Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
196 changes: 196 additions & 0 deletions .github/workflows/fix-encoding-errors.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,196 @@
name: Build FixEncodingErrors (SE5)

on:
push:
branches: [main]
paths:
- 'se5/FixEncodingErrors/**'
- 'se5/FixEncodingErrors.Tests/**'
- 'se5/Plugin-Shared/**'
- '.github/workflows/fix-encoding-errors.yml'
pull_request:
paths:
- 'se5/FixEncodingErrors/**'
- 'se5/FixEncodingErrors.Tests/**'
- 'se5/Plugin-Shared/**'
- '.github/workflows/fix-encoding-errors.yml'
workflow_dispatch:
inputs:
release:
description: 'Publish a GitHub release. Bumps the minor in plugin.json and commits it back.'
type: boolean
default: false
version:
description: 'Optional explicit version (e.g. 1.2.0). Overrides the auto-minor-bump.'
required: false
type: string

permissions:
contents: write

jobs:
prepare:
runs-on: ubuntu-latest
outputs:
version: ${{ steps.set.outputs.version }}
tag: ${{ steps.set.outputs.tag }}
ref: ${{ steps.set.outputs.ref }}
release: ${{ steps.set.outputs.release }}
steps:
- name: Checkout
uses: actions/checkout@v4

- name: Resolve version (and bump on release)
id: set
run: |
set -e
current=$(jq -r .version se5/FixEncodingErrors/plugin.json)
release_flag="false"
ref="${{ github.sha }}"
version="$current"
tag=""

if [ "${{ github.event_name }}" = "workflow_dispatch" ] && [ "${{ inputs.release }}" = "true" ]; then
release_flag="true"
if [ -n "${{ inputs.version }}" ]; then
version="${{ inputs.version }}"
else
IFS=. read -r major minor patch <<< "$current"
version="$major.$((minor + 1)).0"
fi

tmp=$(mktemp)
jq --arg v "$version" '.version = $v' se5/FixEncodingErrors/plugin.json > "$tmp"
mv "$tmp" se5/FixEncodingErrors/plugin.json

git config user.name "github-actions[bot]"
git config user.email "41898282+github-actions[bot]@users.noreply.github.com"
git add se5/FixEncodingErrors/plugin.json
git commit -m "Bump FixEncodingErrors to $version [skip ci]"

# Retry-with-rebase: if a concurrent release dispatched for a different
# plugin won the push first, pull --rebase and try again.
attempts=0
while ! git push 2>/tmp/push.err; do
attempts=$((attempts + 1))
if [ "$attempts" -ge 5 ]; then
echo "Failed to push after $attempts attempts:" >&2
cat /tmp/push.err >&2
exit 1
fi
echo "Push rejected (attempt $attempts), pulling --rebase and retrying..."
git pull --rebase --no-edit
done
ref=$(git rev-parse HEAD)

IFS=. read -r major minor patch <<< "$version"
tag="se5-fix-encoding-errors-v$major.$minor"
fi

echo "version=$version" >> "$GITHUB_OUTPUT"
echo "tag=$tag" >> "$GITHUB_OUTPUT"
echo "ref=$ref" >> "$GITHUB_OUTPUT"
echo "release=$release_flag" >> "$GITHUB_OUTPUT"

test:
runs-on: ubuntu-latest
steps:
- name: Checkout
uses: actions/checkout@v4

- name: Setup .NET 8
uses: actions/setup-dotnet@v4
with:
dotnet-version: '8.0.x'

- name: Test
run: dotnet test se5/FixEncodingErrors.Tests/FixEncodingErrors.Tests.csproj -c Release

build:
needs: [prepare, test]
runs-on: ubuntu-latest
strategy:
fail-fast: false
matrix:
include:
- rid: win-x64
os: windows
exe: FixEncodingErrors.exe
- rid: win-arm64
os: windows
exe: FixEncodingErrors.exe
- rid: linux-x64
os: linux
exe: FixEncodingErrors
- rid: linux-arm64
os: linux
exe: FixEncodingErrors
- rid: osx-x64
os: macos
exe: FixEncodingErrors
- rid: osx-arm64
os: macos
exe: FixEncodingErrors

steps:
- name: Checkout
uses: actions/checkout@v4
with:
ref: ${{ needs.prepare.outputs.ref }}

- name: Setup .NET 8
uses: actions/setup-dotnet@v4
with:
dotnet-version: '8.0.x'

- name: Publish self-contained (${{ matrix.rid }})
working-directory: se5/FixEncodingErrors
run: |
dotnet publish FixEncodingErrors.csproj \
-c Release \
-r ${{ matrix.rid }} \
--self-contained true \
-p:DebugType=None \
-p:DebugSymbols=false \
-o staging/FixEncodingErrors

- name: Rewrite plugin.json for ${{ matrix.os }}
working-directory: se5/FixEncodingErrors
run: |
jq '
del(.runtime, .entry) |
.executables = { "${{ matrix.os }}": "${{ matrix.exe }}" }
' plugin.json > staging/FixEncodingErrors/plugin.json

- name: Package zip
working-directory: se5/FixEncodingErrors/staging
run: zip -r "$GITHUB_WORKSPACE/FixEncodingErrors-${{ matrix.rid }}.zip" FixEncodingErrors

- name: Upload artifact
uses: actions/upload-artifact@v4
with:
name: FixEncodingErrors-${{ matrix.rid }}
path: FixEncodingErrors-${{ matrix.rid }}.zip

release:
name: Publish GitHub Release
needs: [prepare, build]
if: needs.prepare.outputs.release == 'true'
runs-on: ubuntu-latest
steps:
- name: Download all zips
uses: actions/download-artifact@v4
with:
path: dist
pattern: FixEncodingErrors-*
merge-multiple: true

- name: Create release and upload zips
env:
GH_TOKEN: ${{ github.token }}
run: |
gh release create "${{ needs.prepare.outputs.tag }}" dist/FixEncodingErrors-*.zip \
--repo "${{ github.repository }}" \
--target "${{ needs.prepare.outputs.ref }}" \
--title "FixEncodingErrors v${{ needs.prepare.outputs.version }}" \
--notes "Self-contained FixEncodingErrors plugin builds for win/linux/osx (x64 + arm64)."
1 change: 1 addition & 0 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@ index [`se5-plugins.json`](se5-plugins.json).
| [ARTE check](se5/ArteCheck) | Checks and fixes EBU STL teletext subtitles for ARTE delivery. |
| [American to British](se5/AmericanToBritish) | Converts American English spellings to British English. |
| [British to American](se5/BritishToAmerican) | Converts British English spellings to American English. |
| [Fix encoding errors](se5/FixEncodingErrors) | Repairs UTF-8 text read with the wrong code page: résumé → résumé, don’t → don’t. |
| [Haxor](se5/Haxor) | Translates text to "haxor" - the minimal reference plugin. |
| [Italic each line](se5/ItalicEachLine) | Switches italic tags between one tag per subtitle and a tag on each line. |
| [Persian Subtitle Fixes](se5/PersianErrors) | Fixes common errors in Persian (Farsi) subtitles - 1,500+ rules in 12 groups. |
Expand Down
140 changes: 140 additions & 0 deletions se5/FixEncodingErrors.Tests/EncodingFixerTests.cs
Original file line number Diff line number Diff line change
@@ -0,0 +1,140 @@
using SubtitleEdit.Plugins.FixEncodingErrors;
using System.Text;

namespace FixEncodingErrors.Tests;

public class EncodingFixerTests
{
static EncodingFixerTests() => Encoding.RegisterProvider(CodePagesEncodingProvider.Instance);

/// <summary>What a UTF-8 text looks like after being read with <paramref name="codePage"/>.</summary>
private static string Break(string text, int codePage = 1252) =>
Encoding.GetEncoding(codePage).GetString(Encoding.UTF8.GetBytes(text));

[Theory]
[InlineData("résumé", "résumé")]
[InlineData("don’t", "don’t")]
[InlineData("“Quoteâ€\u009D", "“Quote”")]
[InlineData("“Quoteâ€", "“Quote”")] // 0x9D lost
[InlineData("25°C", "25°C")]
[InlineData("Hello", "Hello")]
[InlineData("Wait à bientôt", "Wait à bientôt")] // A0 turned into a space
[InlineData("😀 hi", "😀 hi")]
public void FixesWestern(string broken, string expected)
{
Assert.Equal(expected, EncodingFixer.Fix(broken, 1252, confirmed: true));
}

[Theory]
[InlineData("cache of such communiquã©s,", "cache of such communiqués,")]
[InlineData("and dust off my rã©sumã©.", "and dust off my résumé.")]
[InlineData("madame la secrã©taire.", "madame la secrétaire.")]
public void FixesTextLowercasedAfterItBroke(string broken, string expected)
{
// Ed's Madam Secretary subtitles: © alone tells the subtitle is broken.
Assert.Equal(expected, EncodingFixer.Fix(broken, 1252, confirmed: false));
}

[Theory]
[InlineData("Ça va? Où êtes-vous? À bientôt, Œdipe!")]
[InlineData("Größe, Übermaß, naïve, Señor, Ñandú, Ærø")]
[InlineData("“Smart quotes” – dashes — and… ellipsis ‘ok’ 😀")]
public void RoundTripsWestern(string text)
{
Assert.Equal(text, EncodingFixer.Fix(Break(text), 1252, confirmed: true));
}

[Theory]
[InlineData("Привет, как дела? Всё хорошо.", 1251)]
[InlineData("Příliš žluťoučký kůň úpěl ďábelské ódy.", 1250)]
[InlineData("Καλημέρα κόσμε", 1253)]
[InlineData("Günaydın, nasılsın? Şimdi değil.", 1254)]
public void RoundTripsOtherCodePages(string text, int codePage)
{
// Detection may pick a code page that reads the same bytes the same way (1252/1254 for Turkish).
var broken = Break(text, codePage);
var detected = EncodingFixer.Detect(new[] { broken }).CodePage;
Assert.Equal(text, EncodingFixer.Fix(broken, detected, confirmed: true));
Assert.Equal(text, EncodingFixer.Fix(broken, codePage, confirmed: true));
}

[Fact]
public void DetectsCyrillic()
{
Assert.Equal(1251, EncodingFixer.Detect(new[] { Break("Привет, как дела?", 1251) }).CodePage);
}

[Theory]
[InlineData("Привет, как дела?")]
[InlineData("Καλημέρα κόσμε")]
[InlineData("שלום עולם")]
[InlineData("مرحبا بالعالم")]
public void FixesOtherScriptsMisreadAsWestern(string text)
{
var broken = Break(text);
Assert.Equal(1252, EncodingFixer.Detect(new[] { broken }).CodePage);
Assert.Equal(text, EncodingFixer.Fix(broken, 1252, EncodingFixer.IsBroken(new[] { broken }, 1252)));
}

[Fact]
public void FixesDoubleBroken()
{
Assert.Equal("café", EncodingFixer.Fix(Break(Break("café")), 1252, confirmed: true));
}

[Theory]
[InlineData("Não”")]
[InlineData("A minha irmã… e a mãe")]
[InlineData("«Não»")]
[InlineData("Irmã Lúcia")]
[InlineData("Ça coûte 25 €, café»")]
[InlineData("Plain ASCII text")]
public void LeavesRealTextAlone(string text)
{
Assert.Equal(0, EncodingFixer.CountTelling(new[] { text }, 1252));
Assert.Equal(text, EncodingFixer.Fix(text, 1252, confirmed: false));
}

[Theory]
[InlineData("Привет, Сёма! В доме.")]
[InlineData("Příliš žluťoučký kůň")]
public void LeavesRealTextAloneInItsOwnCodePage(string text)
{
Assert.Equal(1252, EncodingFixer.Detect(new[] { text }).CodePage);
Assert.Equal(text, EncodingFixer.Fix(text, 1252, confirmed: false));
}

[Fact]
public void AmbiguousMatchNeedsConfirmation()
{
// "Ö" (Ö) at a word end could be real text; fixed once the subtitle is known to be broken.
Assert.Equal("BÖ", EncodingFixer.Fix("BÖ", 1252, confirmed: false));
Assert.Equal("BÖ", EncodingFixer.Fix("BÖ", 1252, confirmed: true));
Assert.True(EncodingFixer.IsBroken(new[] { "Österreich", "café" }, 1252));
Assert.False(EncodingFixer.IsBroken(new[] { "Österreich" }, 1252));
}

[Fact]
public void DetectsWesternForEd()
{
var lines = new[] { "and dust off my rã©sumã©.", "I'll send you his rã©sumã©." };
Assert.Equal(1252, EncodingFixer.Detect(lines).CodePage);
Assert.Equal(4, EncodingFixer.CountTelling(lines, 1252));
}

[Fact]
public void RealMacedonianIsNotBroken()
{
// ј, џ, ќ, ѓ are continuation bytes in Windows-1251: "ој" must not become Greek "μ", "рџ" not "п".
var lines = new[] { "Тој е мојот брат.", "Џорџ веќе учење, која боја." };
Assert.False(EncodingFixer.IsBroken(lines, 1251));
Assert.Equal(1252, EncodingFixer.Detect(lines).CodePage);
Assert.All(lines, l => Assert.Equal(l, EncodingFixer.Fix(l, 1251, confirmed: false)));
}

[Fact]
public void KeepsTagsAndLineBreaks()
{
Assert.Equal("<i>résumé</i>\r\n{\\an8}café", EncodingFixer.Fix("<i>résumé</i>\r\n{\\an8}café", 1252, confirmed: false));
}
}
27 changes: 27 additions & 0 deletions se5/FixEncodingErrors.Tests/FixEncodingErrors.Tests.csproj
Original file line number Diff line number Diff line change
@@ -0,0 +1,27 @@
<Project Sdk="Microsoft.NET.Sdk">

<PropertyGroup>
<TargetFramework>net8.0</TargetFramework>
<Nullable>enable</Nullable>
<ImplicitUsings>enable</ImplicitUsings>
<IsPackable>false</IsPackable>
</PropertyGroup>

<ItemGroup>
<Using Include="Xunit" />
</ItemGroup>

<PropertyGroup>
</PropertyGroup>

<ItemGroup>
<PackageReference Include="Microsoft.NET.Test.Sdk" Version="17.11.1" />
<PackageReference Include="xunit" Version="2.9.2" />
<PackageReference Include="xunit.runner.visualstudio" Version="2.8.2" />
</ItemGroup>

<ItemGroup>
<ProjectReference Include="..\FixEncodingErrors\FixEncodingErrors.csproj" />
</ItemGroup>

</Project>
Loading
Loading