Overview

Summary

Objects in the ethdebug/format/info schema organize debugging data associated with a particular compilation. This includes lookup tables of types and pointers by their unique name identifiers, as well as optionally may include additional information such as complete program listings and compiled source contents.

This schema serves as a root schema, in that it (a) aggregates all other schemas in this format (either directly or indirectly), and (b) in that no other schema aggregates it.

This schema serves as a suitable target for JSON Schema bundling, e.g., for implementations that want to reduce complexity around validating ethdebug/format records and don't want to handle reference resolution across schema YAML files.

warning

Note, however, that "root schema" does not imply "objects in the schema always contain all relevant debugging data". Due to widespread practice among EVM language compilers to produce their own structured JSON output, objects in this schema are not necessarily complete records of debug information.

Compilers may choose to divide distinct portions of debugging data into distinct sections of compiler JSON output. For instance, it likely makes sense for a compiler to produce debugging data about a particular contract's bytecode alongside that contract bytecode's other output information.

At least until tooling accommodates this concern, debugging implementers must be aware that they will likely need to aggregate debugging data from multiple sections of each supported compiler's own output format.

However, this schema is suitable for representing a complete and standalone debugging data record for a particular compilation, but note the optionality of fields such as compilation and programs. Implementations looking to produce such self-contained debugging data objects should study the structure of these (and any other optional fields) and ensure that such fields are fully populated.