Add apptainer container. Clean up repo
This commit is contained in:
Kylian Schmidt
+4
-1
@@ -1 +1,4 @@
|
||||
__pycache__
|
||||
**__pycache__**
|
||||
.vscode
|
||||
*.sif
|
||||
*.ipynb
|
||||
Vendored
-6
@@ -1,6 +0,0 @@
|
||||
{
|
||||
"flake8.args": [
|
||||
"--max-line-length=120",
|
||||
"--ignore=W293,E123,W503",
|
||||
],
|
||||
}
|
||||
@@ -0,0 +1,20 @@
|
||||
Bootstrap: docker
|
||||
From: python:3.11-slim
|
||||
|
||||
%post
|
||||
apt-get update
|
||||
apt-get install -y --no-install-recommends imagemagick
|
||||
rm -rf /var/lib/apt/lists/*
|
||||
|
||||
pip install --no-cache-dir jinja2 pyyaml
|
||||
mkdir -p /src
|
||||
|
||||
%files
|
||||
. /src
|
||||
|
||||
%environment
|
||||
export PYTHONPATH=/src
|
||||
|
||||
%runscript
|
||||
cd /src
|
||||
exec python3 generate_gallery.py "$@"
|
||||
Binary file not shown.
@@ -16,8 +16,6 @@ class PathConfig:
|
||||
"""Configuration for system paths and directories."""
|
||||
work_dir: str
|
||||
web_folder: str
|
||||
cgi_script: str
|
||||
config_path: str
|
||||
|
||||
|
||||
@dataclass
|
||||
|
||||
@@ -1,205 +0,0 @@
|
||||
"""
|
||||
PDF Export Module for Scientific Gallery Generator
|
||||
|
||||
This module handles exporting and merging multiple plots into a single PDF
|
||||
using PyPDF2 and reportlab for layout.
|
||||
"""
|
||||
|
||||
import io
|
||||
import tempfile
|
||||
from pathlib import Path
|
||||
from typing import List, Dict, Any
|
||||
import subprocess
|
||||
|
||||
try:
|
||||
from PyPDF2 import PdfReader, PdfWriter
|
||||
from reportlab.pdfgen import canvas
|
||||
from reportlab.lib.pagesizes import letter, A4
|
||||
from reportlab.lib.utils import ImageReader
|
||||
HAS_PDF_LIBS = True
|
||||
except ImportError:
|
||||
HAS_PDF_LIBS = False
|
||||
|
||||
|
||||
class PDFExporter:
|
||||
"""Handles exporting multiple plots to a merged PDF"""
|
||||
|
||||
def __init__(self):
|
||||
self.page_size = A4
|
||||
self.margin = 50
|
||||
|
||||
def merge_plots(self, plot_paths: List[str], layout: Dict[str, int],
|
||||
output_path: str = None) -> bytes:
|
||||
"""
|
||||
Merge multiple PDF plots into a single PDF with grid layout.
|
||||
|
||||
Args:
|
||||
plot_paths: List of paths to PDF files
|
||||
layout: Dictionary with 'rows' and 'cols' keys
|
||||
output_path: Optional output file path
|
||||
|
||||
Returns:
|
||||
PDF bytes
|
||||
"""
|
||||
if not HAS_PDF_LIBS:
|
||||
return self._merge_with_pdfjam(plot_paths, layout, output_path)
|
||||
|
||||
return self._merge_with_pypdf(plot_paths, layout, output_path)
|
||||
|
||||
def _merge_with_pypdf(self, plot_paths: List[str], layout: Dict[str, int],
|
||||
output_path: str = None) -> bytes:
|
||||
"""Merge PDFs using PyPDF2 and reportlab"""
|
||||
from reportlab.pdfgen import canvas
|
||||
from reportlab.lib.pagesizes import A4
|
||||
|
||||
# Create a new PDF with the layout
|
||||
buffer = io.BytesIO()
|
||||
c = canvas.Canvas(buffer, pagesize=A4)
|
||||
page_width, page_height = A4
|
||||
|
||||
rows = layout['rows']
|
||||
cols = layout['cols']
|
||||
|
||||
# Calculate dimensions for each plot
|
||||
plot_width = (page_width - 2 * self.margin) / cols
|
||||
plot_height = (page_height - 2 * self.margin) / rows
|
||||
|
||||
# Place each plot in the grid
|
||||
for i, plot_path in enumerate(plot_paths):
|
||||
if i >= rows * cols:
|
||||
break
|
||||
|
||||
row = i // cols
|
||||
col = i % cols
|
||||
|
||||
# Calculate position
|
||||
x = self.margin + col * plot_width
|
||||
y = page_height - self.margin - (row + 1) * plot_height
|
||||
|
||||
try:
|
||||
# Read the source PDF
|
||||
with open(plot_path, 'rb') as f:
|
||||
reader = PdfReader(f)
|
||||
if len(reader.pages) > 0:
|
||||
page = reader.pages[0]
|
||||
|
||||
# Convert PDF page to image and place it
|
||||
# This is a simplified approach - in practice you'd want
|
||||
# to properly scale and position the PDF content
|
||||
self._draw_pdf_placeholder(c, x, y, plot_width, plot_height,
|
||||
Path(plot_path).stem)
|
||||
|
||||
except Exception as e:
|
||||
print(f"Error processing {plot_path}: {e}")
|
||||
self._draw_error_placeholder(c, x, y, plot_width, plot_height)
|
||||
|
||||
c.save()
|
||||
|
||||
if output_path:
|
||||
with open(output_path, 'wb') as f:
|
||||
f.write(buffer.getvalue())
|
||||
|
||||
return buffer.getvalue()
|
||||
|
||||
def _merge_with_pdfjam(self, plot_paths: List[str], layout: Dict[str, int],
|
||||
output_path: str = None) -> bytes:
|
||||
"""Merge PDFs using pdfjam (requires pdfpages LaTeX package)"""
|
||||
rows = layout['rows']
|
||||
cols = layout['cols']
|
||||
|
||||
# Create temporary output file
|
||||
with tempfile.NamedTemporaryFile(suffix='.pdf', delete=False) as tmp_file:
|
||||
temp_output = tmp_file.name
|
||||
|
||||
try:
|
||||
# Build pdfjam command
|
||||
cmd = [
|
||||
'pdfjam',
|
||||
'--nup', f'{cols}x{rows}',
|
||||
'--landscape' if cols > rows else '--no-landscape',
|
||||
'--frame', 'true',
|
||||
'--delta', '10pt 10pt',
|
||||
'--offset', '0pt 0pt',
|
||||
'--outfile', temp_output
|
||||
]
|
||||
|
||||
# Add input files
|
||||
cmd.extend(plot_paths[:rows * cols])
|
||||
|
||||
# Run pdfjam
|
||||
result = subprocess.run(cmd, capture_output=True, text=True)
|
||||
|
||||
if result.returncode != 0:
|
||||
raise Exception(f"pdfjam failed: {result.stderr}")
|
||||
|
||||
# Read the output file
|
||||
with open(temp_output, 'rb') as f:
|
||||
pdf_bytes = f.read()
|
||||
|
||||
if output_path:
|
||||
with open(output_path, 'wb') as f:
|
||||
f.write(pdf_bytes)
|
||||
|
||||
return pdf_bytes
|
||||
|
||||
finally:
|
||||
# Clean up temporary file
|
||||
Path(temp_output).unlink(missing_ok=True)
|
||||
|
||||
def _draw_pdf_placeholder(self, canvas, x: float, y: float, width: float,
|
||||
height: float, plot_name: str):
|
||||
"""Draw a placeholder for a PDF plot"""
|
||||
# Draw border
|
||||
canvas.setStrokeColorRGB(0.5, 0.5, 0.5)
|
||||
canvas.setLineWidth(1)
|
||||
canvas.rect(x, y, width, height)
|
||||
|
||||
# Draw plot name
|
||||
canvas.setFillColorRGB(0, 0, 0)
|
||||
canvas.setFont("Helvetica", 10)
|
||||
text_width = canvas.stringWidth(plot_name, "Helvetica", 10)
|
||||
text_x = x + (width - text_width) / 2
|
||||
text_y = y + height / 2
|
||||
canvas.drawString(text_x, text_y, plot_name)
|
||||
|
||||
def _draw_error_placeholder(self, canvas, x: float, y: float, width: float,
|
||||
height: float):
|
||||
"""Draw an error placeholder"""
|
||||
# Draw red border
|
||||
canvas.setStrokeColorRGB(1, 0, 0)
|
||||
canvas.setLineWidth(2)
|
||||
canvas.rect(x, y, width, height)
|
||||
|
||||
# Draw error text
|
||||
canvas.setFillColorRGB(1, 0, 0)
|
||||
canvas.setFont("Helvetica-Bold", 12)
|
||||
error_text = "Error loading plot"
|
||||
text_width = canvas.stringWidth(error_text, "Helvetica-Bold", 12)
|
||||
text_x = x + (width - text_width) / 2
|
||||
text_y = y + height / 2
|
||||
canvas.drawString(text_x, text_y, error_text)
|
||||
|
||||
|
||||
def check_dependencies() -> Dict[str, bool]:
|
||||
"""Check if required dependencies are available"""
|
||||
deps = {
|
||||
'pypdf2': HAS_PDF_LIBS,
|
||||
'pdfjam': False
|
||||
}
|
||||
|
||||
# Check for pdfjam
|
||||
try:
|
||||
result = subprocess.run(['pdfjam', '--version'],
|
||||
capture_output=True, text=True)
|
||||
deps['pdfjam'] = result.returncode == 0
|
||||
except FileNotFoundError:
|
||||
pass
|
||||
|
||||
return deps
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
# Test the exporter
|
||||
exporter = PDFExporter()
|
||||
deps = check_dependencies()
|
||||
print("Available dependencies:", deps)
|
||||
@@ -7,6 +7,10 @@ authors = [
|
||||
]
|
||||
readme = "README.md"
|
||||
requires-python = ">=3.9"
|
||||
dependencies = [
|
||||
"jinja2",
|
||||
"pyyaml",
|
||||
]
|
||||
|
||||
[tool.black]
|
||||
line-length = 120
|
||||
|
||||
-235
@@ -1,235 +0,0 @@
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 2,
|
||||
"id": "aedbb9f5",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from dataclasses import dataclass, field, asdict\n",
|
||||
"from pathlib import Path\n",
|
||||
"\n",
|
||||
"import yaml\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"@dataclass\n",
|
||||
"class GalleryItem:\n",
|
||||
" name: str\n",
|
||||
" path: Path\n"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 5,
|
||||
"id": "66624faf",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"yaml_file = Path(\"config.yaml\")\n",
|
||||
"\n",
|
||||
"with open(yaml_file, \"r\") as f:\n",
|
||||
" new_config: dict = yaml.safe_load(f)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 8,
|
||||
"id": "40a472f8",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"{'name': 'test_1_plot',\n",
|
||||
" 'path': 'data/cf_store/test_analysis/cf.PlotVariables1D/run2_2016_nano_v9/calib__test/sel__test/prod__test/weight__test/nominal/datasets_ttbar_dl/dev1/plot__proc_3_5606769559__cat_incl__var_jet1_pt.pdf'}"
|
||||
]
|
||||
},
|
||||
"execution_count": 8,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"new_config[\"sources\"][0]"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 10,
|
||||
"id": "745c97ad",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"[GalleryItem(name='test_1_plot', path=PosixPath('data/cf_store/test_analysis/cf.PlotVariables1D/run2_2016_nano_v9/calib__test/sel__test/prod__test/weight__test/nominal/datasets_ttbar_dl/dev1/plot__proc_3_5606769559__cat_incl__var_jet1_pt.pdf')),\n",
|
||||
" GalleryItem(name='test_2_dir', path=PosixPath('data/cf_store/test_analysis/cf.PlotVariables1D/run2_2016_nano_v9/calib__test/sel__test/prod__test/weight__test/nominal/datasets_ttbar_dl/dev1'))]"
|
||||
]
|
||||
},
|
||||
"execution_count": 10,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"sources = [\n",
|
||||
" GalleryItem(name=src[\"name\"], path=Path(src[\"path\"]))\n",
|
||||
" for src in new_config.get(\"sources\", False)\n",
|
||||
"]\n",
|
||||
"sources"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 15,
|
||||
"id": "a0102f9d",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"Config(web_folder='', backup_folder='', png_dpi=400, plot_root='gallery', sources=[{'name': 'test_1_plot', 'path': 'data/cf_store/test_analysis/cf.PlotVariables1D/run2_2016_nano_v9/calib__test/sel__test/prod__test/weight__test/nominal/datasets_ttbar_dl/dev1/plot__proc_3_5606769559__cat_incl__var_jet1_pt.pdf'}, {'name': 'test_2_dir', 'path': 'data/cf_store/test_analysis/cf.PlotVariables1D/run2_2016_nano_v9/calib__test/sel__test/prod__test/weight__test/nominal/datasets_ttbar_dl/dev1/'}])"
|
||||
]
|
||||
},
|
||||
"execution_count": 15,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"\n",
|
||||
"\n",
|
||||
"@dataclass\n",
|
||||
"class Config:\n",
|
||||
" web_folder: str = \"\"\n",
|
||||
" backup_folder: str = \"\"\n",
|
||||
" png_dpi: int = 400\n",
|
||||
" plot_root: str = \"gallery\"\n",
|
||||
" sources: list[GalleryItem] = field(default_factory=list)\n",
|
||||
"\n",
|
||||
" @classmethod\n",
|
||||
" def from_yaml(cls, yaml_file: str, strict: bool = False) -> \"Config\":\n",
|
||||
" \"\"\"\n",
|
||||
" Load configuration from a YAML file.\n",
|
||||
"\n",
|
||||
" Args:\n",
|
||||
" yaml_file (str): Path to the YAML file.\n",
|
||||
" strict (bool):\n",
|
||||
" If True, raises an error if a key in the YAML file does not exist in the Config class.\n",
|
||||
" If False (default), adds all keys as attributes\n",
|
||||
"\n",
|
||||
" Returns:\n",
|
||||
" Config: Instance of this class\n",
|
||||
" \"\"\"\n",
|
||||
"\n",
|
||||
" with open(yaml_file, \"r\") as f:\n",
|
||||
" new_config: dict = yaml.safe_load(f)\n",
|
||||
"\n",
|
||||
" new_config[\"sources\"] = [\n",
|
||||
" GalleryItem(name=src[\"name\"], path=Path(src[\"path\"]))\n",
|
||||
" for src in new_config.get(\"sources\", False)\n",
|
||||
" ]\n",
|
||||
"\n",
|
||||
" instance = cls(**{\n",
|
||||
" k: v\n",
|
||||
" for k, v in new_config.items()\n",
|
||||
" if hasattr(cls, k) or not strict\n",
|
||||
" })\n",
|
||||
"\n",
|
||||
" for key in new_config.keys():\n",
|
||||
" if strict and not hasattr(instance, key):\n",
|
||||
" raise KeyError(f\"Key '{key}' not found in Config class\")\n",
|
||||
"\n",
|
||||
" return instance\n",
|
||||
"\n",
|
||||
" def to_yaml(self, yaml_file: str) -> None:\n",
|
||||
" \"\"\"\n",
|
||||
" Save the current configuration to a YAML file.\n",
|
||||
"\n",
|
||||
" Args:\n",
|
||||
" yaml_file (str): Path to the YAML file.\n",
|
||||
" \"\"\"\n",
|
||||
" with open(yaml_file, \"w\") as f:\n",
|
||||
" yaml.dump(asdict(self), f, default_flow_style=False)\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"instance = Config(**{\n",
|
||||
" k: v\n",
|
||||
" for k, v in new_config.items()\n",
|
||||
" #if hasattr(Config, k)\n",
|
||||
"})\n",
|
||||
"instance"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 17,
|
||||
"id": "0b516ae9",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"[{'name': 'test_1_plot',\n",
|
||||
" 'path': 'data/cf_store/test_analysis/cf.PlotVariables1D/run2_2016_nano_v9/calib__test/sel__test/prod__test/weight__test/nominal/datasets_ttbar_dl/dev1/plot__proc_3_5606769559__cat_incl__var_jet1_pt.pdf'},\n",
|
||||
" {'name': 'test_2_dir',\n",
|
||||
" 'path': 'data/cf_store/test_analysis/cf.PlotVariables1D/run2_2016_nano_v9/calib__test/sel__test/prod__test/weight__test/nominal/datasets_ttbar_dl/dev1/'}]"
|
||||
]
|
||||
},
|
||||
"execution_count": 17,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"new_config[\"sources\"]"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 18,
|
||||
"id": "b047f0c5",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"[{'name': 'test_1_plot',\n",
|
||||
" 'path': 'data/cf_store/test_analysis/cf.PlotVariables1D/run2_2016_nano_v9/calib__test/sel__test/prod__test/weight__test/nominal/datasets_ttbar_dl/dev1/plot__proc_3_5606769559__cat_incl__var_jet1_pt.pdf'},\n",
|
||||
" {'name': 'test_2_dir',\n",
|
||||
" 'path': 'data/cf_store/test_analysis/cf.PlotVariables1D/run2_2016_nano_v9/calib__test/sel__test/prod__test/weight__test/nominal/datasets_ttbar_dl/dev1/'}]"
|
||||
]
|
||||
},
|
||||
"execution_count": 18,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"instance.sources"
|
||||
]
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
"kernelspec": {
|
||||
"display_name": "Python 3",
|
||||
"language": "python",
|
||||
"name": "python3"
|
||||
},
|
||||
"language_info": {
|
||||
"codemirror_mode": {
|
||||
"name": "ipython",
|
||||
"version": 3
|
||||
},
|
||||
"file_extension": ".py",
|
||||
"mimetype": "text/x-python",
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.12.1"
|
||||
}
|
||||
},
|
||||
"nbformat": 4,
|
||||
"nbformat_minor": 5
|
||||
}
|
||||
Reference in New Issue
Block a user