7 Commits
Author SHA1 Message Date
kato 76bc3d8185 ci: add nodejs24 and nodejs24-bin to CI dependencies
Tests / Go and GTK tests (push) Successful in 1m38s
Release / Tests before release (push) Successful in 2m5s
Release / Build and publish release (push) Successful in 3m23s
2026-07-16 20:48:42 +03:00
kato 449626bffa docs: add CI/release documentation and ignore dist directory
Tests / Go and GTK tests (push) Failing after 1m8s
Add `/dist/` to `.gitignore` to exclude locally built distribution
artifacts. Document the Gitea Actions workflow, release tagging, and
branch protection setup in the README.
2026-07-16 19:40:21 +03:00
kato 13ddebbf98 feat(overlay): improve provisional text display and line splitting 2026-07-16 19:27:39 +03:00
kato 2f5ca41ac3 feat: add --overlay-font-family option for caption font customization
Allow users to specify a custom font family for the overlay captions via
a new --overlay-font-family flag, defaulting to "Sans". Validation
ensures the font family is not empty. Updated README, config, and tests.
2026-07-16 18:09:09 +03:00
kato 192666dd43 fix(overlay): enforce text wrap width to prevent overflow beyond screen
Add `gtk_label_set_width_chars` to set a concrete width for label allocation,
ensuring long sentences wrap within the configured max-chars bubble width
instead of extending off-screen.
2026-07-16 18:06:00 +03:00
kato bd2fc175a8 feat(overlay): add gradual fade effect for aging caption lines
Updated the overlay sink and state to apply a gradual fade from white to gray for completed caption lines over their display lifetime. The `FinalMarkup` method replaces plain `FinalText`, generating Pango markup with decreasing brightness. A periodic refresh timer (200 ms) recalculates colors and expiry, improving visual clarity of line recency. README updated to document the new fade behavior.
2026-07-16 18:03:23 +03:00
kato 55181d6397 feat: add six-line caption cap and enforce 3s minimum final timeout
- Reduce default overlay-final-timeout from 4s to 3s and validate
  that non-zero timeout is at least 3s to prevent captions from
  disappearing too quickly.
- Limit completed captions to six lines and keep each visible for
  at least 3s, improving readability during fast speech.
- Update README to document new VAD behavior and changed defaults.
2026-07-16 18:00:11 +03:00
13 changed files with 750 additions and 64 deletions
+115
View File
@@ -0,0 +1,115 @@
name: Release
on:
push:
tags:
- 'v*'
permissions:
code: read
jobs:
test:
name: Tests before release
runs-on: ubuntu-latest
container: registry.fedoraproject.org/fedora:44
steps:
- name: Validate release tag
shell: bash
run: |
set -Eeuo pipefail
[[ "${GITHUB_REF_NAME}" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]] || {
printf 'Tag must match vMAJOR.MINOR.BUG; got %s\n' "${GITHUB_REF_NAME}" >&2
exit 1
}
- name: Install system dependencies
run: >-
dnf install -y
bzip2 curl gcc gcc-c++ git gtk4-devel gtk4-layer-shell-devel
jq libX11-devel libXfixes-devel nodejs24 nodejs24-bin pkgconf-pkg-config
pulseaudio-utils ShellCheck tar gzip binutils
- name: Check out repository
uses: https://github.com/actions/checkout@v4
- name: Install the Go version from go.mod
shell: bash
run: |
set -Eeuo pipefail
go_version="$(awk '$1 == "go" { print $2 }' go.mod)"
curl --fail --location --silent --show-error \
"https://go.dev/dl/go${go_version}.linux-amd64.tar.gz" \
--output /tmp/go.tar.gz
rm -rf /usr/local/go
tar -C /usr/local -xzf /tmp/go.tar.gz
ln -sf /usr/local/go/bin/go /usr/local/bin/go
ln -sf /usr/local/go/bin/gofmt /usr/local/bin/gofmt
go version
- name: Run tests and static checks
run: ./scripts/ci/test.sh
publish:
name: Build and publish release
needs: test
permissions:
code: read
releases: write
runs-on: ubuntu-latest
container: registry.fedoraproject.org/fedora:44
steps:
- name: Install system dependencies
run: >-
dnf install -y
bzip2 curl gcc gcc-c++ git gtk4-devel gtk4-layer-shell-devel
jq libX11-devel libXfixes-devel nodejs24 nodejs24-bin pkgconf-pkg-config
pulseaudio-utils ShellCheck tar gzip binutils
- name: Check out repository and tags
uses: https://github.com/actions/checkout@v4
with:
fetch-depth: 0
- name: Install the Go version from go.mod
shell: bash
run: |
set -Eeuo pipefail
go_version="$(awk '$1 == "go" { print $2 }' go.mod)"
curl --fail --location --silent --show-error \
"https://go.dev/dl/go${go_version}.linux-amd64.tar.gz" \
--output /tmp/go.tar.gz
rm -rf /usr/local/go
tar -C /usr/local -xzf /tmp/go.tar.gz
ln -sf /usr/local/go/bin/go /usr/local/bin/go
ln -sf /usr/local/go/bin/gofmt /usr/local/bin/gofmt
go version
- name: Build portable release archive
run: ./scripts/ci/package-release.sh "${GITHUB_REF_NAME}"
- name: Generate release notes
shell: bash
run: |
set -Eeuo pipefail
previous_tag="$(git describe --tags --abbrev=0 "${GITHUB_REF_NAME}^" --match 'v[0-9]*' 2>/dev/null || true)"
{
printf 'Automated Captioneer release for `%s`.\n\n' "${GITHUB_REF_NAME}"
printf 'The Linux x86_64 archive contains both Captioneer commands and their Sherpa/ONNX native libraries. Models are downloaded separately with `scripts/download-models.sh`.\n'
if [[ -n "$previous_tag" ]]; then
printf '\nChanges since `%s`:\n\n' "$previous_tag"
git log --no-merges --pretty='- %s (`%h`)' "${previous_tag}..${GITHUB_REF_NAME}"
fi
} > dist/release-notes.md
- name: Publish Gitea release
env:
GITEA_API_URL: ${{ gitea.api_url }}
GITEA_REPOSITORY: ${{ gitea.repository }}
GITEA_TOKEN: ${{ secrets.GITEA_TOKEN }}
RELEASE_TAG: ${{ gitea.ref_name }}
RELEASE_NOTES: dist/release-notes.md
run: >-
./scripts/ci/publish-release.sh
"dist/captioneer-${GITHUB_REF_NAME}-linux-amd64.tar.gz"
"dist/captioneer-${GITHUB_REF_NAME}-linux-amd64.tar.gz.sha256"
+43
View File
@@ -0,0 +1,43 @@
name: Tests
on:
pull_request:
push:
branches:
- '**'
permissions:
code: read
jobs:
test:
name: Go and GTK tests
runs-on: ubuntu-latest
container: registry.fedoraproject.org/fedora:44
steps:
- name: Install system dependencies
run: >-
dnf install -y
bzip2 curl gcc gcc-c++ git gtk4-devel gtk4-layer-shell-devel
jq libX11-devel libXfixes-devel nodejs24 nodejs24-bin pkgconf-pkg-config
pulseaudio-utils ShellCheck tar gzip binutils
- name: Check out repository
uses: https://github.com/actions/checkout@v4
- name: Install the Go version from go.mod
shell: bash
run: |
set -Eeuo pipefail
go_version="$(awk '$1 == "go" { print $2 }' go.mod)"
curl --fail --location --silent --show-error \
"https://go.dev/dl/go${go_version}.linux-amd64.tar.gz" \
--output /tmp/go.tar.gz
rm -rf /usr/local/go
tar -C /usr/local -xzf /tmp/go.tar.gz
ln -sf /usr/local/go/bin/go /usr/local/bin/go
ln -sf /usr/local/go/bin/gofmt /usr/local/bin/gofmt
go version
- name: Run tests and static checks
run: ./scripts/ci/test.sh
+1
View File
@@ -5,3 +5,4 @@ models/silero_vad_v5.onnx
# Locally built executable.
/captioneer
/build/
/dist/
+23 -5
View File
@@ -64,7 +64,9 @@ Use `--output=both` to keep terminal output as well:
./scripts/distrobox/run.sh --mode=vad --output=both
```
The overlay shows the latest final caption above the active provisional caption. Final captions remain for four seconds by default. The dark background is 90% opaque, text is centered and limited to two lines per row, and pointer input passes through the window by default. Empty completed captions hide the overlay.
In VAD mode, a 500 ms pause finalizes the current speech segment. Captioneer splits that transcript again at sentence punctuation and at word-safe boundaries of at most 64 characters, so continuous speech does not become one enormous paragraph. The live provisional view keeps the latest three digestible chunks and grows vertically without ellipsizing them. Completed chunks remain visible for at least three seconds, receive extra time based on their word count, and gradually fade from white to a restrained gray as they age. When several chunks finalize together, their removal is staggered in reading order instead of deleting the whole block at once. Up to six completed chunks remain above the provisional text; if a seventh arrives, the oldest is removed. Blank recognition results never erase completed text before its reading lifetime expires.
Text is left-aligned inside the centered, fixed-width dark bubble and wraps at its configured width instead of extending off-screen. Pointer input passes through the window by default.
To build outside the helper script, install GTK4 and GTK4 Layer Shell development packages, then run:
@@ -105,18 +107,19 @@ Layer Shell and X11 hints are requests, not universal guarantees. Same-layer win
| `--overlay-backend=auto|wayland|x11` | `auto` | Let GTK detect the backend, or require one explicitly. |
| `--overlay-opacity=0.90` | `0.90` | Dark bubble opacity from `0` to `1`. |
| `--overlay-font-size=28` | `28` | Font size in logical pixels. |
| `--overlay-font-family=Sans` | `Sans` | GTK/Pango font family for final and provisional captions. |
| `--overlay-bottom-margin=100` | `100` | Bottom margin in logical pixels. |
| `--overlay-max-width=1200` | `1200` | Fixed caption-bubble width; it is capped at 80% of the selected monitor. |
| `--overlay-monitor=auto` | `auto` | `auto`, a zero-based monitor index, or a connector name such as `DP-1`. |
| `--overlay-final-timeout=4s` | `4s` | Time before the final row disappears. `0s` keeps it until replaced or hidden. |
| `--overlay-final-timeout=3s` | `3s` | Completed-caption lifetime; it must be `0s` or at least `3s`. `0s` keeps lines until the six-line cap removes them. |
| `--overlay-click-through=true` | `true` | Use an empty input region where the display backend supports it. |
Examples:
```bash
# Larger text, longer final-caption visibility
# Larger text, a custom font, and longer final-caption visibility
./scripts/distrobox/run.sh --mode=vad --output=overlay \
--overlay-font-size=34 --overlay-final-timeout=6s
--overlay-font-size=34 --overlay-font-family="Noto Sans" --overlay-final-timeout=6s
# Explicit XWayland/X11 route on the cursor's monitor
./scripts/distrobox/run.sh --mode=vad --output=both --overlay-backend=x11
@@ -198,8 +201,23 @@ distrobox enter --no-workdir captioneer-dev -- bash -lc '
Run this from the repository root. The build and run helpers resolve that path automatically.
## Gitea Actions and releases
The `Tests` workflow runs the normal and GTK-tagged Go tests, `go vet`, and ShellCheck for every pull request and branch push. To make it mandatory for pull requests, open the repository's **Settings → Branches**, edit the protection rule for `master`, enable required status checks, and select **Tests / Go and GTK tests** after its first run. Workflow files can report the check, but the branch-protection rule is what prevents merging when it fails.
Pushing a tag that exactly follows `vMAJOR.MINOR.BUG` runs the same test suite first, then builds and publishes a Gitea release:
```bash
git tag v1.3.519
git push origin v1.3.519
```
The release contains a checksummed Linux x86_64 archive with `captioneer`, `captioneer-desktop`, the required Sherpa/ONNX shared libraries, the model download script, the README, and the license. The desktop command still requires GTK4 and GTK4 Layer Shell runtime libraries on the destination system.
The release workflow uses Gitea's built-in `GITEA_TOKEN`; no personal token is required. In **Settings → Actions → General**, allow the workflow token `Read` access to code and `Write` access to releases. Gitea clamps workflow permissions to those repository settings, so a more restrictive maximum produces an actionable HTTP 403 during publishing.
## License
Captioneer is licensed under the [Apache License 2.0](LICENSE). Sherpa-Onnx is Apache-2.0 licensed, GTK4 is LGPL-2.1-or-later, and GTK4 Layer Shell is MIT licensed. Captioneer links to the system GTK libraries rather than copying their source into the project.
Version 1 intentionally has no animation, caption history, runtime dragging, background-free mode, capture exclusion, Flatpak, desktop entry, or autostart integration.
Version 1 intentionally has no animation, persisted caption history, runtime dragging, background-free mode, capture exclusion, Flatpak, desktop entry, or autostart integration.
+51
View File
@@ -0,0 +1,51 @@
#!/usr/bin/env bash
set -Eeuo pipefail
repo_root="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")/../.." && pwd -P)"
cd -- "$repo_root"
tag=${1:-}
if [[ ! "$tag" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
printf 'captioneer: release tag must match vMAJOR.MINOR.BUG, got %q\n' "$tag" >&2
exit 1
fi
platform=linux-amd64
archive_name="captioneer-${tag}-${platform}.tar.gz"
package_name="captioneer-${tag}-${platform}"
package_dir="dist/$package_name"
rm -rf -- "$package_dir"
rm -f -- "dist/$archive_name" "dist/$archive_name.sha256"
mkdir -p -- "$package_dir/lib" "$package_dir/scripts"
# The Sherpa Go module embeds an absolute build-machine RPATH. Add a portable
# fallback so the bundled native libraries resolve beside the extracted tools.
# shellcheck disable=SC2016 # $ORIGIN must remain literal for the ELF loader.
export CGO_LDFLAGS='-Wl,-rpath,$ORIGIN/lib'
go build -trimpath -o "$package_dir/captioneer" ./src/cmd/captioneer
go build -trimpath -tags gtk -o "$package_dir/captioneer-desktop" ./src/cmd/captioneer-desktop
sherpa_dir="$(go list -m -f '{{.Dir}}' github.com/k2-fsa/sherpa-onnx-go-linux)"
sherpa_lib_dir="$sherpa_dir/lib/x86_64-unknown-linux-gnu"
cp -- "$sherpa_lib_dir/libsherpa-onnx-c-api.so" "$package_dir/lib/"
cp -- "$sherpa_lib_dir/libonnxruntime.so" "$package_dir/lib/"
cp -- README.md LICENSE "$package_dir/"
cp -- scripts/download-models.sh "$package_dir/scripts/"
for executable in captioneer captioneer-desktop; do
# shellcheck disable=SC2016 # Verify the literal ELF-loader token.
if ! readelf -d "$package_dir/$executable" | grep -Fq '$ORIGIN/lib'; then
printf 'captioneer: %s is missing the portable native-library RPATH\n' "$executable" >&2
exit 1
fi
done
tar -C dist -czf "dist/$archive_name" "$package_name"
(
cd -- dist
sha256sum "$archive_name" > "$archive_name.sha256"
)
printf '%s\n' "$repo_root/dist/$archive_name"
printf '%s\n' "$repo_root/dist/$archive_name.sha256"
+80
View File
@@ -0,0 +1,80 @@
#!/usr/bin/env bash
set -Eeuo pipefail
: "${GITEA_API_URL:?GITEA_API_URL is required}"
: "${GITEA_REPOSITORY:?GITEA_REPOSITORY is required}"
: "${GITEA_TOKEN:?GITEA_TOKEN is required}"
: "${RELEASE_TAG:?RELEASE_TAG is required}"
: "${RELEASE_NOTES:?RELEASE_NOTES is required}"
if [[ ! "$RELEASE_TAG" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
printf 'captioneer: release tag must match vMAJOR.MINOR.BUG, got %q\n' "$RELEASE_TAG" >&2
exit 1
fi
if (( $# == 0 )); then
printf 'captioneer: at least one release asset is required\n' >&2
exit 1
fi
for asset in "$@"; do
[[ -f "$asset" ]] || { printf 'captioneer: release asset not found: %s\n' "$asset" >&2; exit 1; }
done
api_base="${GITEA_API_URL%/}/repos/$GITEA_REPOSITORY"
authorization="Authorization: token $GITEA_TOKEN"
release_response="$(mktemp)"
trap 'rm -f -- "$release_response"' EXIT
status="$(curl --silent --show-error --output "$release_response" --write-out '%{http_code}' \
--header "$authorization" \
"$api_base/releases/tags/$RELEASE_TAG")"
release_payload="$(jq -n \
--arg tag "$RELEASE_TAG" \
--arg name "Captioneer $RELEASE_TAG" \
--rawfile body "$RELEASE_NOTES" \
'{tag_name: $tag, name: $name, body: $body, draft: false, prerelease: false}')"
case "$status" in
200)
release_id="$(jq -er '.id' "$release_response")"
curl --fail-with-body --silent --show-error \
--request PATCH \
--header "$authorization" \
--header 'Content-Type: application/json' \
--data "$release_payload" \
"$api_base/releases/$release_id" > "$release_response"
;;
404)
curl --fail-with-body --silent --show-error \
--request POST \
--header "$authorization" \
--header 'Content-Type: application/json' \
--data "$release_payload" \
"$api_base/releases" > "$release_response"
release_id="$(jq -er '.id' "$release_response")"
;;
*)
printf 'captioneer: Gitea release lookup failed with HTTP %s\n' "$status" >&2
cat "$release_response" >&2
exit 1
;;
esac
for asset in "$@"; do
asset_name="$(basename -- "$asset")"
while IFS= read -r attachment_id; do
[[ -n "$attachment_id" ]] || continue
curl --fail-with-body --silent --show-error \
--request DELETE \
--header "$authorization" \
"$api_base/releases/$release_id/assets/$attachment_id" >/dev/null
done < <(jq -r --arg name "$asset_name" '.assets[]? | select(.name == $name) | .id' "$release_response")
curl --fail-with-body --silent --show-error \
--request POST \
--header "$authorization" \
--form "attachment=@$asset" \
"$api_base/releases/$release_id/assets?name=$asset_name" >/dev/null
printf 'Uploaded %s\n' "$asset_name"
done
+19
View File
@@ -0,0 +1,19 @@
#!/usr/bin/env bash
set -Eeuo pipefail
repo_root="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")/../.." && pwd -P)"
cd -- "$repo_root"
go test ./...
go test -tags gtk ./...
go vet ./...
go vet -tags gtk ./...
shellcheck \
scripts/download-models.sh \
scripts/distrobox/common.sh \
scripts/distrobox/build.sh \
scripts/distrobox/run.sh \
scripts/ci/test.sh \
scripts/ci/package-release.sh \
scripts/ci/publish-release.sh
+15 -2
View File
@@ -4,10 +4,15 @@ package config
import (
"errors"
"flag"
"strings"
"time"
)
const DefaultModelsDir = "models"
const (
DefaultModelsDir = "models"
DefaultOverlayFontFamily = "Sans"
MinimumOverlayFinalLifetime = 3 * time.Second
)
type Mode string
@@ -57,6 +62,7 @@ type OverlaySettings struct {
Backend OverlayBackend
Opacity float64
FontSize int
FontFamily string
BottomMargin int
MaxWidth int
Monitor string
@@ -85,10 +91,11 @@ func ParseDesktop() DesktopSettings {
flags.StringVar(&backend, "overlay-backend", "auto", "display backend: auto, wayland, or x11")
flags.Float64Var(&desktop.Overlay.Opacity, "overlay-opacity", 0.90, "caption background opacity from 0 to 1")
flags.IntVar(&desktop.Overlay.FontSize, "overlay-font-size", 28, "caption font size in logical pixels")
flags.StringVar(&desktop.Overlay.FontFamily, "overlay-font-family", DefaultOverlayFontFamily, "caption font family")
flags.IntVar(&desktop.Overlay.BottomMargin, "overlay-bottom-margin", 100, "distance from the monitor bottom in logical pixels")
flags.IntVar(&desktop.Overlay.MaxWidth, "overlay-max-width", 1200, "fixed caption width in logical pixels before the monitor safety cap")
flags.StringVar(&desktop.Overlay.Monitor, "overlay-monitor", "auto", "monitor selection: auto, numeric index, or connector name")
flags.DurationVar(&desktop.Overlay.FinalTimeout, "overlay-final-timeout", 4*time.Second, "how long a final caption remains; zero waits for replacement")
flags.DurationVar(&desktop.Overlay.FinalTimeout, "overlay-final-timeout", MinimumOverlayFinalLifetime, "completed-caption lifetime; zero keeps lines until the six-line cap removes them")
flags.BoolVar(&desktop.Overlay.ClickThrough, "overlay-click-through", true, "pass pointer input through the caption window")
flag.Parse()
desktop.Mode = Mode(mode)
@@ -126,6 +133,9 @@ func (s DesktopSettings) Validate() error {
if s.Overlay.FontSize <= 0 {
return errors.New("--overlay-font-size must be positive")
}
if strings.TrimSpace(s.Overlay.FontFamily) == "" {
return errors.New("--overlay-font-family cannot be empty")
}
if s.Overlay.BottomMargin < 0 {
return errors.New("--overlay-bottom-margin cannot be negative")
}
@@ -135,6 +145,9 @@ func (s DesktopSettings) Validate() error {
if s.Overlay.FinalTimeout < 0 {
return errors.New("--overlay-final-timeout cannot be negative")
}
if s.Overlay.FinalTimeout > 0 && s.Overlay.FinalTimeout < MinimumOverlayFinalLifetime {
return errors.New("--overlay-final-timeout must be zero or at least 3s")
}
if s.Overlay.Monitor == "" {
return errors.New("--overlay-monitor cannot be empty")
}
+3 -1
View File
@@ -20,7 +20,7 @@ func validDesktopSettings() DesktopSettings {
Settings: Settings{Mode: ModeVAD, ChunkDuration: time.Second, Threads: 2},
Output: OutputOverlay,
Overlay: OverlaySettings{
Backend: BackendAuto, Opacity: 0.9, FontSize: 28, BottomMargin: 100,
Backend: BackendAuto, Opacity: 0.9, FontSize: 28, FontFamily: DefaultOverlayFontFamily, BottomMargin: 100,
MaxWidth: 1200, Monitor: "auto", FinalTimeout: 4 * time.Second, ClickThrough: true,
},
}
@@ -40,10 +40,12 @@ func TestDesktopSettingsValidate(t *testing.T) {
{"negative opacity", func(s *DesktopSettings) { s.Overlay.Opacity = -0.1 }},
{"large opacity", func(s *DesktopSettings) { s.Overlay.Opacity = 1.1 }},
{"zero font", func(s *DesktopSettings) { s.Overlay.FontSize = 0 }},
{"empty font family", func(s *DesktopSettings) { s.Overlay.FontFamily = " " }},
{"negative margin", func(s *DesktopSettings) { s.Overlay.BottomMargin = -1 }},
{"zero width", func(s *DesktopSettings) { s.Overlay.MaxWidth = 0 }},
{"empty monitor", func(s *DesktopSettings) { s.Overlay.Monitor = "" }},
{"negative timeout", func(s *DesktopSettings) { s.Overlay.FinalTimeout = -time.Second }},
{"short timeout", func(s *DesktopSettings) { s.Overlay.FinalTimeout = time.Second }},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
+1
View File
@@ -132,6 +132,7 @@ func newVAD(model string, threads int) *sherpa.VoiceActivityDetector {
SileroVad: sherpa.SileroVadModelConfig{
Model: model,
Threshold: 0.5,
// A 500 ms pause ends the current caption line and starts the next one.
MinSilenceDuration: 0.5,
MinSpeechDuration: 0.25,
WindowSize: 512,
+91 -15
View File
@@ -215,7 +215,7 @@ static gboolean captioneer_apply_update(gpointer data) {
CaptioneerOverlay *overlay = update->overlay;
overlay->final_visible = update->final_text[0] != '\0';
gtk_label_set_text(GTK_LABEL(overlay->final_label), update->final_text);
gtk_label_set_markup(GTK_LABEL(overlay->final_label), update->final_text);
gtk_widget_set_visible(overlay->final_label, overlay->final_visible);
overlay->preview_visible = update->preview_text[0] != '\0';
@@ -246,10 +246,27 @@ static gboolean captioneer_window_close(GtkWindow *window, gpointer data) {
return TRUE;
}
static char *captioneer_font_family_rule(const char *font_family) {
GString *rule = g_string_new("font-family: \"");
for (const char *character = font_family; *character != '\0'; character++) {
if (*character == '\\' || *character == '\"') {
g_string_append_c(rule, '\\');
}
if (*character == '\n' || *character == '\r' || *character == '\t') {
g_string_append_c(rule, ' ');
} else {
g_string_append_c(rule, *character);
}
}
g_string_append(rule, "\";");
return g_string_free(rule, FALSE);
}
static CaptioneerOverlay *captioneer_overlay_new(
const char *backend,
double opacity,
int font_size,
const char *font_family,
int bottom_margin,
int max_width,
const char *monitor_selection,
@@ -354,28 +371,39 @@ static CaptioneerOverlay *captioneer_overlay_new(
gtk_label_set_xalign(GTK_LABEL(labels[i]), 0.0f);
gtk_label_set_wrap(GTK_LABEL(labels[i]), TRUE);
gtk_label_set_wrap_mode(GTK_LABEL(labels[i]), PANGO_WRAP_WORD_CHAR);
gtk_label_set_lines(GTK_LABEL(labels[i]), 2);
gtk_label_set_ellipsize(GTK_LABEL(labels[i]), PANGO_ELLIPSIZE_END);
// width-chars gives GTK a concrete allocation width, so wrapping happens
// before a long sentence can widen the bubble beyond the screen.
gtk_label_set_width_chars(GTK_LABEL(labels[i]), max_chars);
gtk_label_set_max_width_chars(GTK_LABEL(labels[i]), max_chars);
gtk_widget_set_halign(labels[i], GTK_ALIGN_FILL);
gtk_box_append(GTK_BOX(box), labels[i]);
gtk_widget_set_visible(labels[i], FALSE);
}
// Completed-caption history grows vertically for every visible entry. The
// view state, rather than GTK ellipsizing, enforces its six-entry limit.
gtk_label_set_lines(GTK_LABEL(overlay->final_label), -1);
gtk_label_set_ellipsize(GTK_LABEL(overlay->final_label), PANGO_ELLIPSIZE_NONE);
// Provisional text is already split and bounded by the Go view state. Let
// GTK allocate all of those lines instead of replacing overflow with "…".
gtk_label_set_lines(GTK_LABEL(overlay->preview_label), -1);
gtk_label_set_ellipsize(GTK_LABEL(overlay->preview_label), PANGO_ELLIPSIZE_NONE);
gtk_widget_add_css_class(overlay->final_label, "captioneer-final");
gtk_widget_add_css_class(overlay->preview_label, "captioneer-preview");
gtk_window_set_child(overlay->window, box);
char *font_family_rule = captioneer_font_family_rule(font_family);
char *css = g_strdup_printf(
"window.captioneer-overlay { background-color: transparent; box-shadow: none; }"
".captioneer-bubble { background-color: rgba(0, 0, 0, %.3f); border-radius: 12px; padding: 14px 22px; }"
".captioneer-final { color: rgba(255, 255, 255, 1.0); font-size: %dpx; }"
".captioneer-preview { color: rgba(255, 255, 255, 0.78); font-size: %dpx; }",
opacity, font_size, font_size);
".captioneer-final { color: rgba(255, 255, 255, 1.0); font-size: %dpx; %s }"
".captioneer-preview { color: rgba(255, 255, 255, 0.78); font-size: %dpx; %s }",
opacity, font_size, font_family_rule, font_size, font_family_rule);
GtkCssProvider *provider = gtk_css_provider_new();
gtk_css_provider_load_from_string(provider, css);
gtk_style_context_add_provider_for_display(display, GTK_STYLE_PROVIDER(provider), GTK_STYLE_PROVIDER_PRIORITY_APPLICATION);
g_object_unref(provider);
g_free(css);
g_free(font_family_rule);
gtk_widget_realize(GTK_WIDGET(overlay->window));
GdkSurface *surface = gtk_native_get_surface(GTK_NATIVE(overlay->window));
@@ -445,10 +473,18 @@ import (
"tea.chunkbyte.com/kato/captioneer/src/config"
)
const (
finalFadeRefreshInterval = 200 * time.Millisecond
averageGlyphWidthFactor = 0.58
minimumLineCharacters = 20
maximumLineCharacters = 64
)
type Sink struct {
native *C.CaptioneerOverlay
finalTimeout time.Duration
finalTimer *time.Timer
lineCharacters int
state viewState
mu sync.RWMutex
closed bool
@@ -456,8 +492,10 @@ type Sink struct {
func New(settings config.OverlaySettings) (*Sink, string, error) {
backend := C.CString(string(settings.Backend))
fontFamily := C.CString(settings.FontFamily)
monitor := C.CString(settings.Monitor)
defer C.free(unsafe.Pointer(backend))
defer C.free(unsafe.Pointer(fontFamily))
defer C.free(unsafe.Pointer(monitor))
var warningMessage *C.char
@@ -466,6 +504,7 @@ func New(settings config.OverlaySettings) (*Sink, string, error) {
backend,
C.double(settings.Opacity),
C.int(settings.FontSize),
fontFamily,
C.int(settings.BottomMargin),
C.int(settings.MaxWidth),
monitor,
@@ -481,7 +520,11 @@ func New(settings config.OverlaySettings) (*Sink, string, error) {
}
return nil, warning, errors.New(message)
}
return &Sink{native: native, finalTimeout: settings.FinalTimeout}, warning, nil
return &Sink{
native: native,
finalTimeout: settings.FinalTimeout,
lineCharacters: overlayLineCharacters(settings),
}, warning, nil
}
func (s *Sink) Publish(event captions.Event) error {
@@ -490,29 +533,42 @@ func (s *Sink) Publish(event captions.Event) error {
if s.closed {
return errors.New("overlay is closed")
}
s.state.Apply(event, time.Now(), s.finalTimeout)
if event.Kind == captions.Final || event.Kind == captions.Hide {
now := time.Now()
s.state.Apply(event, now, s.finalTimeout, s.lineCharacters)
if event.Kind == captions.Final {
s.stopFinalTimer()
}
if event.Kind == captions.Final && s.finalTimeout > 0 {
s.finalTimer = time.AfterFunc(s.finalTimeout, s.expireFinal)
s.scheduleFinalRefresh(now)
}
s.postState()
return nil
}
func (s *Sink) expireFinal() {
func overlayLineCharacters(settings config.OverlaySettings) int {
characters := int(float64(settings.MaxWidth) / (float64(settings.FontSize) * averageGlyphWidthFactor))
if characters < minimumLineCharacters {
return minimumLineCharacters
}
if characters > maximumLineCharacters {
return maximumLineCharacters
}
return characters
}
func (s *Sink) refreshFinals() {
s.mu.Lock()
defer s.mu.Unlock()
if s.closed || !s.state.ExpireFinal(time.Now()) {
if s.closed {
return
}
now := time.Now()
s.state.ExpireFinal(now)
s.finalTimer = nil
s.scheduleFinalRefresh(now)
s.postState()
}
func (s *Sink) postState() {
finalText := C.CString(s.state.Final)
finalText := C.CString(s.state.FinalMarkup(time.Now(), s.finalTimeout))
previewText := C.CString(s.state.Provisional)
defer C.free(unsafe.Pointer(finalText))
defer C.free(unsafe.Pointer(previewText))
@@ -526,6 +582,26 @@ func (s *Sink) stopFinalTimer() {
}
}
func (s *Sink) scheduleFinalRefresh(now time.Time) {
if len(s.state.Finals) == 0 {
return
}
if s.state.NextFinalExpiry().IsZero() && !s.state.NeedsFadeRefresh(now, s.finalTimeout) {
return
}
nextRefresh := now.Add(finalFadeRefreshInterval)
expiresAt := s.state.NextFinalExpiry()
if !expiresAt.IsZero() && expiresAt.Before(nextRefresh) {
nextRefresh = expiresAt
}
delay := nextRefresh.Sub(now)
if delay < 0 {
delay = 0
}
s.finalTimer = time.AfterFunc(delay, s.refreshFinals)
}
func (s *Sink) Run() error {
s.mu.RLock()
if s.closed {
+204 -16
View File
@@ -1,39 +1,227 @@
package overlay
import (
"fmt"
"html"
"math"
"strings"
"time"
"unicode/utf8"
"tea.chunkbyte.com/kato/captioneer/src/captions"
)
type viewState struct {
Final string
Provisional string
FinalExpiresAt time.Time
const (
maxVisibleFinalLines = 6
maxVisiblePreviewLines = 3
oldestLineBrightness = 0.65
defaultLineFadeDuration = 3 * time.Second
readingWordsPerSecond = 3.0
)
type finalLine struct {
CreatedAt time.Time
Text string
ExpiresAt time.Time
}
func (s *viewState) Apply(event captions.Event, now time.Time, finalTimeout time.Duration) {
type viewState struct {
Finals []finalLine
Provisional string
}
func (s *viewState) Apply(event captions.Event, now time.Time, finalTimeout time.Duration, maxLineCharacters int) {
switch event.Kind {
case captions.Provisional:
s.Provisional = strings.TrimSpace(event.Text)
case captions.Final:
s.Final = strings.TrimSpace(event.Text)
s.Provisional = ""
s.FinalExpiresAt = time.Time{}
if finalTimeout > 0 {
s.FinalExpiresAt = now.Add(finalTimeout)
chunks := splitCaptionText(event.Text, maxLineCharacters)
if len(chunks) > maxVisiblePreviewLines {
chunks = chunks[len(chunks)-maxVisiblePreviewLines:]
}
s.Provisional = strings.Join(chunks, "\n")
case captions.Final:
chunks := splitCaptionText(event.Text, maxLineCharacters)
if len(chunks) == 0 {
return
}
if len(chunks) > maxVisibleFinalLines {
chunks = chunks[len(chunks)-maxVisibleFinalLines:]
}
expiresAt := now
for _, text := range chunks {
line := finalLine{CreatedAt: now, Text: text}
if finalTimeout > 0 {
expiresAt = expiresAt.Add(readingLifetime(text, finalTimeout))
line.ExpiresAt = expiresAt
}
s.Finals = append(s.Finals, line)
}
if len(s.Finals) > maxVisibleFinalLines {
s.Finals = s.Finals[len(s.Finals)-maxVisibleFinalLines:]
}
s.Provisional = ""
case captions.Hide:
*s = viewState{}
// A blank recognition result must not erase completed captions before
// their individual reading lifetimes have elapsed.
s.Provisional = ""
}
}
func (s *viewState) ExpireFinal(now time.Time) bool {
if s.FinalExpiresAt.IsZero() || now.Before(s.FinalExpiresAt) {
firstVisible := 0
for firstVisible < len(s.Finals) {
expiresAt := s.Finals[firstVisible].ExpiresAt
if expiresAt.IsZero() || now.Before(expiresAt) {
break
}
firstVisible++
}
if firstVisible == 0 {
return false
}
s.Final = ""
s.FinalExpiresAt = time.Time{}
s.Finals = s.Finals[firstVisible:]
return true
}
func (s viewState) FinalText() string {
lines := make([]string, len(s.Finals))
for i, line := range s.Finals {
lines[i] = line.Text
}
return strings.Join(lines, "\n")
}
func (s viewState) FinalMarkup(now time.Time, lineLifetime time.Duration) string {
fadeDuration := lineLifetime
if fadeDuration <= 0 {
fadeDuration = defaultLineFadeDuration
}
var markup strings.Builder
for i, line := range s.Finals {
if i > 0 {
markup.WriteByte('\n')
}
lineFadeDuration := fadeDuration
if !line.ExpiresAt.IsZero() {
lineFadeDuration = line.ExpiresAt.Sub(line.CreatedAt)
}
brightness := lineBrightness(now.Sub(line.CreatedAt), lineFadeDuration)
channel := int(brightness*255 + 0.5)
fmt.Fprintf(&markup, `<span foreground="#%02X%02X%02X">`, channel, channel, channel)
markup.WriteString(html.EscapeString(line.Text))
markup.WriteString("</span>")
}
return markup.String()
}
func (s viewState) NeedsFadeRefresh(now time.Time, lineLifetime time.Duration) bool {
fadeDuration := lineLifetime
if fadeDuration <= 0 {
fadeDuration = defaultLineFadeDuration
}
for _, line := range s.Finals {
lineFadeDuration := fadeDuration
if !line.ExpiresAt.IsZero() {
lineFadeDuration = line.ExpiresAt.Sub(line.CreatedAt)
}
if now.Sub(line.CreatedAt) < lineFadeDuration {
return true
}
}
return false
}
func splitCaptionText(text string, maxCharacters int) []string {
if maxCharacters < 1 {
maxCharacters = 1
}
words := strings.Fields(text)
if len(words) == 0 {
return nil
}
var chunks []string
var current strings.Builder
currentCharacters := 0
flush := func() {
if currentCharacters == 0 {
return
}
chunks = append(chunks, current.String())
current.Reset()
currentCharacters = 0
}
for _, word := range words {
wordCharacters := utf8.RuneCountInString(word)
if wordCharacters > maxCharacters {
flush()
runes := []rune(word)
for len(runes) > maxCharacters {
chunks = append(chunks, string(runes[:maxCharacters]))
runes = runes[maxCharacters:]
}
if len(runes) > 0 {
current.WriteString(string(runes))
currentCharacters = len(runes)
}
} else {
separatorCharacters := 0
if currentCharacters > 0 {
separatorCharacters = 1
}
if currentCharacters+separatorCharacters+wordCharacters > maxCharacters {
flush()
separatorCharacters = 0
}
if separatorCharacters > 0 {
current.WriteByte(' ')
currentCharacters++
}
current.WriteString(word)
currentCharacters += wordCharacters
}
if endsSentence(word) {
flush()
}
}
flush()
return chunks
}
func endsSentence(word string) bool {
trimmed := strings.TrimRight(word, `"'”’)]}`)
return strings.HasSuffix(trimmed, ".") || strings.HasSuffix(trimmed, "!") || strings.HasSuffix(trimmed, "?")
}
func readingLifetime(text string, minimum time.Duration) time.Duration {
if minimum <= 0 {
return 0
}
words := len(strings.Fields(text))
readingSeconds := math.Ceil(float64(words) / readingWordsPerSecond)
readingTime := time.Duration(readingSeconds * float64(time.Second))
if readingTime < minimum {
return minimum
}
return readingTime
}
func (s viewState) NextFinalExpiry() time.Time {
if len(s.Finals) == 0 {
return time.Time{}
}
return s.Finals[0].ExpiresAt
}
func lineBrightness(age, fadeDuration time.Duration) float64 {
if age <= 0 || fadeDuration <= 0 {
return 1
}
progress := float64(age) / float64(fadeDuration)
if progress > 1 {
progress = 1
}
return 1 - progress*(1-oldestLineBrightness)
}
+96 -17
View File
@@ -7,44 +7,123 @@ import (
"tea.chunkbyte.com/kato/captioneer/src/captions"
)
func TestViewStateKeepsFinalAboveNewProvisional(t *testing.T) {
const testLineCharacters = 80
func TestViewStateKeepsCompletedLinesAboveNewProvisional(t *testing.T) {
now := time.Unix(100, 0)
var state viewState
state.Apply(captions.Event{Kind: captions.Provisional, Text: " first draft "}, now, 4*time.Second)
state.Apply(captions.Event{Kind: captions.Final, Text: " final "}, now, 4*time.Second)
state.Apply(captions.Event{Kind: captions.Provisional, Text: " next draft "}, now, 4*time.Second)
state.Apply(captions.Event{Kind: captions.Provisional, Text: " first draft "}, now, 4*time.Second, testLineCharacters)
state.Apply(captions.Event{Kind: captions.Final, Text: " first final "}, now, 4*time.Second, testLineCharacters)
state.Apply(captions.Event{Kind: captions.Final, Text: " second final "}, now, 4*time.Second, testLineCharacters)
state.Apply(captions.Event{Kind: captions.Provisional, Text: " next draft "}, now, 4*time.Second, testLineCharacters)
if state.Final != "final" || state.Provisional != "next draft" {
if state.FinalText() != "first final\nsecond final" || state.Provisional != "next draft" {
t.Fatalf("unexpected rows: %#v", state)
}
state.Apply(captions.Event{Kind: captions.Hide}, now, 4*time.Second)
if state.Final != "" || state.Provisional != "" {
t.Fatalf("hide did not clear rows: %#v", state)
state.Apply(captions.Event{Kind: captions.Hide}, now, 4*time.Second, testLineCharacters)
if state.FinalText() != "first final\nsecond final" || state.Provisional != "" {
t.Fatalf("blank result erased completed captions: %#v", state)
}
}
func TestViewStateFinalTimeout(t *testing.T) {
func TestViewStateExpiresEachCompletedLineAfterItsOwnLifetime(t *testing.T) {
now := time.Unix(100, 0)
var state viewState
state.Apply(captions.Event{Kind: captions.Final, Text: "final"}, now, 4*time.Second)
state.Apply(captions.Event{Kind: captions.Provisional, Text: "draft"}, now, 4*time.Second)
state.Apply(captions.Event{Kind: captions.Final, Text: "first"}, now, 3*time.Second, testLineCharacters)
state.Apply(captions.Event{Kind: captions.Final, Text: "second"}, now.Add(time.Second), 3*time.Second, testLineCharacters)
state.Apply(captions.Event{Kind: captions.Provisional, Text: "draft"}, now, 3*time.Second, testLineCharacters)
if state.ExpireFinal(now.Add(3999 * time.Millisecond)) {
t.Fatal("final caption expired early")
if state.ExpireFinal(now.Add(2999 * time.Millisecond)) {
t.Fatal("completed line expired early")
}
if !state.ExpireFinal(now.Add(4 * time.Second)) {
t.Fatal("final caption did not expire")
if !state.ExpireFinal(now.Add(3 * time.Second)) {
t.Fatal("first completed line did not expire")
}
if state.Final != "" || state.Provisional != "draft" {
if state.FinalText() != "second" || state.Provisional != "draft" {
t.Fatalf("expiry changed the wrong row: %#v", state)
}
}
func TestViewStateLimitsVisibleCompletedLines(t *testing.T) {
now := time.Unix(100, 0)
var state viewState
for i := 0; i <= maxVisibleFinalLines; i++ {
state.Apply(captions.Event{Kind: captions.Final, Text: string(rune('a' + i))}, now, 3*time.Second, testLineCharacters)
}
if len(state.Finals) != maxVisibleFinalLines || state.FinalText() != "b\nc\nd\ne\nf\ng" {
t.Fatalf("unexpected completed-line limit: %#v", state)
}
}
func TestViewStateZeroTimeoutDoesNotExpire(t *testing.T) {
now := time.Unix(100, 0)
var state viewState
state.Apply(captions.Event{Kind: captions.Final, Text: "final"}, now, 0)
state.Apply(captions.Event{Kind: captions.Final, Text: "final"}, now, 0, testLineCharacters)
if state.ExpireFinal(now.Add(24 * time.Hour)) {
t.Fatal("zero-timeout caption expired")
}
}
func TestViewStateFinalMarkupFadesAndEscapesText(t *testing.T) {
now := time.Unix(100, 0)
var state viewState
state.Apply(captions.Event{Kind: captions.Final, Text: "<hello & goodbye>"}, now, 3*time.Second, testLineCharacters)
if got := state.FinalMarkup(now, 3*time.Second); got != `<span foreground="#FFFFFF">&lt;hello &amp; goodbye&gt;</span>` {
t.Fatalf("new line markup = %q", got)
}
if got := state.FinalMarkup(now.Add(3*time.Second), 3*time.Second); got != `<span foreground="#A6A6A6">&lt;hello &amp; goodbye&gt;</span>` {
t.Fatalf("old line markup = %q", got)
}
}
func TestSplitCaptionTextUsesSentencesAndWordBoundaries(t *testing.T) {
got := splitCaptionText("Hello there. This sentence should wrap cleanly at words.", 18)
want := []string{"Hello there.", "This sentence", "should wrap", "cleanly at words."}
if len(got) != len(want) {
t.Fatalf("chunks = %#v, want %#v", got, want)
}
for i := range want {
if got[i] != want[i] {
t.Fatalf("chunks = %#v, want %#v", got, want)
}
}
}
func TestProvisionalCaptionKeepsOnlyLatestDigestibleChunks(t *testing.T) {
now := time.Unix(100, 0)
var state viewState
state.Apply(captions.Event{Kind: captions.Provisional, Text: "One. Two. Three. Four."}, now, 3*time.Second, testLineCharacters)
if state.Provisional != "Two.\nThree.\nFour." {
t.Fatalf("provisional text = %q", state.Provisional)
}
}
func TestReadingLifetimeGrowsWithWordCount(t *testing.T) {
text := "one two three four five six seven eight nine ten eleven twelve"
if got := readingLifetime(text, 3*time.Second); got != 4*time.Second {
t.Fatalf("reading lifetime = %s, want 4s", got)
}
if got := readingLifetime("short caption", 3*time.Second); got != 3*time.Second {
t.Fatalf("short reading lifetime = %s, want 3s", got)
}
}
func TestFinalizedChunksExpireInReadingOrder(t *testing.T) {
now := time.Unix(100, 0)
var state viewState
state.Apply(captions.Event{Kind: captions.Final, Text: "First sentence. Second sentence."}, now, 3*time.Second, testLineCharacters)
if len(state.Finals) != 2 {
t.Fatalf("final chunks = %#v", state.Finals)
}
if state.Finals[0].ExpiresAt != now.Add(3*time.Second) {
t.Fatalf("first expiry = %s", state.Finals[0].ExpiresAt)
}
if state.Finals[1].ExpiresAt != now.Add(6*time.Second) {
t.Fatalf("second expiry = %s", state.Finals[1].ExpiresAt)
}
if !state.ExpireFinal(now.Add(3*time.Second)) || state.FinalText() != "Second sentence." {
t.Fatalf("unexpected state after first reading window: %#v", state)
}
}