diff --git a/base_report_paper_muncher/README.rst b/base_report_paper_muncher/README.rst
new file mode 100644
index 00000000000..02532148f95
--- /dev/null
+++ b/base_report_paper_muncher/README.rst
@@ -0,0 +1,199 @@
+============================
+Report Engine: Paper Muncher
+============================
+
+..
+ !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
+ !! This file is generated by oca-gen-addon-readme !!
+ !! changes will be overwritten. !!
+ !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
+ !! source digest: sha256:b1d3faf687c4a3a6d43429969fc2e2e38e46f4fc707ef22f9d88f4cd9b94cf4a
+ !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
+
+.. |badge1| image:: https://img.shields.io/badge/maturity-Beta-yellow.png
+ :target: https://odoo-community.org/page/development-status
+ :alt: Beta
+.. |badge2| image:: https://img.shields.io/badge/licence-LGPL--3-blue.png
+ :target: http://www.gnu.org/licenses/lgpl-3.0-standalone.html
+ :alt: License: LGPL-3
+.. |badge3| image:: https://img.shields.io/badge/github-OCA%2Fserver--tools-lightgray.png?logo=github
+ :target: https://github.com/OCA/server-tools/tree/17.0/base_report_paper_muncher
+ :alt: OCA/server-tools
+.. |badge4| image:: https://img.shields.io/badge/weblate-Translate%20me-F47D42.png
+ :target: https://translation.odoo-community.org/projects/server-tools-17-0/server-tools-17-0-base_report_paper_muncher
+ :alt: Translate me on Weblate
+.. |badge5| image:: https://img.shields.io/badge/runboat-Try%20me-875A7B.png
+ :target: https://runboat.odoo-community.org/builds?repo=OCA/server-tools&target_branch=17.0
+ :alt: Try me on Runboat
+
+|badge1| |badge2| |badge3| |badge4| |badge5|
+
+This module integrates `Paper
+Muncher `__ as the PDF rendering
+engine for QWeb reports. When the ``paper-muncher`` binary is available
+on the system, it replaces wkhtmltopdf as the default engine for PDF
+generation.
+
+Main features:
+
+- Automatic detection of the ``paper-muncher`` binary in ``PATH`` or in
+ ``/opt/paper-muncher/bin/paper-muncher``
+- Globally configurable PDF engine via the ``report.pdf_engine`` system
+ parameter
+- Multi-page header and footer support and ``report.paperformat``
+ integration
+- Transparent fallback to wkhtmltopdf when Paper Muncher is not
+ installed
+- HTTP-over-pipe communication between Odoo and the Paper Muncher
+ subprocess, so report assets are served with Odoo permissions
+
+This module depends on:
+
+- Odoo modules: ``base``, ``web``
+- Python package: ``h11``
+- System binary: ``paper-muncher`` from `GitHub
+ releases `__
+
+**Table of contents**
+
+.. contents::
+ :local:
+
+Installation
+============
+
+Install the ``paper-muncher`` system binary before enabling this module.
+Example for Ubuntu 22.04 (Jammy):
+
+.. code:: bash
+
+ curl -fsSL -o paper-muncher.deb \
+ "https://github.com/odoo/paper-muncher/releases/download/v0.3.1-1/paper-muncher_v0.3.1-1_jammy_amd64.deb"
+ sudo apt install ./paper-muncher.deb
+ paper-muncher --help
+
+If the binary is not installed via the ``.deb`` package, add
+``/opt/paper-muncher/bin`` to the ``PATH`` of the Odoo process.
+
+The Python dependency ``h11`` is declared in the module manifest and
+listed in ``requirements.txt``. Install it in the Odoo Python
+environment if needed:
+
+.. code:: bash
+
+ pip install h11
+
+Then install the module as any other Odoo addon:
+
+.. code:: bash
+
+ odoo -d -i base_report_paper_muncher --stop-after-init
+
+Paper Muncher is only supported on Linux and macOS.
+
+Configuration
+=============
+
+The system parameter ``report.pdf_engine`` controls which PDF engine is
+used:
+
++--------------------+-------------------------------------------------+
+| Value | Behavior |
++====================+=================================================+
+| ``auto`` (default) | Use Paper Muncher when the binary is available, |
+| | otherwise wkhtmltopdf |
++--------------------+-------------------------------------------------+
+| ``paper-muncher`` | Force Paper Muncher; raise an error if the |
+| | binary is missing |
++--------------------+-------------------------------------------------+
+| ``wkhtmltopdf`` | Always use wkhtmltopdf |
++--------------------+-------------------------------------------------+
+
+Change it from code:
+
+.. code:: python
+
+ env["ir.config_parameter"].sudo().set_param("report.pdf_engine", "auto")
+
+Or from **Settings > Technical > Parameters > System Parameters**, key
+``report.pdf_engine``.
+
+Optional environment variables for the Odoo process:
+
+- ``ODOO_PAPER_MUNCHER_FEATURE=1``: pass ``--feature *=on`` to the
+ binary
+- ``ODOO_PAPER_MUNCHER_DEBUG=1``: pass ``--debug http-client`` to the
+ binary. This flag is not sent by default in production. It is also
+ enabled automatically when the module logger is set to ``DEBUG``.
+
+Usage
+=====
+
+Once installed and configured, QWeb PDF reports use Paper Muncher
+automatically when the binary is available and ``report.pdf_engine`` is
+set to ``auto`` or ``paper-muncher``.
+
+Check the engine status:
+
+.. code:: python
+
+ env["ir.actions.report"].get_wkhtmltopdf_state()
+
+Returns ``ok`` when Paper Muncher is available with ``auto`` or
+``paper-muncher``. Returns ``install`` when ``paper-muncher`` is forced
+but the binary is missing.
+
+When generating PDFs, look for log lines prefixed with ``PDF engine:``:
+
+- ``PDF engine: Paper-Muncher (...)``: rendering started with Paper
+ Muncher
+- ``PDF engine: Paper-Muncher completed (...)``: PDF generated
+ successfully
+- ``PDF engine: wkhtmltopdf (...)``: fallback to wkhtmltopdf
+
+Low-level HTTP-over-pipe details are logged at ``DEBUG`` level only.
+
+Known issues / Roadmap
+======================
+
+- Paper Muncher is only supported on Linux and macOS (not Windows).
+- WebSocket requests are rejected by the HTTP-over-pipe server.
+- Rendering timeout: 15 minutes (``SERVE_TIMEOUT``).
+- Pipe write timeout: 15 seconds (``WRITE_TIMEOUT``).
+- The binary must be installed on the system; this module only
+ integrates the communication layer.
+
+Bug Tracker
+===========
+
+Bugs are tracked on `GitHub Issues `_.
+In case of trouble, please check there if your issue has already been reported.
+If you spotted it first, help us to smash it by providing a detailed and welcomed
+`feedback `_.
+
+Do not contact contributors directly about support or help with technical issues.
+
+Credits
+=======
+
+Contributors
+------------
+
+- Felix Coca
+
+Maintainers
+-----------
+
+This module is maintained by the OCA.
+
+.. image:: https://odoo-community.org/logo.png
+ :alt: Odoo Community Association
+ :target: https://odoo-community.org
+
+OCA, or the Odoo Community Association, is a nonprofit organization whose
+mission is to support the collaborative development of Odoo features and
+promote its widespread use.
+
+This module is part of the `OCA/server-tools `_ project on GitHub.
+
+You are welcome to contribute. To learn how please visit https://odoo-community.org/page/Contribute.
diff --git a/base_report_paper_muncher/__init__.py b/base_report_paper_muncher/__init__.py
new file mode 100644
index 00000000000..d6210b1285d
--- /dev/null
+++ b/base_report_paper_muncher/__init__.py
@@ -0,0 +1,3 @@
+# Part of Odoo. See LICENSE file for full copyright and licensing details.
+
+from . import models
diff --git a/base_report_paper_muncher/__manifest__.py b/base_report_paper_muncher/__manifest__.py
new file mode 100644
index 00000000000..6b046b8c635
--- /dev/null
+++ b/base_report_paper_muncher/__manifest__.py
@@ -0,0 +1,27 @@
+# Part of Odoo. See LICENSE file for full copyright and licensing details.
+
+{
+ "name": "Report Engine: Paper Muncher",
+ "summary": (
+ "Paper Muncher PDF rendering engine for QWeb reports, "
+ "replacing wkhtmltopdf when the binary is available"
+ ),
+ "version": "17.0.1.0.0",
+ "development_status": "Beta",
+ "category": "Hidden/Tools",
+ "website": "https://github.com/OCA/server-tools",
+ "author": "Odoo Community Association (OCA)",
+ "depends": [
+ "base",
+ "web",
+ ],
+ "data": [
+ "data/ir_config_parameter_data.xml",
+ ],
+ "external_dependencies": {
+ "python": ["h11"],
+ },
+ "license": "LGPL-3",
+ "installable": True,
+ "application": False,
+}
diff --git a/base_report_paper_muncher/data/ir_config_parameter_data.xml b/base_report_paper_muncher/data/ir_config_parameter_data.xml
new file mode 100644
index 00000000000..a0bd62e7774
--- /dev/null
+++ b/base_report_paper_muncher/data/ir_config_parameter_data.xml
@@ -0,0 +1,7 @@
+
+
+
+ report.pdf_engine
+ auto
+
+
diff --git a/base_report_paper_muncher/models/__init__.py b/base_report_paper_muncher/models/__init__.py
new file mode 100644
index 00000000000..7fc40541f9a
--- /dev/null
+++ b/base_report_paper_muncher/models/__init__.py
@@ -0,0 +1,3 @@
+# Part of Odoo. See LICENSE file for full copyright and licensing details.
+
+from . import ir_actions_report
diff --git a/base_report_paper_muncher/models/ir_actions_report.py b/base_report_paper_muncher/models/ir_actions_report.py
new file mode 100644
index 00000000000..1bd3dfc5b3b
--- /dev/null
+++ b/base_report_paper_muncher/models/ir_actions_report.py
@@ -0,0 +1,363 @@
+# Part of Odoo. See LICENSE file for full copyright and licensing details.
+
+import logging
+import os
+import re
+import subprocess
+from collections.abc import Sequence
+from contextlib import ExitStack
+from typing import NamedTuple
+from urllib.parse import urlsplit
+
+import lxml.html
+
+from odoo import _, api, models
+from odoo.exceptions import UserError
+from odoo.http import request, root
+from odoo.service import security
+
+from ..paper_muncher import PaperMuncherInfo, PaperMuncherServer, paper_muncher
+
+_logger = logging.getLogger(__name__)
+
+_BODY_TAG_RE = re.compile(r"]*>", re.IGNORECASE)
+
+
+class PdfEngineResolution(NamedTuple):
+ engine: str
+ use_paper_muncher: bool
+ pm_info: PaperMuncherInfo
+
+
+def _extract_div_fragments(body_content: str) -> list[str]:
+ """Extract top-level div fragments from HTML body content."""
+ if not body_content:
+ return []
+ wrapper = lxml.html.fromstring(f"{body_content}
")
+ return [
+ lxml.html.tostring(div, encoding="unicode") for div in wrapper.findall("./div")
+ ]
+
+
+def partition_on_body(html: str) -> tuple[str, str, str]:
+ """Split HTML into pre-body, body content, and post-body."""
+ match = _BODY_TAG_RE.search(html)
+ if not match:
+ return html, "", ""
+ pre_body = html[: match.end()]
+ rest = html[match.end() :]
+ body, sep, post_body = rest.rpartition(" ")
+ if not sep:
+ return html, "", ""
+ return pre_body, body, sep + post_body
+
+
+def make_multi_docs_html(
+ bodies: Sequence[str], header: str = "", footer: str = ""
+) -> list[str]:
+ """Inject per-page header/footer fragments into each body HTML document."""
+ footer_body = partition_on_body(footer)[1]
+ footers = _extract_div_fragments(footer_body)
+
+ header_body = partition_on_body(header)[1]
+ headers = _extract_div_fragments(header_body)
+
+ is_same_length_header = len(headers) == len(bodies)
+ if headers and not is_same_length_header:
+ _logger.warning(
+ "Header fragments count (%d) does not match body count (%d); "
+ "reusing the first header fragment where needed.",
+ len(headers),
+ len(bodies),
+ )
+
+ is_same_length_footer = len(footers) == len(bodies)
+ if footers and not is_same_length_footer:
+ _logger.warning(
+ "Footer fragments count (%d) does not match body count (%d); "
+ "reusing the first footer fragment where needed.",
+ len(footers),
+ len(bodies),
+ )
+
+ documents = []
+ for i, body in enumerate(bodies):
+ pre_body, body_content, post_body = partition_on_body(body)
+ header_fragment = (
+ headers[i] if is_same_length_header else (headers[0] if headers else "")
+ )
+ footer_fragment = (
+ footers[i] if is_same_length_footer else (footers[0] if footers else "")
+ )
+ documents.append(
+ f"{pre_body}{header_fragment}{body_content}{footer_fragment}{post_body}\n"
+ )
+
+ return documents
+
+
+def _paper_muncher_debug_enabled():
+ return os.getenv("ODOO_PAPER_MUNCHER_DEBUG") == "1" or _logger.isEnabledFor(
+ logging.DEBUG
+ )
+
+
+class IrActionsReport(models.Model):
+ _inherit = "ir.actions.report"
+
+ @api.model
+ def _get_pdf_engine_config(self):
+ return (
+ self.env["ir.config_parameter"]
+ .sudo()
+ .get_param("report.pdf_engine", "auto")
+ )
+
+ @api.model
+ def _resolve_pdf_engine(self) -> PdfEngineResolution:
+ engine = self._get_pdf_engine_config()
+ pm_info = paper_muncher()
+ if engine == "wkhtmltopdf":
+ return PdfEngineResolution(engine, False, pm_info)
+ if engine == "paper-muncher":
+ if pm_info.state != "ok":
+ raise UserError(
+ _(
+ "Paper-Muncher is not installed on this system. "
+ "Install it from "
+ "https://github.com/odoo/paper-muncher/releases "
+ "or set report.pdf_engine to 'auto' or 'wkhtmltopdf'."
+ )
+ )
+ return PdfEngineResolution(engine, True, pm_info)
+ return PdfEngineResolution(engine, pm_info.state == "ok", pm_info)
+
+ @api.model
+ def _should_use_paper_muncher(self):
+ return self._resolve_pdf_engine().use_paper_muncher
+
+ @api.model
+ def get_wkhtmltopdf_state(self):
+ engine = self._get_pdf_engine_config()
+ pm_info = paper_muncher()
+ if engine == "wkhtmltopdf":
+ return super().get_wkhtmltopdf_state()
+ if pm_info.state == "ok":
+ return "ok"
+ if engine == "paper-muncher":
+ return "install"
+ return super().get_wkhtmltopdf_state()
+
+ @api.model
+ def _resolve_report_sudo(self, report_ref):
+ if not report_ref:
+ return None
+ try:
+ return self._get_report(report_ref)
+ except ValueError:
+ return None
+
+ @api.model
+ def _get_report_log_label(self, report_ref=False, report_sudo=None):
+ if report_sudo:
+ return report_sudo.report_name or report_sudo.display_name
+ if not report_ref:
+ return "unknown"
+ try:
+ report = self._get_report(report_ref)
+ return report.report_name or report.display_name
+ except ValueError:
+ return str(report_ref)
+
+ @api.model
+ def _run_wkhtmltopdf(
+ self,
+ bodies,
+ report_ref=False,
+ header=None,
+ footer=None,
+ landscape=False,
+ specific_paperformat_args=None,
+ set_viewport_size=False,
+ ):
+ resolution = self._resolve_pdf_engine()
+ report_sudo = self._resolve_report_sudo(report_ref)
+ report_label = self._get_report_log_label(report_ref, report_sudo=report_sudo)
+
+ if resolution.use_paper_muncher:
+ _logger.info(
+ (
+ "PDF engine: Paper-Muncher "
+ "(report=%s, config=%s, binary=%s, version=%s)"
+ ),
+ report_label,
+ resolution.engine,
+ resolution.pm_info.bin,
+ resolution.pm_info.version,
+ )
+ return self._run_paper_muncher(
+ bodies,
+ report_ref=report_ref,
+ header=header,
+ footer=footer,
+ landscape=landscape,
+ specific_paperformat_args=specific_paperformat_args,
+ report_sudo=report_sudo,
+ report_label=report_label,
+ pm_info=resolution.pm_info,
+ )
+ if resolution.engine == "auto" and resolution.pm_info.state != "ok":
+ _logger.info(
+ (
+ "PDF engine: wkhtmltopdf "
+ "(report=%s, config=auto, Paper-Muncher not available)"
+ ),
+ report_label,
+ )
+ elif resolution.engine == "wkhtmltopdf":
+ _logger.info(
+ "PDF engine: wkhtmltopdf (report=%s, config=wkhtmltopdf)",
+ report_label,
+ )
+ return super()._run_wkhtmltopdf(
+ bodies,
+ report_ref=report_ref,
+ header=header,
+ footer=footer,
+ landscape=landscape,
+ specific_paperformat_args=specific_paperformat_args,
+ set_viewport_size=set_viewport_size,
+ )
+
+ @api.model
+ def _run_paper_muncher(
+ self,
+ bodies,
+ report_ref=False,
+ header=None,
+ footer=None,
+ landscape=False,
+ specific_paperformat_args=None,
+ scale=72,
+ report_sudo=None,
+ report_label=None,
+ pm_info=None,
+ ):
+ """Render a PDF from HTML content using Paper Muncher subprocess."""
+ pm_info = pm_info or paper_muncher()
+ report_label = report_label or self._get_report_log_label(
+ report_ref, report_sudo=report_sudo
+ )
+
+ paperformat = (
+ report_sudo.get_paperformat() if report_sudo else self.get_paperformat()
+ )
+ header = header or ""
+ footer = footer or ""
+
+ if not isinstance(bodies, list | tuple):
+ bodies = list(bodies)
+
+ documents = make_multi_docs_html(bodies, header, footer)
+
+ names = [f"pipe:/paper-muncher/{i}.html" for i in range(len(documents))]
+ extra_args = self._build_paper_muncher_extra_args(
+ landscape=landscape,
+ paperformat=paperformat,
+ scale=scale,
+ specific_paperformat_args=specific_paperformat_args,
+ )
+
+ os_env = os.environ.copy()
+ os_env["NO_COLOR"] = "1"
+
+ try:
+ with ExitStack() as stack:
+ wsgi_environ = {}
+ if request and request.db:
+ temp_session = root.session_store.new()
+ temp_session.update(
+ {
+ **request.session,
+ "debug": "",
+ "_trace_disable": True,
+ }
+ )
+ if temp_session.uid:
+ temp_session.session_token = security.compute_session_token(
+ temp_session, self.env
+ )
+ root.session_store.save(temp_session)
+ stack.callback(root.session_store.delete, temp_session)
+ url = urlsplit(self._get_report_url())
+ wsgi_environ["HTTP_HOST"] = url.netloc
+ wsgi_environ["HTTP_COOKIE"] = (
+ f"session_id={temp_session.sid}; HttpOnly; "
+ f"domain={url.hostname}; path=/;"
+ )
+ else:
+ wsgi_environ["HTTP_X_ODOO_DATABASE"] = self.env.cr.dbname
+
+ with PaperMuncherServer(
+ args=[
+ pm_info.bin,
+ *names,
+ "-o",
+ "pipe:/paper-muncher/output.pdf",
+ *extra_args,
+ ],
+ os_env=os_env,
+ wsgi_environ=wsgi_environ,
+ ) as server:
+ pdf = server.serve(documents)
+ _logger.info(
+ (
+ "PDF engine: Paper-Muncher completed "
+ "(report=%s, pages=%d, size=%d bytes)"
+ ),
+ report_label,
+ len(documents),
+ len(pdf),
+ )
+ return pdf
+ except (subprocess.CalledProcessError, TimeoutError, RuntimeError) as exc:
+ message = _("Paper-Muncher failed. Message: %s", str(exc)[-1000:])
+ _logger.warning(message)
+ raise UserError(message) from exc
+
+ @api.model
+ def _build_paper_muncher_extra_args(
+ self,
+ landscape=False,
+ paperformat=None,
+ scale=72,
+ specific_paperformat_args=None,
+ ):
+ """Build paper-muncher CLI arguments (exposed for testing)."""
+ if specific_paperformat_args:
+ if not landscape and specific_paperformat_args.get("data-report-landscape"):
+ landscape = specific_paperformat_args["data-report-landscape"]
+ if specific_paperformat_args.get("data-report-dpi"):
+ scale = int(specific_paperformat_args["data-report-dpi"])
+
+ extra_args = [
+ "--scale",
+ f"{scale}dpi",
+ "--margins",
+ "none",
+ ]
+ if landscape:
+ extra_args += ["--orientation", "landscape"]
+ elif paperformat and paperformat.orientation:
+ extra_args += ["--orientation", paperformat.orientation.lower()]
+ if os.getenv("ODOO_PAPER_MUNCHER_FEATURE") == "1":
+ extra_args += ["--feature", "*=on"]
+ if paperformat and paperformat.format:
+ if paperformat.format != "custom":
+ extra_args += ["--paper", paperformat.format]
+ elif paperformat.page_height and paperformat.page_width:
+ extra_args += ["--width", f"{paperformat.page_width}mm"]
+ extra_args += ["--height", f"{paperformat.page_height}mm"]
+ if _paper_muncher_debug_enabled():
+ extra_args += ["--debug", "http-client"]
+ return extra_args
diff --git a/base_report_paper_muncher/paper_muncher.py b/base_report_paper_muncher/paper_muncher.py
new file mode 100644
index 00000000000..ac57129b2c6
--- /dev/null
+++ b/base_report_paper_muncher/paper_muncher.py
@@ -0,0 +1,385 @@
+# Part of Odoo. See LICENSE file for full copyright and licensing details.
+
+import datetime as dt
+import logging
+import os
+import os.path
+import re
+import selectors
+import subprocess as sp
+import sys
+import threading
+import time
+from collections.abc import Sequence
+from email.utils import format_datetime
+from functools import cache
+from io import DEFAULT_BUFFER_SIZE, BytesIO
+from typing import Literal, NamedTuple
+from urllib.parse import unquote
+
+import h11
+
+from odoo.http import root
+from odoo.tools.misc import find_in_path
+
+__all__ = ["PaperMuncherInfo", "PaperMuncherServer", "paper_muncher"]
+
+_logger = logging.getLogger(__name__)
+_logger_pipe = _logger.getChild("pipe")
+_logger_process = _logger.getChild("process")
+
+SERVER_SOFTWARE = "Odoo"
+SERVER_AGENT = "Odoo"
+
+FALLBACK_BIN_PATH = "/opt/paper-muncher/bin/paper-muncher"
+WRITE_TIMEOUT = 15
+SERVE_TIMEOUT = 15 * 60
+CHUNK_SIZE = 8192
+MAX_INCOMPLETE_EVENT_SIZE = 8192
+GET_DOCUMENT_RE = re.compile(rb"^/paper-muncher/(\.|[0-9]+)\.(?:html|xhtml|xml)$")
+
+
+class PaperMuncherServer:
+ __slots__ = (
+ "_args",
+ "_conn",
+ "_deadline",
+ "_documents",
+ "_os_env",
+ "_pdf",
+ "_process",
+ "_request",
+ "_request_body",
+ "_selector",
+ "_wsgi_environ",
+ )
+
+ def __init__(self, args, os_env=None, wsgi_environ=None):
+ self._args = args
+ self._os_env = os_env
+ self._wsgi_environ = wsgi_environ or {}
+ self._process = None
+
+ def __enter__(self):
+ if self._process:
+ raise RuntimeError("process started already")
+
+ self._process = sp.Popen(
+ self._args,
+ stdin=sp.PIPE,
+ stdout=sp.PIPE,
+ stderr=(
+ sys.stderr
+ if logging.NOTSET < _logger_process.level <= logging.DEBUG
+ else sp.DEVNULL
+ ),
+ env=self._os_env,
+ )
+
+ self._conn = h11.Connection(
+ h11.SERVER,
+ max_incomplete_event_size=MAX_INCOMPLETE_EVENT_SIZE,
+ )
+ return self
+
+ def __exit__(self, *_):
+ if self._process and not self._process.poll():
+ try:
+ self._process.terminate()
+ self._process.wait(1)
+ except sp.TimeoutExpired:
+ self._process.kill()
+ self._process = None
+
+ def serve(self, documents: Sequence[str], *, timeout: int = SERVE_TIMEOUT):
+ """Serve Paper Muncher requests until the rendered PDF is returned."""
+ if not self._process:
+ raise RuntimeError(
+ "this function cannot be called outside of the context manager"
+ )
+
+ if not hasattr(threading.current_thread(), "query_count"):
+ threading.current_thread().query_count = 0
+ threading.current_thread().query_time = 0
+
+ _logger.debug("Starting request loop, %d documents available", len(documents))
+ self._deadline = time.monotonic() + timeout
+ self._documents = [
+ doc.encode() if isinstance(doc, str) else doc for doc in documents
+ ]
+ self._selector = selectors.DefaultSelector()
+ with self._selector:
+ self._selector.register(
+ self._process.stdout, selectors.EVENT_READ, data="stdout"
+ )
+
+ while self._process.poll() is None and self._selector.get_map():
+ events = self._selector.select(timeout=_remaining_time(self._deadline))
+ if events:
+ chunk = os.read(self._process.stdout.fileno(), CHUNK_SIZE)
+ if logging.NOTSET < _logger_pipe.level <= logging.DEBUG:
+ _logger_pipe.debug("read %d bytes:\n%s", len(chunk), chunk)
+ else:
+ _logger.debug("read %d bytes", len(chunk))
+ self._conn.receive_data(chunk)
+ self._process_data()
+
+ if exit_code := self._process.poll():
+ raise sp.CalledProcessError(exit_code, self._args)
+
+ return self._pdf
+
+ def _process_data(self):
+ while True:
+ event = self._conn.next_event()
+ _logger.debug(
+ "h11 current-state=%s event=%s", self._conn.states, type(event).__name__
+ )
+ if event is h11.NEED_DATA:
+ break
+ if isinstance(event, h11.Request):
+ _logger.debug(
+ "[REQ] %s %s", event.method.decode(), event.target.decode()
+ )
+ self._request = event
+ self._request_body = bytearray()
+ elif isinstance(event, h11.Data):
+ self._request_body += event.data
+ elif isinstance(event, h11.EndOfMessage):
+ try:
+ self._process_request()
+ except Exception as exc:
+ exc.add_note("upon processing %s" % self._request)
+ raise
+ if self._conn.our_state is h11.MUST_CLOSE:
+ self._selector.unregister(self._process.stdout)
+ break
+ self._conn.start_next_cycle()
+ elif isinstance(event, h11.ConnectionClosed):
+ self._selector.unregister(self._process.stdout)
+ break
+ else:
+ raise TypeError(f"unexpected {event=} in states={self._conn.states}")
+
+ def _process_request(self):
+ if self._request.method == b"GET" and (
+ match := GET_DOCUMENT_RE.match(self._request.target)
+ ):
+ self._handle_get_document(match[1])
+ elif (
+ self._request.method == b"PUT"
+ and self._request.target == b"/paper-muncher/output.pdf"
+ ):
+ self._handle_put(self._request_body)
+ _logger.debug("Got a PDF of %s bytes", len(self._request_body))
+ else:
+ self._handle_fallback(self._request, self._request_body)
+
+ def _handle_get_document(self, document_index):
+ """Serve one GET document request from the worker."""
+ index = int(document_index) if document_index != b"." else 0
+ content = self._documents[index]
+
+ response = h11.Response(
+ status_code=200,
+ headers=[
+ (
+ b"Date",
+ format_datetime(dt.datetime.now(dt.timezone.utc), usegmt=True),
+ ),
+ (b"Content-Length", str(len(content))),
+ (b"Content-Type", "text/html; charset=utf-8"),
+ (b"Server", SERVER_SOFTWARE),
+ ],
+ )
+ self._send(response)
+ self._send(h11.Data(data=content))
+ self._send(h11.EndOfMessage())
+
+ def _handle_put(self, body: bytes):
+ assert body.startswith(b"%PDF-"), body
+ self._pdf = body
+ response = h11.Response(
+ status_code=200,
+ headers=[
+ (
+ b"Date",
+ format_datetime(dt.datetime.now(dt.timezone.utc), usegmt=True),
+ ),
+ (b"Server", SERVER_SOFTWARE),
+ (b"Content-Length", "0"),
+ (b"Connection", "close"),
+ ],
+ )
+ self._send(response)
+ self._send(h11.EndOfMessage())
+ self._process.stdin.close()
+
+ def _handle_fallback(self, request: h11.Request, body: bytes):
+ assert request.target.startswith(b"/"), request.target
+ request_uri = request.target.decode("ascii")
+ path_quoted, _, query = request_uri.partition("?")
+ environ = {
+ "REQUEST_METHOD": request.method.decode("ascii"),
+ "SCRIPT_NAME": "",
+ "PATH_INFO": unquote(path_quoted, "latin-1"),
+ "QUERY_STRING": query,
+ "REQUEST_URI": request_uri,
+ "RAW_URI": request_uri,
+ "SERVER_PROTOCOL": "HTTP/1.0",
+ "SERVER_SOFTWARE": SERVER_SOFTWARE,
+ "wsgi.version": (1, 0),
+ "wsgi.url_scheme": "http",
+ "wsgi.input": BytesIO(body),
+ "wsgi.errors": sys.stderr,
+ "wsgi.multithread": False,
+ "wsgi.multiprocess": False,
+ "wsgi.run_once": False,
+ }
+ headers = {
+ "HTTP_" + header.upper().replace(b"-", b"_").decode("ascii"): value.decode(
+ "latin-1"
+ )
+ for header, value in request.headers
+ }
+ if content_type := headers.pop("HTTP_CONTENT_TYPE", ""):
+ environ["CONTENT_TYPE"] = content_type
+ if content_length := headers.pop("HTTP_CONTENT_LENGTH", ""):
+ environ["CONTENT_LENGTH"] = content_length
+ environ.update(headers)
+ environ.update(self._wsgi_environ)
+
+ response = None
+ x_sendfile = None
+
+ def start_response(status, res_headers, exc_info=None):
+ nonlocal response, x_sendfile
+ status_code = int(status.partition(" ")[0])
+ res_headers = [(_normalize_header(h), v) for h, v in res_headers]
+
+ def find_header(header):
+ return next((v for h, v in res_headers if h == header), None)
+
+ if find_header(b"Connection"):
+ raise ValueError("the WSGI app cannot set the Connection header")
+ if find_header(b"Upgrade"):
+ raise ValueError("paper-muncher does not support websocket")
+ if not find_header(b"Date"):
+ res_headers.insert(
+ 0,
+ (
+ b"Date",
+ format_datetime(dt.datetime.now(dt.timezone.utc), usegmt=True),
+ ),
+ )
+ if not find_header(b"Server"):
+ res_headers.append((b"Server", SERVER_AGENT))
+ x_sendfile = find_header(b"X-Sendfile")
+ if x_sendfile:
+ index = next(
+ (
+ i
+ for i, (h, v) in enumerate(res_headers)
+ if h == b"Content-Length"
+ )
+ )
+ res_headers[index] = (
+ b"Content-Length",
+ str(os.path.getsize(x_sendfile)),
+ )
+
+ response = h11.Response(status_code=status_code, headers=res_headers)
+ _logger.debug(
+ "[RES] %s %s", request.method.decode(), request.target.decode()
+ )
+
+ response_body = root(environ, start_response)
+ deadline = time.monotonic() + WRITE_TIMEOUT
+ self._send(response, deadline=deadline)
+
+ try:
+ if x_sendfile:
+ response_chunks = list(response_body)
+ assert not any(response_chunks), response_chunks
+ with open(x_sendfile, "rb") as f:
+ while chunk := f.read(DEFAULT_BUFFER_SIZE):
+ self._send(h11.Data(data=chunk), deadline=deadline)
+ else:
+ for chunk in response_body:
+ self._send(h11.Data(data=chunk), deadline=deadline)
+ if hasattr(response_body, "close"):
+ response_body.close()
+ self._send(h11.EndOfMessage(), deadline=deadline)
+ except Exception:
+ self._conn.send_failed()
+ raise
+
+ def _send(self, event, *, deadline=None) -> None:
+ data = self._conn.send(event)
+ memview = memoryview(data)
+ bytes_written = 0
+
+ if deadline is None:
+ deadline = time.monotonic() + WRITE_TIMEOUT
+
+ with selectors.DefaultSelector() as selector:
+ selector.register(self._process.stdin.fileno(), selectors.EVENT_WRITE)
+ while bytes_written < len(data):
+ events = selector.select(timeout=_remaining_time(deadline))
+ if not events:
+ raise TimeoutError("Timeout exceeded while writing to subprocess")
+ bytes_written += os.write(
+ self._process.stdin.fileno(), memview[bytes_written:]
+ )
+ self._process.stdin.flush()
+
+ if logging.NOTSET < _logger_pipe.level <= logging.DEBUG:
+ _logger_pipe.debug("wrote %d bytes:\n%s", bytes_written, data)
+ else:
+ _logger.debug("wrote %d bytes", bytes_written)
+
+
+def _remaining_time(deadline: float) -> float:
+ remaining = deadline - time.monotonic()
+ if remaining <= 0:
+ raise TimeoutError
+ return max(1, remaining)
+
+
+def _normalize_header(header: str | bytes) -> bytes:
+ if isinstance(header, bytes):
+ header = header.decode("ascii")
+ return header.replace("-", " ").title().replace(" ", "-").encode("ascii")
+
+
+class PaperMuncherInfo(NamedTuple):
+ state: Literal["ok", "install"]
+ bin: str
+ version: str
+
+
+@cache
+def paper_muncher() -> PaperMuncherInfo:
+ bin_path = ""
+ version = ""
+ try:
+ try:
+ bin_path = find_in_path("paper-muncher")
+ except OSError as exc:
+ if not os.path.isfile(FALLBACK_BIN_PATH):
+ raise RuntimeError("paper-muncher binary not found in PATH") from exc
+ bin_path = FALLBACK_BIN_PATH
+
+ result = sp.run(
+ [bin_path, "--version"], stdout=sp.PIPE, stderr=sp.DEVNULL, check=True
+ )
+ version = result.stdout.decode("utf-8", errors="replace").strip()
+ except (RuntimeError, OSError, sp.SubprocessError):
+ _logger.info(
+ "You need paper-muncher to print a pdf version of the reports.",
+ exc_info=_logger.isEnabledFor(logging.DEBUG),
+ )
+ return PaperMuncherInfo(state="install", bin=bin_path, version=version)
+
+ _logger.info("Will use the paper-muncher binary at %s", bin_path)
+ return PaperMuncherInfo(state="ok", bin=bin_path, version=version)
diff --git a/base_report_paper_muncher/pyproject.toml b/base_report_paper_muncher/pyproject.toml
new file mode 100644
index 00000000000..4231d0cccb3
--- /dev/null
+++ b/base_report_paper_muncher/pyproject.toml
@@ -0,0 +1,3 @@
+[build-system]
+requires = ["whool"]
+build-backend = "whool.buildapi"
diff --git a/base_report_paper_muncher/readme/CONFIGURE.md b/base_report_paper_muncher/readme/CONFIGURE.md
new file mode 100644
index 00000000000..dfe7ae69601
--- /dev/null
+++ b/base_report_paper_muncher/readme/CONFIGURE.md
@@ -0,0 +1,23 @@
+The system parameter `report.pdf_engine` controls which PDF engine is used:
+
+| Value | Behavior |
+| ----- | -------- |
+| `auto` (default) | Use Paper Muncher when the binary is available, otherwise wkhtmltopdf |
+| `paper-muncher` | Force Paper Muncher; raise an error if the binary is missing |
+| `wkhtmltopdf` | Always use wkhtmltopdf |
+
+Change it from code:
+
+```python
+env["ir.config_parameter"].sudo().set_param("report.pdf_engine", "auto")
+```
+
+Or from **Settings > Technical > Parameters > System Parameters**, key
+`report.pdf_engine`.
+
+Optional environment variables for the Odoo process:
+
+- `ODOO_PAPER_MUNCHER_FEATURE=1`: pass `--feature *=on` to the binary
+- `ODOO_PAPER_MUNCHER_DEBUG=1`: pass `--debug http-client` to the binary. This
+ flag is not sent by default in production. It is also enabled automatically
+ when the module logger is set to `DEBUG`.
diff --git a/base_report_paper_muncher/readme/CONTRIBUTORS.md b/base_report_paper_muncher/readme/CONTRIBUTORS.md
new file mode 100644
index 00000000000..957ad99ee32
--- /dev/null
+++ b/base_report_paper_muncher/readme/CONTRIBUTORS.md
@@ -0,0 +1 @@
+- Felix Coca \<