feat: add binary title ID and save ID extraction during scan

DEVELOPER_SETUP.md

@@ -100,6 +100,20 @@ source .venv/bin/activate
 uv sync --all-extras --dev
 ```
 
+#### - Build and install sigil (optional)
+
+```sh
+# This is only required for title ID extraction; the feature no-ops without it
+# Users on macOS can skip this step, like RAHasher
+# Requires cmake, and RomM's Python 3.13 venv (created above) must be active
+git clone --recurse-submodules https://github.com/rommforge/argosy-sigil.git ../argosy-sigil
+cd ../argosy-sigil/bindings/python
+uv pip install cffi setuptools
+make build
+uv pip install -e .
+cd -
+```
+
 #### - Spin up the database and other services
 
 ```sh

Dockerfile

@@ -10,6 +10,7 @@ ENV DEBIAN_FRONTEND=noninteractive
 RUN apt-get update && apt-get install -y --no-install-recommends \
     git \
     make \
+    cmake \
     gcc \
     g++ \
     libmariadb3 \
@@ -77,6 +78,24 @@ RUN uv sync --all-extras
 
 ENV PATH="/app/.venv/bin:${PATH}"
 
+# Build and install sigil (optional, for title ID extraction)
+# Placed after `uv sync` because the extension is compiled with the venv's
+# Python so the ABI matches.
+ARG SIGIL_VERSION=98f3c626eecac48aa99209414d88a15ebe9b70c4
+RUN git clone --recurse-submodules https://github.com/rommforge/argosy-sigil.git /tmp/argosy-sigil
+WORKDIR /tmp/argosy-sigil
+RUN git checkout "${SIGIL_VERSION}" \
+    && git submodule update --init --recursive \
+    && cmake -B ./build-python -S . -DSIGIL_BUILD_CLI=OFF -DSIGIL_BUILD_TESTS=OFF \
+    && cmake --build ./build-python --target sigil
+WORKDIR /tmp/argosy-sigil/bindings/python
+RUN uv pip install --python /app/.venv/bin/python cffi setuptools \
+    && /app/.venv/bin/python build_sigil.py \
+    && mkdir -p /app/.venv/lib/python3.13/site-packages/sigil \
+    && cp ./sigil/*.py ./sigil/_sigil.*.so /app/.venv/lib/python3.13/site-packages/sigil/
+WORKDIR /app
+RUN rm -rf /tmp/argosy-sigil
+
 # Copy entrypoint script
 COPY entrypoint.sh /entrypoint.sh
 RUN chmod +x /entrypoint.sh

backend/adapters/services/sigil.py

@@ -0,0 +1,99 @@
+import asyncio
+from dataclasses import dataclass
+from typing import Any, Final
+
+from handler.metadata.base_handler import UniversalPlatformSlug as UPS
+from logger.logger import log
+
+try:
+    import sigil
+except ImportError:
+    sigil = None  # type: ignore[assignment]
+
+SWITCH_SIGIL_SLUG: Final = "switch"
+
+SIGIL_PLATFORM_SLUGS: Final[dict[UPS, str]] = {
+    UPS.PSP: "psp",
+    UPS.PSX: "psx",
+    UPS.PS2: "ps2",
+    UPS.PSVITA: "psvita",
+    UPS.SWITCH: SWITCH_SIGIL_SLUG,
+    UPS.SWITCH_2: SWITCH_SIGIL_SLUG,
+    UPS.N3DS: "3ds",
+    UPS.WII: "wii",
+    UPS.WIIU: "wiiu",
+    UPS.NGC: "gamecube",
+}
+
+# Errors that are expected for arbitrary library files (no title id present,
+# format sigil can't parse, missing decryption keys). Logged at debug level.
+ROUTINE_SIGIL_ERROR_CODES: Final = frozenset(
+    {"NOT_FOUND", "UNSUPPORTED", "UNSUPPORTED_FORMAT", "NEEDS_KEY"}
+)
+
+_missing_binding_logged = False
+
+
+class SigilServiceError(Exception): ...
+
+
+@dataclass(frozen=True)
+class SigilExtractionResult:
+    title_id: str
+    save_id: str
+    usage: str
+    content_type: str | None = None
+    version: int | None = None
+
+
+class SigilService:
+    """Service to extract platform-native title ids from ROM binaries via the
+    optional `sigil` cffi binding."""
+
+    async def extract_title_id(
+        self,
+        platform_slug: UPS | str,
+        file_path: str,
+        prod_keys_path: str | None = None,
+    ) -> SigilExtractionResult | None:
+        global _missing_binding_logged
+
+        if sigil is None:
+            if not _missing_binding_logged:
+                log.debug("sigil binding not installed, skipping title id extraction")
+                _missing_binding_logged = True
+            return None
+
+        sigil_slug = SIGIL_PLATFORM_SLUGS.get(platform_slug)  # type: ignore[arg-type]
+        if sigil_slug is None:
+            return None
+
+        kwargs: dict[str, Any] = {"platform": sigil_slug, "filename_fallback": False}
+        if sigil_slug == SWITCH_SIGIL_SLUG:
+            kwargs["prod_keys_path"] = prod_keys_path
+
+        try:
+            result = await asyncio.to_thread(sigil.extract, file_path, **kwargs)
+        except Exception as exc:
+            code = getattr(exc, "code", None)
+            is_sigil_error = isinstance(exc, getattr(sigil, "SigilError", ()))
+            if is_sigil_error and code in ROUTINE_SIGIL_ERROR_CODES:
+                log.debug(f"Sigil found no title id for {file_path}: {code}")
+            else:
+                log.error(f"Sigil extraction failed for {file_path}: {exc}")
+            return None
+
+        raw_content_type = getattr(result, "switch_content_type", None)
+        content_type = (
+            raw_content_type if raw_content_type not in (None, "", "unknown") else None
+        )
+
+        return SigilExtractionResult(
+            title_id=result.title_id,
+            save_id=result.save_id,
+            usage=result.usage,
+            content_type=content_type,
+            # Version 0 is a valid base-game version, so keep the int as-is;
+            # a missing field (non-Switch, older binding) maps to None.
+            version=getattr(result, "title_version", None),
+        )

backend/alembic/versions/0104_sigil_title_ids.py

@@ -0,0 +1,93 @@
+"""Add sigil-extracted title id columns: per-file title_id/save_id on
+rom_files, plus rom-level title_id/save_id/save_usage on roms.
+
+Revision ID: 0104_sigil_title_ids
+Revises: 0103_roms_facets_provider_ids
+Create Date: 2026-07-23 00:00:00.000000
+
+"""
+
+import sqlalchemy as sa
+from alembic import op
+from sqlalchemy.dialects.postgresql import ENUM
+
+from utils.database import is_postgresql
+
+# revision identifiers, used by Alembic.
+revision = "0104_sigil_title_ids"
+down_revision = "0103_roms_facets_provider_ids"
+branch_labels = None
+depends_on = None
+
+SAVE_USAGE_VALUES = ("FOLDER_EXACT", "FOLDER_PREFIX", "FILE_EXACT", "FILE_PREFIX")
+
+
+def _save_usage_enum(connection) -> sa.Enum:
+    if is_postgresql(connection):
+        enum = ENUM(*SAVE_USAGE_VALUES, name="saveusage", create_type=False)
+        enum.create(connection, checkfirst=True)
+        return enum
+    return sa.Enum(*SAVE_USAGE_VALUES, name="saveusage")
+
+
+def upgrade() -> None:
+    connection = op.get_bind()
+    save_usage_enum = _save_usage_enum(connection)
+
+    op.add_column(
+        "rom_files",
+        sa.Column("title_id", sa.String(length=100), nullable=True),
+        if_not_exists=True,
+    )
+    op.add_column(
+        "rom_files",
+        sa.Column("save_id", sa.String(length=100), nullable=True),
+        if_not_exists=True,
+    )
+    op.add_column(
+        "roms",
+        sa.Column("title_id", sa.String(length=100), nullable=True),
+        if_not_exists=True,
+    )
+    op.add_column(
+        "roms",
+        sa.Column("save_id", sa.String(length=100), nullable=True),
+        if_not_exists=True,
+    )
+    op.add_column(
+        "roms",
+        sa.Column("save_usage", save_usage_enum, nullable=True),
+        if_not_exists=True,
+    )
+
+    with op.batch_alter_table("rom_files", schema=None) as batch_op:
+        batch_op.create_index(
+            "idx_rom_files_title_id",
+            ["title_id"],
+            unique=False,
+            if_not_exists=True,
+        )
+    with op.batch_alter_table("roms", schema=None) as batch_op:
+        batch_op.create_index(
+            "idx_roms_title_id",
+            ["title_id"],
+            unique=False,
+            if_not_exists=True,
+        )
+
+
+def downgrade() -> None:
+    with op.batch_alter_table("roms", schema=None) as batch_op:
+        batch_op.drop_index("idx_roms_title_id", if_exists=True)
+    with op.batch_alter_table("rom_files", schema=None) as batch_op:
+        batch_op.drop_index("idx_rom_files_title_id", if_exists=True)
+
+    op.drop_column("roms", "save_usage", if_exists=True)
+    op.drop_column("roms", "save_id", if_exists=True)
+    op.drop_column("roms", "title_id", if_exists=True)
+    op.drop_column("rom_files", "save_id", if_exists=True)
+    op.drop_column("rom_files", "title_id", if_exists=True)
+
+    connection = op.get_bind()
+    if is_postgresql(connection):
+        ENUM(name="saveusage").drop(connection, checkfirst=True)

backend/alembic/versions/0105_sigil_title_version.py

@@ -0,0 +1,29 @@
+"""Add sigil-extracted title_version column to rom_files (Switch title
+versions are u32, stored as BigInteger).
+
+Revision ID: 0105_sigil_title_version
+Revises: 0104_sigil_title_ids
+Create Date: 2026-07-24 00:00:00.000000
+
+"""
+
+import sqlalchemy as sa
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision = "0105_sigil_title_version"
+down_revision = "0104_sigil_title_ids"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "rom_files",
+        sa.Column("title_version", sa.BigInteger(), nullable=True),
+        if_not_exists=True,
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("rom_files", "title_version", if_exists=True)

backend/alembic/versions/0106_sigil_folder_split.py

@@ -0,0 +1,59 @@
+"""Add FOLDER_SPLIT to the saveusage enum (3DS saves nest the id across
+two directory levels).
+
+Revision ID: 0106_sigil_folder_split
+Revises: 0105_sigil_title_version
+Create Date: 2026-07-29 00:00:00.000000
+
+"""
+
+import sqlalchemy as sa
+from alembic import op
+
+from utils.database import is_postgresql
+
+# revision identifiers, used by Alembic.
+revision = "0106_sigil_folder_split"
+down_revision = "0105_sigil_title_version"
+branch_labels = None
+depends_on = None
+
+SAVE_USAGE_VALUES = (
+    "FOLDER_EXACT",
+    "FOLDER_PREFIX",
+    "FILE_EXACT",
+    "FILE_PREFIX",
+    "FOLDER_SPLIT",
+)
+
+
+def upgrade() -> None:
+    connection = op.get_bind()
+
+    if is_postgresql(connection):
+        # ALTER TYPE ... ADD VALUE must run outside alembic's transaction.
+        with op.get_context().autocommit_block():
+            op.execute("ALTER TYPE saveusage ADD VALUE IF NOT EXISTS 'FOLDER_SPLIT'")
+    else:
+        save_usage_enum = sa.Enum(*SAVE_USAGE_VALUES, name="saveusage")
+        with op.batch_alter_table("roms", schema=None) as batch_op:
+            batch_op.alter_column("save_usage", type_=save_usage_enum, nullable=True)
+
+
+def downgrade() -> None:
+    # PostgreSQL cannot drop enum values; leave FOLDER_SPLIT in place. On
+    # MariaDB, narrow the enum back after clearing any rows that used it.
+    connection = op.get_bind()
+    if is_postgresql(connection):
+        return
+
+    op.execute("UPDATE roms SET save_usage = NULL WHERE save_usage = 'FOLDER_SPLIT'")
+    reverted_enum = sa.Enum(
+        "FOLDER_EXACT",
+        "FOLDER_PREFIX",
+        "FILE_EXACT",
+        "FILE_PREFIX",
+        name="saveusage",
+    )
+    with op.batch_alter_table("roms", schema=None) as batch_op:
+        batch_op.alter_column("save_usage", type_=reverted_enum, nullable=True)

backend/config/config_manager.py

@@ -185,6 +185,8 @@ class Config:
     ROMS_FOLDER_NAME: str
     FIRMWARE_FOLDER_NAME: str
     SKIP_HASH_CALCULATION: bool
+    SKIP_TITLE_ID_EXTRACTION: bool
+    EMBED_SWITCH_TITLE_IDS: bool
     EJS_DEBUG: bool
     EJS_CACHE_LIMIT: int | None
     EJS_DISABLE_AUTO_UNLOAD: bool
@@ -437,6 +439,12 @@ def _parse_config(self):
             SKIP_HASH_CALCULATION=pydash.get(
                 self._raw_config, "filesystem.skip_hash_calculation", False
             ),
+            SKIP_TITLE_ID_EXTRACTION=pydash.get(
+                self._raw_config, "filesystem.skip_title_id_extraction", False
+            ),
+            EMBED_SWITCH_TITLE_IDS=pydash.get(
+                self._raw_config, "filesystem.embed_switch_title_ids", False
+            ),
             EJS_DEBUG=pydash.get(self._raw_config, "emulatorjs.debug", False),
             EJS_CACHE_LIMIT=pydash.get(
                 self._raw_config, "emulatorjs.cache_limit", None
@@ -873,6 +881,8 @@ def _update_config_file(self) -> None:
                 "roms_folder": self.config.ROMS_FOLDER_NAME,
                 "firmware_folder": self.config.FIRMWARE_FOLDER_NAME,
                 "skip_hash_calculation": self.config.SKIP_HASH_CALCULATION,
+                "skip_title_id_extraction": self.config.SKIP_TITLE_ID_EXTRACTION,
+                "embed_switch_title_ids": self.config.EMBED_SWITCH_TITLE_IDS,
             },
             "system": {
                 "platforms": self.config.PLATFORMS_BINDING,