MorphQL Language Reference

This document provides a complete reference for the Morph Query Language (MorphQL), a declarative DSL for structural data transformation.

Query Structure

A MorphQL query follows this basic structure:

from <source_format> to <target_format>
[transform [unsafe]
  <actions>
]

Supported formats: json, xml, object, csv, plaintext

The transform block is optional—omit it for pure format conversion:

from json to xml

Safe Mode vs Unsafe Mode

By default, MorphQL generates safe code that uses optional chaining (?.) to prevent crashes when accessing properties on null/undefined values. For performance-critical scenarios with validated data, you can use the unsafe keyword:

// Safe mode (default) - uses optional chaining
from object to object
transform
  set result = price / quantity

// Unsafe mode - no optional chaining, maximum performance
from object to object
transform unsafe
  set result = price / quantity

WARNING

Use unsafe only with validated/trusted input data. Unsafe mode will crash if it encounters null/undefined values.

---

Context Identifiers

MorphQL provides special identifiers to control which data context you're reading from or writing to.

Default Contexts

• source - The input data (implicit when reading in set)
• target - The output data (implicit when reading in modify or return)

// These are equivalent - 'source' is implicit in set
set name = firstName
set name = source.firstName

// These are equivalent - 'target' is implicit in modify
modify total = total * 1.1
modify total = target.total * 1.1

Explicit Context Access

You can explicitly specify which context to read from:

// Read from source explicitly
set result = source.price

// Read from target (useful in set to reference previously set fields)
set first = value
set second = target.first  // References the field we just set

// Mix contexts
set total = source.price + target.markup

Root Context Identifiers

When working inside nested sections, the source and target identifiers refer to the current scope. To access the root/parent scope, use _source and _target:

from json to object
transform
  set globalId = id  // Root source.id

  section multiple items(
    set itemId = id           // Current item's id
    set parentId = _source.id // Root source.id (parent scope)
    set globalRef = _target.globalId // Access root target field
  ) from list

Use cases for root identifiers:

• Access parent data from within nested sections
• Reference root-level computed fields
• Cross-scope data relationships

TIP

_source and _target always refer to the outermost transformation scope, even when deeply nested in multiple sections or subqueries.

Iteration Index: _key

When inside a section multiple block, you can use the _key identifier to access the current iteration index (starting from 0).

section multiple items(
  set rank = _key + 1
  set name = productName
) from products

The _key is also available in the where clause of a section to filter by position:

section multiple firstThree(
  set val = source
) from data where _key < 3

---

Actions

Actions are the commands used inside the transform block or section blocks.

set

Sets a property on the target object by reading from the source data.

set fullName = firstName + " " + lastName
set total = (price * quantity) - discount
set shortSku = substring(sku, 0, 3)

modify

Modifies a property on the target object by reading from the target object itself. This is useful for post-processing values that have already been mapped.

set total = price * quantity
modify total = total * 1.10   # Apply a 10% markup to the already calculated total

section

Creates a nested object or array. Supports optional format conversion via subqueries.

Syntax:

section [multiple] <name>( [subquery] <actions> ) [from <path>] [where <condition>]

Parameters:

• multiple: Treats the source as an array and maps each item
• subquery: Optional nested query for format conversion
• from <expression>: Shifts the context to a specific source data source. Can be a path, the parent keyword, or a function call (e.g., spreadsheet(source)).
• where <condition>: Filters the source data before transformation. For multiple sections, only matching items are included. For single sections, the first matching item is used. You can use _key here to filter by index.

Examples:

// Simple nested object
section header(
  set id = orderId
  set date = orderDate
)

// Array mapping
section multiple items(
  set sku = itemSku
  set qty = quantity
) from orderItems

// Filtering with where clause
section multiple activeUsers(
  set name = userName
  set email = userEmail
) from users where status == "active"

// Single section with where (find first match)
section primaryContact(
  set name = contactName
  set phone = contactPhone
) from contacts where isPrimary == true

// Subquery with format conversion
section metadata(
  from xml to object
  transform
    set name = root.productName
    set price = number(root.cost)
) from xmlDataField

// Using an expression in 'from'
section multiple records(
  set val = source
) from concat(list1, list2)

clone

Clones the entire source object or specific fields into the target.

// Clone everything
clone()

// Clone specific fields
clone(firstName, lastName, email)

delete

Removes a property from the target object. Useful after clone().

clone()
delete password
delete internalId

define

Defines a local variable/alias for use in subsequent expressions within the same scope.

define basePrice = price * quantity
set subtotal = basePrice
set total = basePrice - discount

if (Conditional Block)

Executes actions conditionally.

if (status == "active") (
  set isActive = true
  set badge = "✓"
) else (
  set isActive = false
  set badge = "✗"
)

return

Overwrites the target object and returns immediately from the current scope. Expressions inside return read from the target object by default.

// Top level
set greeting = "Hello " + name
return greeting

// Inside section
section user(
  set id = userId
  set name = userName
  return target
)

---

Escaped Identifiers

Use backticks (`fieldname`) to use reserved keywords or special characters as identifiers:

transform
  set `multiple` = true
  set `order-id` = root.`external-id`
  set `field with spaces` = source.`another field`

Backticks can be escaped with \ when needed in a field name.

---

Functions

Functions are used within expressions to calculate or transform values. For full documentation with examples and edge cases, see the Functions Reference.

Function	Description
substring(str, start, [length])	Extracts a part of a string. Supports negative indices.
replace(str, search, replacement)	Replaces the first occurrence in a string.
split(str, [sep], [limit])	Splits a string into an array.
uppercase(str) / lowercase(str)	Converts case.
trim(str)	Removes leading and trailing whitespace.
padstart(str, len, [char]) / padend(str, len, [char])	Pads a string to a target length.
indexof(str, search)	Returns index of first occurrence, or -1.
startswith(str, prefix) / endswith(str, suffix)	Prefix/suffix check.
extractnumber(str)	Extracts the first numeric sequence from a string.
text(val)	Converts a value to a string.
number(val)	Converts a value to a number.
to_base64(val) / from_base64(val)	Base64 encode/decode (isomorphic).
floor(val) / ceil(val)	Rounds down or up to the nearest integer.
round(val, [mode])	Rounds to nearest integer ("half-up" or "half-even").
abs(val)	Absolute value.
fixed(val, [decimals], [mode])	Formats to fixed decimal places. Returns a string.
min(a, b, ...) / max(a, b, ...)	Smallest/largest of scalar values.
if(cond, trueVal, falseVal)	Inline ternary expression.
lookup(val, "k:v", ...)	Dictionary lookup, returning mapped value or null.
fromUnix(timestamp)	Converts Unix timestamp string to ISO 8601 string.
toUnix(isoString)	Converts ISO 8601 string to Unix timestamp string.
list(v1, ...) / array(v1, ...)	Creates an array literal.
aslist(val)	Ensures a value is an array.
concat(arr1, arr2, ...)	Concatenates arrays.
distinct(array, expr)	Unique values of an expression, first-seen order.
transpose(source, key1, ...)	Zips parallel arrays into row objects.
spreadsheet(array)	Converts CSV-parsed rows using first row as headers.
sum(array, expr)	Sums a numeric expression per item.
avg(array, expr)	Average of a numeric expression per item.
minof(array, expr) / maxof(array, expr)	Min/max of a numeric expression per item.
every(array, expr) / some(array, expr)	Boolean aggregates over an array.
groupby(array, expr)	Groups array into [{ key, items }] objects.
extract(obj, spec...)	Extracts and renames fields from an object.
unpack(str, spec...) / pack(obj, spec...)	Fixed-width string read/write.
xmlnode(val, [attrKey, attrVal...])	Wraps a value for XML output with attributes.

---

Operators

Arithmetic

Operator	Description
+	Addition / String concatenation
-	Subtraction / Unary minus
*	Multiplication
/	Division
%	Modulo / Remainder

Comparison

Operator	Description
==	Loose equality
===	Strict equality
!=	Loose inequality
!==	Strict inequality
<	Less than
>	Greater than
<=	Less than or equal
>=	Greater than or equal

Logical

Operator	Description
&&	Logical AND
||	Logical OR
!	Logical NOT

Grouping

Use parentheses ( ) to control operator precedence:

set total = (price * quantity) - discount
set isValid = (status == "active") && (count > 0)

---

Complete Examples

E-commerce Order Transformation

from json to json
transform
  set orderId = id
  set customerName = customer.firstName + " " + customer.lastName

  section multiple lineItems(
    set sku = substring(productCode, 0, 6)
    set unitPrice = number(price)
    set total = number(price) * quantity
  ) from items

  section summary(
    set itemCount = items.length
    set status = if(isPaid, "Paid", "Pending")
  )

XML to JSON Conversion with Mapping

from xml to json
transform
  define product = root.catalog.product
  set name = product.name
  set price = number(product.price)
  set inStock = product.stock > 0

  section multiple variants(
    set color = variant.color
    set size = uppercase(variant.size)
  ) from product.variants

Spreadsheet Transformation (CSV)

Useful when working with CSV data where the first row contains values that should be mapped as keys for subsequent rows.

from csv to json
transform
  // spreadsheet() converts an array of objects (like CSV rows)
  // into a structured array using the first row values as keys.
  set data = spreadsheet(rows)

Format Conversion Only

from json to xml

from xml to object