Skip to content

grammar.md

Latest commit

 

History

History
108 lines (68 loc) · 2.11 KB

grammar.md

File metadata and controls

108 lines (68 loc) · 2.11 KB

BC3 Grammar Rules

This document defines the core lexical and structural grammar rules used by BC3 (FIEBDC-3) files. These rules are the foundation for tokenization and parsing.

1. Record delimiter

  • Each record starts with ~ (tilde).
  • The first character after ~ defines the record type (e.g. V, C, D, M).
  • Records end at the start of the next ~ or end-of-file.

Example:

~V|FIEBDC-3|2020
~C|01|Concrete

2. Field delimiter (|)

  • Fields inside a record are separated by the pipe character |.
  • Empty fields are allowed and represented by consecutive delimiters.

Example:

~C|01||Concrete

Here:

  • Field 1 = 01
  • Field 2 = empty
  • Field 3 = Concrete

3. Subfield delimiter (\\)

  • Subfields inside a field are separated by a backslash (\\).
  • Subfields preserve order and meaning as defined by the record type.

Example:

~C|01\\A|Concrete

Subfields of field 1:

  • 01
  • A

4. Whitespace handling

  • Leading and trailing whitespace must be ignored for fields and subfields.
  • Whitespace inside text values is preserved.

Example:

~C| 01 |  Concrete block

Parsed as:

  • Code: 01
  • Description: Concrete block

5. Empty and optional values

  • Empty fields and subfields are valid unless explicitly forbidden by the record specification.
  • Parsers must not fail on missing optional fields in lenient mode.

6. Unknown records

  • Records with unknown types must be:
    • Ignored in lenient mode
    • Reported as errors in strict mode

7. Line endings

  • Files may use \n, \r\n, or mixed line endings.
  • Line endings must not affect parsing logic.

8. Summary

Element Symbol
Record start ~
Field separator ` `
Subfield separator \\
Whitespace Trimmed (outer)

These grammar rules apply uniformly to all BC3 record types and are used by the tokenizer stage of the parsing pipeline.