

// OR

Stack.assets()
  .query({uid: {uid}})

// example 2: On singleton content type
Stack.contentType('blog')
  .entry()

// OR

Stack.asset()

Regards,
Ramanathan K

Hello,

We encountered a problem while testing one of our components and traced it back to this file: GitHub link. The issue arises from the merging of lodash, which doesn't merge arrays as expected. You can find more details on the issue here.

Example interface:

{
    deals: {
        content: (rail | dealsSelection)[]
    };
    teaser: {
        content: (usp | teaserText)[]
    }
}

Below is the screenshot of schema created:

As you can see, both "deals" and "teaser" contain "content". When the schema is merged, it only retains the last "content", causing one of the references to be lost.

Our suggested fix, which worked for us while testing:
entryReferences = (0, lodash_1.mergeWith)(entryReferences, schema[this.types.references], (obj, src) => { if (lodash_1.isArray(obj)) { return Array.from(new Set(obj.concat(src))); } } );

Please let me know if this isn't clear. We can arrange a meeting to discuss the issue in more detail if needed. Thank you!

Regards,
Chaitra