What are you trying to achieve?
I already have a JSON file containing ten news articles. I need to translate the title and body text into a new JSON file (with only the title and body text objects). I asked the AI to output the translated JSON files, and they will output them one by one. I should combine these into a new news article ->1 I have tried many ways to achieve the structure of 10, but still cannot. My foundation is too weak. Can you help me? Love comes from China
I hope the translated JSON looks like this
not this
blueprint (5).json (92.4 KB)
1 Like
Hello @Shanta_Fisher,
welcome to the Make community!
It looks like your Text aggregator is not pointing to the correct “Source Module”. Normally, it should point to the module that generated the bundles you want to group in one doc.
So, in your case, it looks like you need to have it point to your Iterator (with id=7).
As a summary:
- you split the initial array into separate bundles (one for each news articles)
- you do some cleaning, you ask Gemini to build some text, and you clean again
- and then you group all the results into one single document, with the Text Parser
- and then you do the extra stuff, such as parsing the resulting JSON String
Be careful, make sure that in your Text Aggregator and in your Parse Json, you add any relevant curly backets so that it becomes a properly formatted Json doc.
Another advice. While you develop your scenario, if you want to avoid consuming too many Make operations or Gemini tokens, you could add a filter to pick only the 2-3 first news articles.
In that case, you just need to add a filter after the Iterator. Like this
If you are still stuck, let me know, we’ll try and extract sample data so that I can test on my side.
Greetings from France! 
5 Likes
thank you Benjamin！！！！！Vous êtes le héros 
2 Likes
Sorry to bother you again, how to complete multiple regularization in one step.
1 Like
Hello @Shanta_Fisher,
can you please be more specific? What are you calling “multiple regularization”?
Benjamin
I want to complete multiple text parsers in one or two steps
@Shanta_Fisher it depends on what you want to achieve, and the complexity of the Replace you want to do.
For instance, if you want to replace data from a single document, maybe different fields, or a single field with multiple “Replace”, you could try and use the replace function instead. This function is directly used in mappings, and it alows you to use Regex as well, but with a few limitations (that’s why it depends on the use-case).
Here is an example where I use the replace function to search for H or E or L or P, replacing each of these characters with an emptystring.
You just need to surround the regex with slash characters, and add a flag such as g, m, s, i
1 Like
