Text Proto Formatter - Configuration

April 20, 2026 · View on GitHub

txtpbfmt provides several configuration options that customize the specifics of the output format. These are configured by adding a comment line to top of the proto file (before the first non-empty non-comment line) of the form:

# txtpbfmt: [config-option]

This doc describes each of these options.

AllowTripleQuotedStrings

# txtpbfmt: allow_triple_quoted_strings

Permit usage of Python-style """ or ''' delimited strings.

AllowUnnamedNodesEverywhere

# txtpbfmt: allow_unnamed_nodes_everywhere

Allow unnamed nodes everywhere. Default is to allow only top-level nodes to be unnamed.

ExpandAllChildren

# txtpbfmt: expand_all_children

Expand all children irrespective of the initial state.

Before formatting

Example

After formatting

Example

PreserveAngleBrackets

# txtpbfmt: preserve_angle_brackets

Whether angle brackets used instead of curly braces should be preserved when outputting a formatted textproto.

Before formatting

Example

After formatting

Example

RemoveDuplicateValuesForRepeatedFields

# txtpbfmt: remove_duplicate_values_for_repeated_fields

Remove lines that have the same field name and scalar value as another.

Before formatting

Example

After formatting

Example

SkipAllColons

# txtpbfmt: skip_all_colons

Skip colons whenever possible.

Before formatting

Example

After formatting

Example

SmartQuotes

# txtpbfmt: smartquotes

Use single quotes around strings that contain double but not single quotes.

Before formatting

Example

After formatting

Example

SortFieldsByFieldName

# txtpbfmt: sort_fields_by_field_name

Sort fields by field name.

Before formatting

Example

After formatting

Example

SortRepeatedFieldsByContent

# txtpbfmt: sort_repeated_fields_by_content

Sort adjacent scalar fields of the same field name by their contents.

Before formatting

Example

After formatting

Example

SortRepeatedFieldsBySubfield

# txtpbfmt: sort_repeated_fields_by_subfield=[subfieldSpec]

Sort adjacent message fields of the given field name by the contents of the given subfield path.

subfieldSpec is of one of two forms:

fieldName.subfieldName1.subfieldName2...subfieldNameN, which will sort fields named fieldName by the value of the final subfield named subfieldNameN.
subfieldName, which will sort any field by its subfield named subfieldName

Before formatting

Example

After formatting

Example

FieldSortOrder

# txtpbfmt: field_sort_order=[NodeName]:[field1] [field2] ...

Specify a custom sort order for fields within a specific node type. NodeName is the name of the message node. Use __ROOT__ to specify the sort order for top-level fields in the file. Fields not listed will bubble to the top (unless require_field_sort_order_to_match_all_fields_in_node is set).

Example

# txtpbfmt: field_sort_order=MyNode:foo bar baz

MyNode {
  bar: 1
  foo: 2
  baz: 3
}

After formatting:

# txtpbfmt: field_sort_order=MyNode:foo bar baz

MyNode {
  foo: 2
  bar: 1
  baz: 3
}

RequireFieldSortOrderToMatchAllFieldsInNode

# txtpbfmt: require_field_sort_order_to_match_all_fields_in_node

Enforce that all fields present in a node must be specified in the field_sort_order for that node. If a field is found that is not in the sort order, formatting will fail with an error.

ReverseSort

# txtpbfmt: reverse_sort

Sorts all sort_* fields in descending order instead of the default ascending order. Does nothing if not used with at least 1 other sort_* field.

Before formatting

Example

After formatting

Example

DNSSortOrder

# txtpbfmt: dns_sort_order

When sorting field contents, sorts in a way intended for DNS names. It splits the contents on '.' characters, reverses the substrings, and sorts on their concatenation. This puts, e.g., a.com and c.com next to each other even if b.au is in the list.