FIX Include settings that change output in MathObfuscation, NegationTrap, SelectiveText and Zalgo converter identifiers

[Shivam Sharma (ShivamSharma43)]

Description

Fixes #3287.

Four text converters left settings that change their output out of their identifiers, so differently configured instances had the same hash and default registry name. For example, NegationTrapConverter(wrong_value="1234") and NegationTrapConverter(wrong_value="hunter2") collided, and registering the second one raised ValueError: Instance 'NegationTrapConverter::89a3d225' already exists.

Each converter now adds the missing settings:

• MathObfuscationConverter: hint_hash and suffix_hash, SHA-256 prefixes of the resolved hint and suffix, the same way AddTextImageConverter identifies text_to_add. An explicit default hint or suffix keeps the default identity, and "" (disabled) gets its own.
• NegationTrapConverter: a new _build_identifier() with wrong_value and trap_template_hash. Before this it had no params at all.
• SelectiveTextConverter: word_separator, only for word-level strategies, because character-level and token strategies never split on it. Their hashes are unchanged.
• ZalgoConverter: seed, like the other seeded converters. None params are not hashed, so unseeded Zalgo converters keep their hash.

Conversion behavior is unchanged. The default hashes of MathObfuscationConverter, NegationTrapConverter and word-level SelectiveTextConverter change once, because they now include these settings. Nothing in the repository pins these hashes.

Tests and Documentation

• New identifier tests in the four converter test files. Each checks that a different setting gets a different hash and that both instances register in ConverterRegistry, and that equal settings keep the same hash. SelectiveTextConverter also checks that character-level strategies ignore word_separator, and ZalgoConverter that an unseeded converter has no seed param. On main, 10 of the new tests fail. With this change, all of them pass.
• pytest tests/unit/converter tests/unit/registry tests/unit/backend/test_converter_service.py: 3734 passed, 38 skipped.
• ruff check, ruff format and ty check pass on the changed files. git diff --check is clean.
• No documentation changes. JupyText not run, since the converter docs do not use identifiers.

🤖 Generated with Claude Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[Manohar Paturi (ManoharPaturi)]

The identifier coverage across these four converters looks very thorough. Gating word separator in SelectiveTextConverter only when using word level selection strategies avoids invalidating hashes for character level transformations where the separator has no effect. Similarly, omitting seed when None preserves the unseeded hash for ZalgoConverter while distinguishing seeded runs. Checked registration against ConverterRegistry and the instance collision tests all pass.