commit template changes
This commit is contained in:
2
docengine/.gitignore
vendored
2
docengine/.gitignore
vendored
@@ -58,7 +58,7 @@ ENV/
|
|||||||
*~
|
*~
|
||||||
.DS_Store
|
.DS_Store
|
||||||
Thumbs.db
|
Thumbs.db
|
||||||
storage/
|
/storage/
|
||||||
*.pid
|
*.pid
|
||||||
celerybeat-schedule
|
celerybeat-schedule
|
||||||
celerybeat.pid
|
celerybeat.pid
|
||||||
|
|||||||
1
docengine/app/storage/__init__.py
Normal file
1
docengine/app/storage/__init__.py
Normal file
@@ -0,0 +1 @@
|
|||||||
|
# Storage module
|
||||||
163
docengine/app/storage/provider.py
Normal file
163
docengine/app/storage/provider.py
Normal file
@@ -0,0 +1,163 @@
|
|||||||
|
from __future__ import annotations
|
||||||
|
|
||||||
|
import hashlib
|
||||||
|
import os
|
||||||
|
import shutil
|
||||||
|
import uuid
|
||||||
|
from abc import ABC, abstractmethod
|
||||||
|
from pathlib import Path
|
||||||
|
|
||||||
|
from app.core.config import settings
|
||||||
|
from app.core.exceptions import StorageError
|
||||||
|
from app.core.logging_config import get_logger
|
||||||
|
|
||||||
|
logger = get_logger(__name__)
|
||||||
|
|
||||||
|
|
||||||
|
class StorageProvider(ABC):
|
||||||
|
"""Abstract base class for storage providers."""
|
||||||
|
|
||||||
|
@abstractmethod
|
||||||
|
def save_file(self, file_data: bytes, directory: str, filename: str | None = None) -> str:
|
||||||
|
"""Save file data and return the storage path."""
|
||||||
|
...
|
||||||
|
|
||||||
|
@abstractmethod
|
||||||
|
def read_file(self, storage_path: str) -> bytes:
|
||||||
|
"""Read file data from storage."""
|
||||||
|
...
|
||||||
|
|
||||||
|
@abstractmethod
|
||||||
|
def delete_file(self, storage_path: str) -> bool:
|
||||||
|
"""Delete a file from storage. Returns True if successful."""
|
||||||
|
...
|
||||||
|
|
||||||
|
@abstractmethod
|
||||||
|
def file_exists(self, storage_path: str) -> bool:
|
||||||
|
"""Check if a file exists in storage."""
|
||||||
|
...
|
||||||
|
|
||||||
|
@abstractmethod
|
||||||
|
def get_file_size(self, storage_path: str) -> int:
|
||||||
|
"""Get the file size in bytes."""
|
||||||
|
...
|
||||||
|
|
||||||
|
@abstractmethod
|
||||||
|
def get_absolute_path(self, storage_path: str) -> str:
|
||||||
|
"""Get the absolute filesystem path for a storage path."""
|
||||||
|
...
|
||||||
|
|
||||||
|
@staticmethod
|
||||||
|
def compute_checksum(data: bytes, algorithm: str = "sha256") -> str:
|
||||||
|
"""Compute checksum of file data."""
|
||||||
|
hasher = hashlib.new(algorithm)
|
||||||
|
hasher.update(data)
|
||||||
|
return hasher.hexdigest()
|
||||||
|
|
||||||
|
@staticmethod
|
||||||
|
def generate_filename(original_filename: str) -> str:
|
||||||
|
"""Generate a unique filename preserving the original extension."""
|
||||||
|
ext = Path(original_filename).suffix.lower()
|
||||||
|
return f"{uuid.uuid4().hex}{ext}"
|
||||||
|
|
||||||
|
|
||||||
|
class LocalStorageProvider(StorageProvider):
|
||||||
|
"""Local filesystem storage provider."""
|
||||||
|
|
||||||
|
def __init__(self, base_path: str | None = None) -> None:
|
||||||
|
self.base_path = Path(base_path or settings.storage_local_path).resolve()
|
||||||
|
self._ensure_directories()
|
||||||
|
|
||||||
|
def _ensure_directories(self) -> None:
|
||||||
|
"""Create required storage directories."""
|
||||||
|
for subdir in ("documents", "templates", "images", "temp", "rendered"):
|
||||||
|
(self.base_path / subdir).mkdir(parents=True, exist_ok=True)
|
||||||
|
|
||||||
|
def _resolve_path(self, storage_path: str) -> Path:
|
||||||
|
"""Resolve a storage path to an absolute path."""
|
||||||
|
resolved = (self.base_path / storage_path).resolve()
|
||||||
|
if not str(resolved).startswith(str(self.base_path)):
|
||||||
|
raise StorageError(f"Path traversal detected: {storage_path}")
|
||||||
|
return resolved
|
||||||
|
|
||||||
|
def save_file(self, file_data: bytes, directory: str, filename: str | None = None) -> str:
|
||||||
|
"""Save file data to local storage."""
|
||||||
|
if filename is None:
|
||||||
|
filename = f"{uuid.uuid4().hex}.bin"
|
||||||
|
|
||||||
|
dir_path = self.base_path / directory
|
||||||
|
dir_path.mkdir(parents=True, exist_ok=True)
|
||||||
|
|
||||||
|
file_path = dir_path / filename
|
||||||
|
try:
|
||||||
|
file_path.write_bytes(file_data)
|
||||||
|
storage_path = str(file_path.relative_to(self.base_path))
|
||||||
|
logger.info("file_saved", storage_path=storage_path, size=len(file_data))
|
||||||
|
return storage_path
|
||||||
|
except OSError as e:
|
||||||
|
raise StorageError(f"Failed to save file: {e}") from e
|
||||||
|
|
||||||
|
def read_file(self, storage_path: str) -> bytes:
|
||||||
|
"""Read file data from local storage."""
|
||||||
|
file_path = self._resolve_path(storage_path)
|
||||||
|
if not file_path.exists():
|
||||||
|
raise StorageError(f"File not found: {storage_path}")
|
||||||
|
try:
|
||||||
|
return file_path.read_bytes()
|
||||||
|
except OSError as e:
|
||||||
|
raise StorageError(f"Failed to read file: {e}") from e
|
||||||
|
|
||||||
|
def delete_file(self, storage_path: str) -> bool:
|
||||||
|
"""Delete a file from local storage."""
|
||||||
|
file_path = self._resolve_path(storage_path)
|
||||||
|
if not file_path.exists():
|
||||||
|
return False
|
||||||
|
try:
|
||||||
|
file_path.unlink()
|
||||||
|
logger.info("file_deleted", storage_path=storage_path)
|
||||||
|
return True
|
||||||
|
except OSError as e:
|
||||||
|
logger.error("file_delete_failed", storage_path=storage_path, error=str(e))
|
||||||
|
raise StorageError(f"Failed to delete file: {e}") from e
|
||||||
|
|
||||||
|
def file_exists(self, storage_path: str) -> bool:
|
||||||
|
"""Check if a file exists in local storage."""
|
||||||
|
file_path = self._resolve_path(storage_path)
|
||||||
|
return file_path.exists()
|
||||||
|
|
||||||
|
def get_file_size(self, storage_path: str) -> int:
|
||||||
|
"""Get the file size in bytes."""
|
||||||
|
file_path = self._resolve_path(storage_path)
|
||||||
|
if not file_path.exists():
|
||||||
|
raise StorageError(f"File not found: {storage_path}")
|
||||||
|
return file_path.stat().st_size
|
||||||
|
|
||||||
|
def get_absolute_path(self, storage_path: str) -> str:
|
||||||
|
"""Get the absolute filesystem path."""
|
||||||
|
return str(self._resolve_path(storage_path))
|
||||||
|
|
||||||
|
def save_temp_file(self, file_data: bytes, filename: str) -> str:
|
||||||
|
"""Save a temporary file."""
|
||||||
|
return self.save_file(file_data, "temp", filename)
|
||||||
|
|
||||||
|
def cleanup_temp(self) -> int:
|
||||||
|
"""Remove all files in the temp directory."""
|
||||||
|
temp_dir = self.base_path / "temp"
|
||||||
|
count = 0
|
||||||
|
if temp_dir.exists():
|
||||||
|
for item in temp_dir.iterdir():
|
||||||
|
if item.is_file():
|
||||||
|
item.unlink()
|
||||||
|
count += 1
|
||||||
|
elif item.is_dir():
|
||||||
|
shutil.rmtree(item)
|
||||||
|
count += 1
|
||||||
|
logger.info("temp_cleanup", files_removed=count)
|
||||||
|
return count
|
||||||
|
|
||||||
|
|
||||||
|
def get_storage_provider() -> StorageProvider:
|
||||||
|
"""Factory function to get the configured storage provider."""
|
||||||
|
if settings.storage_provider == "local":
|
||||||
|
return LocalStorageProvider()
|
||||||
|
raise StorageError(f"Unknown storage provider: {settings.storage_provider}")
|
||||||
@@ -17,6 +17,8 @@ export interface DocumentLayout {
|
|||||||
height: number;
|
height: number;
|
||||||
confidence: number;
|
confidence: number;
|
||||||
sequence_no: number;
|
sequence_no: number;
|
||||||
|
page_width: number;
|
||||||
|
page_height: number;
|
||||||
}
|
}
|
||||||
|
|
||||||
export interface TemplateField {
|
export interface TemplateField {
|
||||||
@@ -66,7 +68,9 @@ export class TemplateService {
|
|||||||
width: tb.width,
|
width: tb.width,
|
||||||
height: tb.height,
|
height: tb.height,
|
||||||
confidence: tb.confidence || 0,
|
confidence: tb.confidence || 0,
|
||||||
sequence_no: tb.sequence || 0
|
sequence_no: tb.sequence || 0,
|
||||||
|
page_width: page.width || 1000,
|
||||||
|
page_height: page.height || 1000
|
||||||
});
|
});
|
||||||
});
|
});
|
||||||
}
|
}
|
||||||
|
|||||||
@@ -25,12 +25,21 @@
|
|||||||
<p>Upload a document to view its extracted layout structure.</p>
|
<p>Upload a document to view its extracted layout structure.</p>
|
||||||
</div>
|
</div>
|
||||||
|
|
||||||
<div *ngFor="let node of layoutNodes" cdkDrag class="layout-node" [ngClass]="node.block_type.toLowerCase()">
|
<div class="document-canvas" [style.aspect-ratio]="pageWidth + ' / ' + pageHeight">
|
||||||
<div class="node-type">{{ node.block_type }}</div>
|
<div *ngFor="let node of layoutNodes"
|
||||||
|
cdkDrag
|
||||||
|
class="layout-node"
|
||||||
|
[ngClass]="node.block_type.toLowerCase()"
|
||||||
|
[style.left.%]="(node.x_coordinate / node.page_width) * 100"
|
||||||
|
[style.top.%]="(node.y_coordinate / node.page_height) * 100"
|
||||||
|
[style.width.%]="(node.width / node.page_width) * 100"
|
||||||
|
[style.height.%]="(node.height / node.page_height) * 100">
|
||||||
|
<div class="node-type" *ngIf="node.height > 20">{{ node.block_type }}</div>
|
||||||
<div class="node-text">{{ node.text_value }}</div>
|
<div class="node-text">{{ node.text_value }}</div>
|
||||||
</div>
|
</div>
|
||||||
</div>
|
</div>
|
||||||
</div>
|
</div>
|
||||||
|
</div>
|
||||||
</ng-template>
|
</ng-template>
|
||||||
|
|
||||||
<!-- RIGHT PANEL: TEMPLATE MAPPING -->
|
<!-- RIGHT PANEL: TEMPLATE MAPPING -->
|
||||||
|
|||||||
@@ -47,19 +47,33 @@
|
|||||||
text-align: center;
|
text-align: center;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
.document-canvas {
|
||||||
|
position: relative;
|
||||||
|
width: 100%;
|
||||||
|
background: white;
|
||||||
|
box-shadow: 0 4px 6px -1px rgba(0, 0, 0, 0.1), 0 2px 4px -1px rgba(0, 0, 0, 0.06);
|
||||||
|
border-radius: 4px;
|
||||||
|
overflow: hidden;
|
||||||
|
}
|
||||||
|
|
||||||
/* Draggable Nodes */
|
/* Draggable Nodes */
|
||||||
.layout-node {
|
.layout-node {
|
||||||
padding: 0.75rem;
|
position: absolute;
|
||||||
margin-bottom: 0.5rem;
|
padding: 0.1rem;
|
||||||
background: white;
|
background: rgba(255, 255, 255, 0.85);
|
||||||
border: 1px solid var(--surface-border);
|
border: 1px solid var(--surface-border);
|
||||||
border-left: 4px solid var(--primary-color);
|
border-radius: 2px;
|
||||||
border-radius: 4px;
|
|
||||||
cursor: grab;
|
cursor: grab;
|
||||||
transition: box-shadow 0.2s;
|
transition: box-shadow 0.2s, border-color 0.2s;
|
||||||
|
overflow: hidden;
|
||||||
|
display: flex;
|
||||||
|
flex-direction: column;
|
||||||
|
justify-content: center;
|
||||||
|
|
||||||
&:hover {
|
&:hover {
|
||||||
box-shadow: 0 2px 4px rgba(0,0,0,0.1);
|
box-shadow: 0 4px 8px rgba(0,0,0,0.2);
|
||||||
|
z-index: 10;
|
||||||
|
border-color: var(--primary-color);
|
||||||
}
|
}
|
||||||
|
|
||||||
&:active {
|
&:active {
|
||||||
@@ -67,25 +81,32 @@
|
|||||||
}
|
}
|
||||||
|
|
||||||
.node-type {
|
.node-type {
|
||||||
font-size: 0.7rem;
|
font-size: 0.5rem;
|
||||||
color: var(--text-color-secondary);
|
color: var(--text-color-secondary);
|
||||||
text-transform: uppercase;
|
text-transform: uppercase;
|
||||||
font-weight: 600;
|
font-weight: 600;
|
||||||
margin-bottom: 0.25rem;
|
margin-bottom: 0.1rem;
|
||||||
|
line-height: 1;
|
||||||
|
white-space: nowrap;
|
||||||
|
overflow: hidden;
|
||||||
|
text-overflow: ellipsis;
|
||||||
}
|
}
|
||||||
|
|
||||||
.node-text {
|
.node-text {
|
||||||
font-size: 0.9rem;
|
font-size: 0.7rem;
|
||||||
color: var(--text-color);
|
color: var(--text-color);
|
||||||
word-break: break-word;
|
line-height: 1.1;
|
||||||
|
white-space: nowrap;
|
||||||
|
overflow: hidden;
|
||||||
|
text-overflow: ellipsis;
|
||||||
}
|
}
|
||||||
|
|
||||||
/* Color coding by block type */
|
/* Color coding by block type - use borders now instead of thick left border */
|
||||||
&.header { border-left-color: #3B82F6; }
|
&.header { border-color: #3B82F6; }
|
||||||
&.vendor { border-left-color: #8B5CF6; }
|
&.vendor { border-color: #8B5CF6; }
|
||||||
&.table_header { border-left-color: #F59E0B; }
|
&.table_header { border-color: #F59E0B; }
|
||||||
&.total { border-left-color: #10B981; }
|
&.total { border-color: #10B981; }
|
||||||
&.tax { border-left-color: #EF4444; }
|
&.tax { border-color: #EF4444; }
|
||||||
}
|
}
|
||||||
|
|
||||||
/* Template Mapping Side */
|
/* Template Mapping Side */
|
||||||
|
|||||||
@@ -36,6 +36,8 @@ export class TemplatesComponent implements OnInit {
|
|||||||
// Document State
|
// Document State
|
||||||
documentId: string | null = null;
|
documentId: string | null = null;
|
||||||
layoutNodes: DocumentLayout[] = [];
|
layoutNodes: DocumentLayout[] = [];
|
||||||
|
pageWidth: number = 800;
|
||||||
|
pageHeight: number = 1100;
|
||||||
|
|
||||||
// Template State
|
// Template State
|
||||||
templateName: string = '';
|
templateName: string = '';
|
||||||
@@ -88,6 +90,10 @@ export class TemplatesComponent implements OnInit {
|
|||||||
return;
|
return;
|
||||||
}
|
}
|
||||||
this.layoutNodes = layouts;
|
this.layoutNodes = layouts;
|
||||||
|
if (layouts.length > 0) {
|
||||||
|
this.pageWidth = layouts[0].page_width;
|
||||||
|
this.pageHeight = layouts[0].page_height;
|
||||||
|
}
|
||||||
// Trigger auto-recognition
|
// Trigger auto-recognition
|
||||||
this.autoRecognize();
|
this.autoRecognize();
|
||||||
}
|
}
|
||||||
|
|||||||
Reference in New Issue
Block a user