Parsers (Import Layer)¶

All parsers inherit from the abstract Parser class and register themselves via @ParserRegistry.register().

Class Hierarchy¶

classDiagram
    class Parser {
        <<abstract>>
        +parse(args, schema)*
        +copy_values(obj, dict)
    }

    class TransformableParser {
        +store_data(schema, table, record)
        +map_record(record, mapper)
    }

    class XMIParser {
        +phase1_process_packages_classes()
        +phase2_process_connectors()
    }

    class EAXMIParser {
        EA-specific XMI extensions
        ~600 lines
    }

    class QEAParser {
        QEA database format
    }

    class JSONParser {
        Remote URL support
        Column mapping
    }

    class XLSXParser {
        Pandas-based
        Sheet per table
    }

    class CSVParser {
        entity_name mapping
    }

    class I18nParser {
        Language-specific extraction
    }

    Parser <|-- XMIParser
    Parser <|-- EAXMIParser
    Parser <|-- QEAParser
    Parser <|-- TransformableParser
    TransformableParser <|-- JSONParser
    TransformableParser <|-- XLSXParser
    TransformableParser <|-- CSVParser
    TransformableParser <|-- I18nParser

Registered Parsers¶

XMI Parser¶

Registration: @ParserRegistry.register("xmi")
File: parsers/xmiparser.py
Function: Standard XMI 2.1 format
Features: Two-phase parsing (structure → relationships), XML encoding detection & recovery

EA XMI Parser¶

Registration: @ParserRegistry.register("eaxmi")
File: parsers/eaxmiparser.py
Function: Enterprise Architect XMI with EA-specific extensions
Features: Processes diagrams, extended tags, EA metadata. Approximately 600 lines.

QEA Parser¶

Registration: @ParserRegistry.register("qea")
File: parsers/qeaparser.py
Function: QEA database format (EA native)

JSON Parser¶

Registration: @ParserRegistry.register("json")
File: parsers/multiple_parsers.py
Function: JSON with table names as keys, arrays of records
Features: Remote URL support, --mapper for column renaming, --update_only mode

XLSX Parser¶

Registration: @ParserRegistry.register("xlsx")
Function: Excel files with one sheet per table
Implementation: Pandas-based row-by-row processing

CSV Parser¶

Registration: @ParserRegistry.register("csv")
Function: Single CSV file with --entity_name for target table

i18n Parser¶

Registration: @ParserRegistry.register("i18n")
Function: Language-specific data extraction, integration with translation fields

CLI Arguments (Import)¶

Argument	Description
`-f / --inputfile`	Path to input file
`-url`	Remote URL (JSON)
`-t / --inputtype`	Parser type (xmi, eaxmi, qea, json, xlsx, csv, i18n)
`--skip_xmi_relations`	Skip phase 2 (structure only)
`--mapper`	JSON string for column renaming
`--update_only`	Update existing records only

Planned Extensions¶

API Endpoint Parser

Direct integration with external model repositories via REST APIs.

Streaming / Chunked Parser

Processing of large files in chunks to limit memory usage.

Adding a New Parser¶

from crunch_uml.parsers.parser import Parser, ParserRegistry

@ParserRegistry.register("my_format", descr="My custom parser")
class MyParser(Parser):
    def parse(self, args, schema):
        # Read input
        # Create ORM objects
        # schema.save(obj, recursive=True)
        pass