Implement audio file compression

README.md

@@ -11,7 +11,10 @@ poetry run repub
 - [x] Offlines RSS feed xml
 - [x] Downloads media and enclosures
 - [x] Rewrites media urls
-- [ ] Media compression
+- [x] Image normalization (JPG, RGB)
+- [x] Audio compression
+- [ ] Image compression
+- [ ] Video compression
 - [ ] Download and rewrite media embedded in content/CDATA fields
 - [ ] Config file to drive the program
 - [ ] Daemonize the program

poetry.lock

@@ -411,6 +411,23 @@ files = [
 [package.dependencies]
 sgmllib3k = "*"
 
+[[package]]
+name = "ffmpeg-python"
+version = "0.2.0"
+description = "Python bindings for FFmpeg - with complex filtering support"
+optional = false
+python-versions = "*"
+files = [
+    {file = "ffmpeg-python-0.2.0.tar.gz", hash = "sha256:65225db34627c578ef0e11c8b1eb528bb35e024752f6f10b78c011f6f64c4127"},
+    {file = "ffmpeg_python-0.2.0-py3-none-any.whl", hash = "sha256:ac441a0404e053f8b6a1113a77c0f452f1cfc62f6344a769475ffdc0f56c23c5"},
+]
+
+[package.dependencies]
+future = "*"
+
+[package.extras]
+dev = ["Sphinx (==2.1.0)", "future (==0.17.1)", "numpy (==1.16.4)", "pytest (==4.6.1)", "pytest-mock (==1.10.4)", "tox (==3.12.1)"]
+
 [[package]]
 name = "filelock"
 version = "3.13.4"
@@ -461,6 +478,17 @@ flake8 = ">=3"
 [package.extras]
 develop = ["build", "twine"]
 
+[[package]]
+name = "future"
+version = "1.0.0"
+description = "Clean single-source support for Python 3 and 2"
+optional = false
+python-versions = ">=2.6, !=3.0.*, !=3.1.*, !=3.2.*"
+files = [
+    {file = "future-1.0.0-py3-none-any.whl", hash = "sha256:929292d34f5872e70396626ef385ec22355a1fae8ad29e1a734c3e43f9fbc216"},
+    {file = "future-1.0.0.tar.gz", hash = "sha256:bd2968309307861edae1458a4f8a4f3598c03be43b97521076aebf5d94c07b05"},
+]
+
 [[package]]
 name = "hyperlink"
 version = "21.0.0"
@@ -1600,4 +1628,4 @@ testing = ["coverage (>=5.0.3)", "zope.event", "zope.testing"]
 [metadata]
 lock-version = "2.0"
 python-versions = "^3.11"
-content-hash = "8b12b19145242fe86f09024453bca29792f6e22b4e63cfc72e2c6e480f38f043"
+content-hash = "4c6b23f66fa6309a313c7a054e640ec31ce9207e5b3dd9301e06ae9b9fb44f79"

pyproject.toml

@@ -17,6 +17,7 @@ colorlog = "^6.8.2"
 feedparser = "^6.0.11"
 lxml = "^5.2.1"
 pillow = "^10.3.0"
+ffmpeg-python = "^0.2.0"
 
 
 [build-system]

repub/media.py

@@ -0,0 +1,67 @@
+import logging
+import math
+
+import ffmpeg
+
+logger = logging.getLogger(__name__)
+
+
+def media_info(file_path):
+    return ffmpeg.probe(file_path)
+
+
+def bitrate(info) -> float:
+    try:
+        return int(info["format"]["bit_rate"])
+    except KeyError | ValueError:
+        return math.inf
+
+
+def format(info):
+    try:
+        return info["format"]["format_name"]
+    except KeyError | ValueError:
+        return None
+
+
+def compression_settings(input_file, settings):
+    info = media_info(input_file)
+    br = settings.get("REPUBLISHER_AUDIO_BITRATE", 96000)
+    fmt = settings.get("REPUBLISHER_AUDIO_FORMAT", "mp3")
+    if bitrate(info) <= br:
+        is_br = True
+    else:
+        is_br = False
+    if format(info) == fmt:
+        is_fmt = True
+    else:
+        is_fmt = False
+
+    if is_br and is_fmt:
+        return None
+
+    if is_br:
+        target_br = bitrate(info)
+    else:
+        target_br = br
+    return {"bitrate": target_br, "ext": "mp3"}
+
+
+def compress_audio(input_file, output_file_base, settings):
+    ext = settings["ext"]
+    br = settings["bitrate"]
+    output_file = f"{output_file_base}.{ext}"
+    try:
+        logger.info(f"Compressing audio {input_file} to {output_file} target_br={br}")
+        out, _ = (
+            ffmpeg.input(input_file)
+            .output(
+                output_file,
+                **{"b:a": f"{br}", "map": "0:a:0"},
+                loglevel="quiet",
+            )
+            .run()
+        )
+        return output_file
+    except ffmpeg.Error as e:
+        raise RuntimeError(f"Failed to load audio: {e.stderr.decode()}") from e

repub/pipelines.py

@@ -1,12 +1,20 @@
+import logging
+import tempfile
+from io import BytesIO
 from os import PathLike
 from pathlib import PurePosixPath
-from typing import IO, DefaultDict, Optional, Set, Union
+from typing import IO, DefaultDict, Dict, Optional, Set, Tuple, Union
 from urllib.parse import urlparse
 
 import repub.utils
+from repub import media
 from repub.exporters import RssExporter
-from scrapy.pipelines.images import FilesPipeline as BaseFilesPipeline
+from scrapy.pipelines.files import FileException
+from scrapy.pipelines.files import FilesPipeline as BaseFilesPipeline
 from scrapy.pipelines.images import ImagesPipeline as BaseImagesPipeline
+from scrapy.utils.misc import md5sum
+
+logger = logging.getLogger(__name__)
 
 
 class ImagePipeline(BaseImagesPipeline):
@@ -32,6 +40,49 @@ class AudioPipeline(BaseFilesPipeline):
     def file_path(self, request, response=None, info=None, *, item=None):
         return repub.utils.local_audio_path(request.url)
 
+    def file_downloaded(self, response, request, info, *, item=None):
+        return self.audio_downloaded(response, request, info, item=item)
+
+    def audio_downloaded(self, response, request, info, *, item=None):
+        checksum = None
+        for path, buf in self.get_audio(response, request, info, item=item):
+            if checksum is None:
+                buf.seek(0)
+                checksum = md5sum(buf)
+            # width, height = image.size
+            self.store.persist_file(
+                path,
+                buf,
+                info,
+                # meta={"width": width, "height": height},
+                headers={"Content-Type": "audio/mp3"},
+            )
+        return checksum
+
+    def get_audio(self, response, request, info, *, item=None):
+        path = self.file_path(request, response=response, info=info, item=item)
+        buf = BytesIO(response.body)
+        with tempfile.TemporaryDirectory() as tmpdir:
+            tmp_file = f"{tmpdir}/file"
+            converted_file_base = f"{tmpdir}/converted"
+            with open(tmp_file, "wb") as f:
+                f.write(buf.read())
+
+            s = media.compression_settings(tmp_file, {})
+            if s is not None:
+                converted_file = media.compress_audio(tmp_file, converted_file_base, s)
+                buf_converted = BytesIO()
+                with open(converted_file, "rb") as f:
+                    buf_converted.write(f.read())
+                buf_converted.seek(0)
+                yield path, buf_converted
+            else:
+                logger.info(
+                    f"Skipping audio compression for {path}, it meets requirements"
+                )
+                buf.seek(0)
+                yield path, buf
+
 
 class VideoPipeline(BaseFilesPipeline):
     def __init__(self, store_uri: Union[str, PathLike], **kwargs):

shell.nix

@@ -1,14 +1,24 @@
-{ system ? "x86_64-linux", pkgs ? import <nixpkgs> { inherit system; }, dev ? true, }:
+{
+  system ? "x86_64-linux",
+  pkgs ? import <nixpkgs> { inherit system; },
+  dev ? true,
+}:
 
 let
   pyCurrent = pkgs.python311;
   poetryExtras = if dev then [ "dev" ] else [ ];
   poetryInstallExtras = (
-    if poetryExtras == [] then ""
+    if poetryExtras == [ ] then
-    else pkgs.lib.concatStrings [ " --with=" (pkgs.lib.concatStringsSep "," poetryExtras) ]
+      ""
+    else
+      pkgs.lib.concatStrings [
+        " --with="
+        (pkgs.lib.concatStringsSep "," poetryExtras)
+      ]
   );
   packages = [
-    pyCurrent
+    pkgs.ffmpeg_5-headless
+    #(pyCurrent (ps: with ps; [ ffmpeg-python ]))
     pkgs.zsh
     (pkgs.poetry.withPlugins (ps: with ps; [ poetry-plugin-up ]))
   ];