如何在DataWeave中不使用列名创建Excel文件的键值映射

To dynamically map Excel rows to objects using column indices (without hardcoding header names) and add validation errors, follow these steps:

1. Understand the Data Structure #

When you read an Excel file into MuleSoft, each row is an object where keys are the Excel column headers, and values are the cell contents. For example:

{"Column A": "valueA", "Column B": "valueB"}

To access columns by index, convert this object into an ordered array of key-value pairs using entriesOf(item). This gives you:

[{"key": "Column A", "value": "valueA"}, {"key": "Column B", "value": "valueB"}]

The order of entries matches the column order in your Excel file (A first, B second, etc.).

2. Dynamic Mapping + Column-Index Validation #

Use entriesOf to iterate over columns by index, perform validation, then reconstruct the final object. Here's a complete example:

%dw 2.0
output application/json

// Helper function to validate columns by index
fun validateColumn(index: Number, header: String, value: Any): String? = 
    index match {
        case 0 -> // Validate Column A (index 0)
            if (value is String and value != "") null else "$header cannot be empty"
        case 1 -> // Validate Column B (index 1)
            if (value is Number) null else "$header must be a number"
        // Add more cases for additional columns as needed
        else -> null
    }
---
payload map (row, rowIndex) -> 
    do {
        // Convert row object to ordered key-value entries
        var columnEntries = entriesOf(row)
        
        // Run validation on each column by index
        var validationErrors = columnEntries 
            map ((entry, colIndex) -> validateColumn(colIndex, entry.key, entry.value))
            filter ($ != null) // Keep only error messages
        
        // Format error string (empty if no errors)
        var errorMessage = if (sizeOf(validationErrors) > 0) joinBy(validationErrors, ", ") else ""
        
        // Reconstruct the row object + add Errors field
        ---
        fromEntries(columnEntries) ++ { Errors: errorMessage }
    }

3. Key Details #

• entriesOf(row): Converts the row object into an ordered array of key-value pairs, letting you access columns by their Excel index (0 for A, 1 for B, etc.).
• fromEntries(columnEntries): Converts the array of key-value pairs back into an object (this replaces your hardcoded key-value mappings dynamically).
• Validation Logic: The validateColumn function uses the column index to apply specific rules. Adjust this function to match your validation needs (e.g., regex checks, length limits, data types).

Why Your pluck $$ Attempt Didn't Work #

pluck $$ returns an array of the object's keys, but pluck itself returns an array of values—not the key-value pairs you need to build the map. entriesOf is the correct tool here because it preserves both the key (header name) and value for each column.

Example Output #

For a row where Column A is empty and Column B is a string:

{"Column A": "", "Column B": "not-a-number", "Errors": "Column A cannot be empty, Column B must be a number"}

内容的提问来源于stack exchange，提问作者anxiousAvocado