feat: cli & export to pdf

Author: oliverlambson

.gitignore

@@ -10,3 +10,4 @@ __pycache__
 node_modules/
 
 boredcharts/static/plotlyjs.min.js
+*.pdf

bored-charts/boredcharts/cli.py

@@ -0,0 +1,66 @@
+import argparse
+import asyncio
+from pathlib import Path
+
+from boredcharts.pdf import UrlToPdfFile, print_to_pdf_manual
+
+
+def init() -> None:
+    """create a new project scaffolding"""
+    raise NotImplementedError
+
+
+def list_() -> None:
+    """list available reports"""
+    raise NotImplementedError
+
+
+def export(report: str, *, writer: UrlToPdfFile = print_to_pdf_manual) -> None:
+    """write to pdf
+
+    TODO:
+    - [x] write to pdf
+    - [ ] spin up server
+    - [ ] provide list of reports
+    """
+    url = f"http://localhost:4000/{report}"  # TODO: spin up server & replace base url
+    file = Path(report).absolute().with_suffix(".pdf")
+    asyncio.run(writer(url, file))
+    print(f"Exported {report} to {file}")
+
+
+def dev() -> None:
+    """run uvicorn with reload"""
+    raise NotImplementedError
+
+
+def serve() -> None:
+    """run uvicorn without reload"""
+    raise NotImplementedError
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(description="boredcharts CLI")
+    subparsers = parser.add_subparsers(dest="command")
+    subparsers.required = True
+
+    parser_init = subparsers.add_parser("init", help="Create a new project scaffolding")
+    parser_init.set_defaults(func=init)
+
+    parser_init = subparsers.add_parser("list", help="List available reports")
+    parser_init.set_defaults(func=list_)
+
+    parser_export = subparsers.add_parser("export", help="Write to PDF")
+    parser_export.add_argument("report", type=str, help="The report to export")
+    parser_export.set_defaults(func=export)
+
+    parser_dev = subparsers.add_parser("dev", help="Run uvicorn with reload")
+    parser_dev.set_defaults(func=dev)
+
+    parser_serve = subparsers.add_parser("serve", help="Run uvicorn without reload")
+    parser_serve.set_defaults(func=serve)
+
+    args = parser.parse_args()
+
+    func_args = {k: v for k, v in vars(args).items() if k != "func" and k != "command"}
+    args.func(**func_args)

bored-charts/boredcharts/pdf.py

@@ -0,0 +1,92 @@
+import asyncio
+import json
+import subprocess
+import tempfile
+from pathlib import Path
+from typing import Protocol
+
+from playwright.async_api import async_playwright
+
+
+class UrlToPdfFile(Protocol):
+    async def __call__(self, url: str, file: Path) -> None: ...
+
+
+async def print_to_pdf_manual(url: str, file: Path) -> None:
+    """this one seems to work the best"""
+    async with async_playwright() as p:
+        args = [
+            "--headless=new",
+            "--virtual-time-budget=10000",  # seems to wait for ajax too?
+            "--run-all-compositor-stages-before-draw",  # also recommended, dunno
+            "--no-pdf-header-footer",
+            f"--print-to-pdf={file.as_posix()}",
+            url,
+        ]
+        process = await asyncio.create_subprocess_exec(
+            p.chromium.executable_path,
+            *args,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.PIPE,
+        )
+        _, stderr = await process.communicate()
+
+        if process.returncode != 0:
+            raise ChildProcessError(f"Could not export to pdf {stderr.decode()}")
+
+
+async def print_to_pdf_pw_adv(url: str, file: Path) -> None:
+    # in headless mode this doesn't seem to actually download the pdf
+    prefs = {
+        "printing": {
+            "print_preview_sticky_settings": {
+                "appState": json.dumps(
+                    {
+                        "version": 2,
+                        "recentDestinations": [
+                            {"id": "Save as PDF", "origin": "local", "account": ""}
+                        ],
+                        "selectedDestinationId": "Save as PDF",
+                        "isHeaderFooterEnabled": False,
+                    }
+                )
+            }
+        },
+    }
+    with tempfile.TemporaryDirectory() as pref_dir:
+        pref_file = Path(pref_dir) / "Default" / "Preferences"
+        pref_file.parent.mkdir(parents=True, exist_ok=True)
+        with pref_file.open("w") as f:
+            json.dump(prefs, f)
+
+        async with async_playwright() as p:
+            context = await p.chromium.launch_persistent_context(
+                user_data_dir=pref_file.parent.parent,
+                ignore_default_args=["--headless"],
+                args=[
+                    "--headless=new",
+                    "--kiosk-printing",
+                ],
+            )
+            page = await context.new_page()
+
+            await page.goto(url, wait_until="networkidle")
+            await page.evaluate("window.print();")
+            await context.close()
+
+
+async def print_to_pdf_pw_basic(url: str, file: Path) -> None:
+    # playwright's built-in pdf export results in text that can't be selected well
+    async with async_playwright() as p:
+        browser = await p.chromium.launch()
+        context = await browser.new_context()
+        page = await context.new_page()
+
+        await page.goto(url, wait_until="networkidle")
+        await page.pdf(
+            path=file,
+            format="A4",
+            margin={"top": "1cm", "right": "1cm", "bottom": "1cm", "left": "1cm"},
+        )
+        await context.close()
+        await browser.close()

bored-charts/pyproject.toml

@@ -12,6 +12,7 @@ dependencies = [
     "plotly>=5.23.0",
     "altair>=5.4.0",
     "seaborn>=0.13.2",
+    "playwright>=1.46.0",
 ]
 readme = "README.md"
 license = "MIT"
@@ -47,6 +48,9 @@ classifiers = [
 Repository = "https://github.com/oliverlambson/bored-charts.git"
 Issues = "https://github.com/oliverlambson/bored-charts/issues"
 
+[project.scripts]
+boredcharts = "boredcharts.cli:main"
+
 [build-system]
 requires = ["hatchling"]
 build-backend = "hatchling.build"