Merge pull request #1 from embeddr-net/dev

v0.1.5

Author: Nynxz

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "embeddr-core"
-version = "0.1.4"
+version = "0.1.5"
 description = "Embeddr Core Library"
 readme = "README.md"
 authors = [{ name = "Nynxz", email = "contact@nynxz.com" }]

src/embeddr_core/models/library.py

@@ -1,6 +1,7 @@
 from datetime import datetime
 
 from sqlmodel import Field, Relationship, SQLModel
+from embeddr_core.models.lineage import ImageLineage
 
 
 class LibraryPath(SQLModel, table=True):
@@ -16,14 +17,39 @@ class LocalImage(SQLModel, table=True):
     id: int | None = Field(default=None, primary_key=True)
     path: str = Field(index=True, unique=True)
     filename: str
-    library_path_id: int | None = Field(default=None, foreign_key="librarypath.id")
+    library_path_id: int | None = Field(
+        default=None, foreign_key="librarypath.id")
     created_at: datetime = Field(default_factory=datetime.utcnow)
 
     # Metadata
     width: int | None = None
     height: int | None = None
     file_size: int | None = None
     mime_type: str | None = None
+    media_type: str = Field(default="image", index=True)
+    duration: float | None = None
+    fps: float | None = None
+    frame_count: int | None = None
     prompt: str | None = None
+    tags: str | None = None
+    phash: str | None = Field(default=None, index=True)
+    is_archived: bool = Field(default=False, index=True)
 
     library: LibraryPath | None = Relationship(back_populates="images")
+
+    parents: list["LocalImage"] = Relationship(
+        back_populates="children",
+        link_model=ImageLineage,
+        sa_relationship_kwargs={
+            "primaryjoin": "LocalImage.id==ImageLineage.child_id",
+            "secondaryjoin": "LocalImage.id==ImageLineage.parent_id",
+        },
+    )
+    children: list["LocalImage"] = Relationship(
+        back_populates="parents",
+        link_model=ImageLineage,
+        sa_relationship_kwargs={
+            "primaryjoin": "LocalImage.id==ImageLineage.parent_id",
+            "secondaryjoin": "LocalImage.id==ImageLineage.child_id",
+        },
+    )

src/embeddr_core/models/lineage.py

@@ -0,0 +1,10 @@
+from datetime import datetime
+from sqlmodel import Field, SQLModel
+
+
+class ImageLineage(SQLModel, table=True):
+    parent_id: int | None = Field(
+        default=None, foreign_key="localimage.id", primary_key=True)
+    child_id: int | None = Field(
+        default=None, foreign_key="localimage.id", primary_key=True)
+    created_at: datetime = Field(default_factory=datetime.utcnow)

src/embeddr_core/services/scanner.py

@@ -3,6 +3,8 @@
 import os
 from pathlib import Path
 
+import imagehash
+from PIL import Image
 from sqlmodel import Session, select
 
 from embeddr_core.models.library import LibraryPath, LocalImage
@@ -18,6 +20,9 @@ def scan_library_path(session: Session, library_path: LibraryPath) -> int:
     Returns the number of new images added.
     """
     root_path = Path(library_path.path)
+    lib_name = library_path.name or library_path.path
+    logger.info(f"Starting scan of library: {lib_name} ({root_path})")
+
     if not root_path.exists():
         logger.warning(f"Library path not found: {root_path}")
         return 0
@@ -28,17 +33,28 @@ def scan_library_path(session: Session, library_path: LibraryPath) -> int:
     # For large libraries, this might need optimization (e.g. set of paths)
     existing_paths = set(
         session.exec(
-            select(LocalImage.path).where(LocalImage.library_path_id == library_path.id)
+            select(LocalImage.path).where(
+                LocalImage.library_path_id == library_path.id)
         ).all()
     )
+    logger.info(
+        f"Found {len(existing_paths)} existing images in database for {lib_name}")
 
+    total_scanned = 0
     for root, dirs, files in os.walk(root_path):
         for file in files:
+            total_scanned += 1
+            if total_scanned % 100 == 0:
+                logger.info(
+                    f"Scanning {lib_name}: Checked {total_scanned} files, found {added_count} new images so far..."
+                )
+
             file_path = Path(root) / file
             if file_path.suffix.lower() in IMAGE_EXTENSIONS:
                 str_path = str(file_path)
 
                 if str_path in existing_paths:
+                    logger.debug(f"Skipping existing image: {file}")
                     continue
 
                 # Basic metadata
@@ -50,20 +66,39 @@ def scan_library_path(session: Session, library_path: LibraryPath) -> int:
 
                 mime_type, _ = mimetypes.guess_type(file_path)
 
+                width = None
+                height = None
+                phash = None
+
+                try:
+                    with Image.open(file_path) as img:
+                        width, height = img.size
+                        phash = str(imagehash.phash(img))
+                except Exception as e:
+                    logger.warning(
+                        f"Failed to process image metadata for {file_path}: {e}")
+
                 # Create image record
                 image = LocalImage(
                     path=str_path,
                     filename=file,
                     library_path_id=library_path.id,
                     file_size=file_size,
                     mime_type=mime_type,
+                    width=width,
+                    height=height,
+                    phash=phash,
                 )
                 session.add(image)
                 added_count += 1
+                logger.debug(f"Added new image: {file}")
 
                 # Commit in batches if needed, but for now simple
 
     session.commit()
+    logger.info(
+        f"Finished scanning {lib_name}. Total files checked: {total_scanned}. New images added: {added_count}."
+    )
     return added_count