Transformation Patterns

Transformation Patterns

Detailed dasel v3 patterns organized by use case. Each pattern follows the structure: problem, dasel command, expected result, caveats.

Table of Contents

• Config File Update Patterns
• Batch Processing Patterns
• Format Migration Patterns
• Data Reshaping Patterns
• Multi-File Operations
• Error Handling

Config File Update Patterns

YAML Config — Update Nested Values

Problem: Update a database host in a multi-level YAML config.

dasel -f config.yaml --root 'database.connection.host = "db-prod.example.com"' > tmp.yaml && mv tmp.yaml config.yaml

Expected: Full document output with only the target field changed.

YAML Config — Add New Section

Problem: Add an entirely new top-level section.

dasel -f config.yaml --root '{ $root..., "monitoring": {"enabled": true, "port": 9090} }' > tmp.yaml && mv tmp.yaml config.yaml

Expected: Original document preserved with new monitoring section appended.

TOML Settings — Modify Table Values

Problem: Update version in pyproject.toml.

dasel -f pyproject.toml --root 'project.version = "2.0.0"' > tmp.toml && mv tmp.toml pyproject.toml

TOML Settings — Add Array-of-Tables Entry

Problem: Add a new [[tool.ruff.per-file-ignores]] entry or similar array-of-tables.

dasel -f pyproject.toml --root 'tool.ruff.per-file-ignores = [$root.tool.ruff.per-file-ignores..., {"pattern": "tests/**", "ignore": ["S101"]}]' > tmp.toml && mv tmp.toml pyproject.toml

Caveat: TOML array-of-tables have specific serialization rules. Verify output format matches TOML spec expectations.

JSON Package Files — Update Version

Problem: Bump version in package.json.

dasel -f package.json --root 'version = "3.1.0"' > tmp.json && mv tmp.json package.json

JSON Package Files — Add Dependency

Problem: Add a new dependency to package.json.

dasel -f package.json --root '{ $root..., "dependencies": { $root.dependencies..., "lodash": "^4.17.21" } }' > tmp.json && mv tmp.json package.json

Batch Processing Patterns

Map Over Arrays

Problem: Extract a derived field from each element.

# Get full names from objects with firstName and lastName
dasel -f users.json 'users.map({ "full": firstName })'

Filter Then Transform

Problem: Find expensive items and apply a discount.

# Chain filter + each
dasel -f products.json --root '$expensive = products.filter(price > 100); $expensive.each(price = price * 0.9)'

Caveat: The each mutates in the query context. Use --root to get the full document with changes applied.

Batch Update with each

Problem: Set all users to inactive.

dasel -f data.json --root 'users.each(active = false)'

Expected: Every element in the users array has active set to false.

Batch Numeric Update

Problem: Increase all prices by 15%.

dasel -f catalog.json --root 'items.each(price = price * 1.15)'

Conditional Batch Update

Problem: Set status based on a threshold.

dasel -f data.json --root 'items.each(status = if(score > 80) { "pass" } else { "fail" })'

Format Migration Patterns

JSON to YAML

cat data.json | dasel -i json -o yaml > data.yaml

Caveats: Clean conversion. No data loss. YAML output uses dasel's default indentation (2 spaces).

YAML to TOML

cat config.yaml | dasel -i yaml -o toml > config.toml

Caveats:

• YAML anchors and aliases are resolved before conversion (dasel reads the resolved document)
• YAML multi-document files (--- separator) are not supported; only the first document is processed
• Null values in YAML may not have a TOML equivalent

JSON to XML

cat data.json | dasel -i json -o xml > data.xml

Caveats:

• XML requires a single root element. If JSON has multiple top-level keys, dasel wraps them in a root element
• JSON arrays map to repeated XML elements
• JSON null values may serialize as empty elements

CSV to JSON

cat data.csv | dasel -i csv -o json > data.json

Caveats:

• CSV header row becomes object keys
• All CSV values are strings; numeric conversion requires post-processing
• Empty cells become empty strings, not null

TOML to JSON

cat config.toml | dasel -i toml -o json > config.json

Caveats: TOML datetime types serialize as ISO 8601 strings in JSON.

Conversion Caveats Summary

Data Reshaping Patterns

Flatten Nested to Flat

Problem: Extract nested fields into a flat structure.

# From: {"user": {"name": "Alice", "address": {"city": "NYC"}}}
# To:   {"name": "Alice", "city": "NYC"}
echo '{"user":{"name":"Alice","address":{"city":"NYC"}}}' | \
  dasel -i json --root '{"name": user.name, "city": user.address.city}'

Nest Flat to Nested

Problem: Group flat fields into nested structure.

# From: {"db_host": "localhost", "db_port": 5432}
# To:   {"db": {"host": "localhost", "port": 5432}}
echo '{"db_host":"localhost","db_port":5432}' | \
  dasel -i json --root '{"db": {"host": db_host, "port": db_port}}'

Extract Subset

Problem: Pull out only specific fields from a large document.

dasel -f large.json '{ name, version, description }'

Array to Object Mapping

Problem: Transform array of key-value pairs to an object.

# Use map to extract, then reconstruct
dasel -f data.json 'settings.map({ key, value })'

Multi-File Operations

Shell Loop for Batch File Updates

Problem: Update a field across multiple YAML files.

for f in configs/*.yaml; do
  dasel -f "$f" --root 'metadata.version = "2.0"' > "${f}.tmp" && mv "${f}.tmp" "$f"
done

Convert Directory of JSON to YAML

for f in data/*.json; do
  cat "$f" | dasel -i json -o yaml > "${f%.json}.yaml"
done

Extract Same Field from Multiple Files

for f in services/*/config.yaml; do
  echo "$f: $(dasel -f "$f" 'service.port')"
done

Error Handling

Common Transformation Errors

Error: unknown key: fieldname

Cause: The field path does not exist in the document.

Fix: Verify path with exploration commands first:

dasel -f data.json 'keys($this)'
dasel -f data.json 'has("fieldname")'

Error: cannot use X as type Y

Cause: Assigning a value of wrong type (e.g., string to integer field).

Fix: Use type casting functions:

dasel -f data.json --root 'port = toInt("9090")'

Error: expected array, got object (or vice versa)

Cause: Path points to different type than expected.

Fix: Check type before operating:

dasel -f data.json 'typeOf(items)'

Error: Truncated/empty output file

Cause: Redirected output to the same input file (> same-file).

Fix: Always use temp file + rename pattern:

dasel -f input.json --root '...' > tmp.json && mv tmp.json input.json

Defensive Transformation Pattern

For critical files, validate before overwriting:

# 1. Transform to temp
dasel -f config.yaml --root 'server.port = 9090' > config_new.yaml

# 2. Verify temp is valid (non-empty, parseable)
dasel -f config_new.yaml 'keys($this)' > /dev/null && mv config_new.yaml config.yaml || echo "Transformation produced invalid output"

Dry Run Pattern

Preview transformation without writing:

# Just print — do not redirect
dasel -f config.yaml --root 'server.port = 9090'

Inspect the output. If correct, re-run with redirect.

Variable and Multi-Statement Patterns

Complex Transformations with Variables

# Store filtered subset, then transform
dasel -f data.json '$active = users.filter(active == true); $names = $active.map(name); $names'

Passing External Variables

Use --var to inject values from the shell:

VERSION="3.0.0"
dasel -f package.json --var "v=$VERSION" --root 'version = $v' > tmp.json && mv tmp.json package.json

Chaining with Shell Pipes

# Read JSON, transform, convert to YAML
cat data.json | dasel -i json --root 'items.each(processed = true)' | dasel -i json -o yaml > output.yaml

References

• Dasel Spread Syntax (fetched 2026-02-19)
• Dasel Conditionals (fetched 2026-02-19)
• Dasel each() (fetched 2026-02-19)
• Dasel filter() (fetched 2026-02-19)
• Dasel map() (fetched 2026-02-19)
• Dasel CHANGELOG (fetched 2026-02-19)