Implement text wrapping element around the textwrap crate

5 files changed, 625 insertions, 0 deletions

diff --git a/text/wrap/Cargo.toml b/text/wrap/Cargo.toml
new file mode 100644
index 000000000..7b25b6112
--- /dev/null
+++ b/text/wrap/Cargo.toml
@@ -0,0 +1,31 @@
+[package]
+name = "gst-plugin-textwrap"
+version = "0.6.0"
+authors = ["Mathieu Duponchelle <mathieu@centricular.com>"]
+license = "LGPL-2.1-or-later"
+edition = "2018"
+description = "Rust Text Wrap Plugin"
+repository = "https://gitlab.freedesktop.org/gstreamer/gst-plugins-rs"
+
+[dependencies]
+glib = { git = "https://github.com/gtk-rs/glib" }
+once_cell = "1.0"
+textwrap = { version = "0.11", features = ["hyphenation"] }
+hyphenation = "0.7.1"
+
+[dependencies.gst]
+git = "https://gitlab.freedesktop.org/gstreamer/gstreamer-rs"
+features = ["v1_12"]
+package="gstreamer"
+
+[lib]
+name = "gstrstextwrap"
+crate-type = ["cdylib", "rlib"]
+path = "src/lib.rs"
+
+[build-dependencies]
+gst-plugin-version-helper = { path="../../version-helper" }
+
+[dev-dependencies.gst-check]
+git = "https://gitlab.freedesktop.org/gstreamer/gstreamer-rs"
+package="gstreamer-check"
diff --git a/text/wrap/build.rs b/text/wrap/build.rs
new file mode 100644
index 000000000..fe307a434
--- /dev/null
+++ b/text/wrap/build.rs
@@ -0,0 +1,5 @@
+use gst_plugin_version_helper;
+
+fn main() {
+    gst_plugin_version_helper::get_info()
+}
diff --git a/text/wrap/src/gsttextwrap.rs b/text/wrap/src/gsttextwrap.rs
new file mode 100644
index 000000000..7644259d5
--- /dev/null
+++ b/text/wrap/src/gsttextwrap.rs
@@ -0,0 +1,423 @@
+// Copyright (C) 2020 Mathieu Duponchelle <mathieu@centricular.com>
+//
+// This library is free software; you can redistribute it and/or
+// modify it under the terms of the GNU Library General Public
+// License as published by the Free Software Foundation; either
+// version 2 of the License, or (at your option) any later version.
+//
+// This library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+// Library General Public License for more details.
+//
+// You should have received a copy of the GNU Library General Public
+// License along with this library; if not, write to the
+// Free Software Foundation, Inc., 51 Franklin Street, Suite 500,
+// Boston, MA 02110-1335, USA.
+
+use glib;
+use glib::prelude::*;
+use glib::subclass;
+use glib::subclass::prelude::*;
+use gst;
+use gst::prelude::*;
+use gst::subclass::prelude::*;
+
+use std::default::Default;
+use std::fs::File;
+use std::io;
+use std::sync::Mutex;
+
+use once_cell::sync::Lazy;
+
+use hyphenation::{Load, Standard};
+use textwrap;
+
+static CAT: Lazy<gst::DebugCategory> = Lazy::new(|| {
+    gst::DebugCategory::new(
+        "textwrap",
+        gst::DebugColorFlags::empty(),
+        Some("Text wrapper element"),
+    )
+});
+
+const DEFAULT_DICTIONARY: Option<String> = None;
+const DEFAULT_COLUMNS: u32 = 32; /* CEA 608 max columns */
+const DEFAULT_LINES: u32 = 0;
+
+static PROPERTIES: [subclass::Property; 3] = [
+    subclass::Property("dictionary", |name| {
+        glib::ParamSpec::string(
+            name,
+            "Dictionary",
+            "Path to a dictionary to load at runtime to perform hyphenation, see \
+                <https://docs.rs/crate/hyphenation/0.7.1> for more information",
+            None,
+            glib::ParamFlags::READWRITE,
+        )
+    }),
+    subclass::Property("columns", |name| {
+        glib::ParamSpec::uint(
+            name,
+            "Columns",
+            "Maximum number of columns for any given line",
+            1,
+            std::u32::MAX,
+            DEFAULT_COLUMNS,
+            glib::ParamFlags::READWRITE,
+        )
+    }),
+    subclass::Property("lines", |name| {
+        glib::ParamSpec::uint(
+            name,
+            "Lines",
+            "Split input buffer into output buffers with max lines (0=do not split)",
+            0,
+            std::u32::MAX,
+            DEFAULT_LINES,
+            glib::ParamFlags::READWRITE,
+        )
+    }),
+];
+
+#[derive(Debug, Clone)]
+struct Settings {
+    dictionary: Option<String>,
+    columns: u32,
+    lines: u32,
+}
+
+impl Default for Settings {
+    fn default() -> Self {
+        Self {
+            dictionary: DEFAULT_DICTIONARY,
+            columns: DEFAULT_COLUMNS, /* CEA 608 max columns */
+            lines: DEFAULT_LINES,
+        }
+    }
+}
+
+#[allow(clippy::large_enum_variant)]
+enum Wrapper {
+    H(textwrap::Wrapper<'static, Standard>),
+    N(textwrap::Wrapper<'static, textwrap::NoHyphenation>),
+}
+
+struct State {
+    wrapper: Option<Wrapper>,
+}
+
+impl Wrapper {
+    fn fill(&self, s: &str) -> String {
+        match *self {
+            Wrapper::H(ref w) => w.fill(s),
+            Wrapper::N(ref w) => w.fill(s),
+        }
+    }
+}
+
+impl Default for State {
+    fn default() -> Self {
+        Self { wrapper: None }
+    }
+}
+
+struct TextWrap {
+    srcpad: gst::Pad,
+    sinkpad: gst::Pad,
+    settings: Mutex<Settings>,
+    state: Mutex<State>,
+}
+
+impl TextWrap {
+    fn set_pad_functions(sinkpad: &gst::Pad) {
+        sinkpad.set_chain_function(|pad, parent, buffer| {
+            TextWrap::catch_panic_pad_function(
+                parent,
+                || Err(gst::FlowError::Error),
+                |textwrap, element| textwrap.sink_chain(pad, element, buffer),
+            )
+        });
+    }
+
+    fn update_wrapper(&self, element: &gst::Element) {
+        let settings = self.settings.lock().unwrap();
+        let mut state = self.state.lock().unwrap();
+
+        if state.wrapper.is_some() {
+            return;
+        }
+
+        state.wrapper = if let Some(dictionary) = &settings.dictionary {
+            let dict_file = match File::open(dictionary) {
+                Err(err) => {
+                    gst_error!(CAT, obj: element, "Failed to open dictionary file: {}", err);
+                    return;
+                }
+                Ok(dict_file) => dict_file,
+            };
+
+            let mut reader = io::BufReader::new(dict_file);
+            let standard = match Standard::any_from_reader(&mut reader) {
+                Err(err) => {
+                    gst_error!(
+                        CAT,
+                        obj: element,
+                        "Failed to load standard from file: {}",
+                        err
+                    );
+                    return;
+                }
+                Ok(standard) => standard,
+            };
+
+            Some(Wrapper::H(textwrap::Wrapper::with_splitter(
+                settings.columns as usize,
+                standard,
+            )))
+        } else {
+            Some(Wrapper::N(textwrap::Wrapper::with_splitter(
+                settings.columns as usize,
+                textwrap::NoHyphenation,
+            )))
+        };
+    }
+
+    fn sink_chain(
+        &self,
+        _pad: &gst::Pad,
+        element: &gst::Element,
+        buffer: gst::Buffer,
+    ) -> Result<gst::FlowSuccess, gst::FlowError> {
+        self.update_wrapper(element);
+
+        let mut pts: gst::ClockTime = buffer
+            .get_pts()
+            .ok_or_else(|| {
+                gst_error!(CAT, obj: element, "Need timestamped buffers");
+                gst::FlowError::Error
+            })?
+            .into();
+
+        let duration: gst::ClockTime = buffer
+            .get_duration()
+            .ok_or_else(|| {
+                gst_error!(CAT, obj: element, "Need buffers with duration");
+                gst::FlowError::Error
+            })?
+            .into();
+
+        let data = buffer.map_readable().map_err(|_| {
+            gst_error!(CAT, obj: element, "Can't map buffer readable");
+
+            gst::FlowError::Error
+        })?;
+
+        let data = std::str::from_utf8(&data).map_err(|err| {
+            gst_error!(CAT, obj: element, "Can't decode utf8: {}", err);
+
+            gst::FlowError::Error
+        })?;
+
+        let lines = self.settings.lock().unwrap().lines;
+
+        let data = {
+            let state = self.state.lock().unwrap();
+            let wrapper = state
+                .wrapper
+                .as_ref()
+                .expect("We should have a wrapper by now");
+            wrapper.fill(data)
+        };
+
+        // If the lines property was set, we want to split the result into buffers
+        // of at most N lines. We compute the duration for each of those based on
+        // the total number of words, and the number of words in each of the split-up
+        // buffers.
+        if lines > 0 {
+            let mut bufferlist = gst::BufferList::new();
+            let duration_per_word: gst::ClockTime =
+                duration / data.split_whitespace().count() as u64;
+
+            for chunk in data.lines().collect::<Vec<&str>>().chunks(lines as usize) {
+                let data = chunk.join("\n");
+                let duration: gst::ClockTime =
+                    duration_per_word * data.split_whitespace().count() as u64;
+                let mut buf = gst::Buffer::from_mut_slice(data.into_bytes());
+
+                {
+                    let buf = buf.get_mut().unwrap();
+
+                    buf.set_pts(pts);
+                    buf.set_duration(duration);
+                    pts += duration;
+                }
+
+                bufferlist.get_mut().unwrap().add(buf);
+            }
+
+            self.srcpad.push_list(bufferlist)
+        } else {
+            let mut buf = gst::Buffer::from_mut_slice(data.into_bytes());
+
+            {
+                let buf = buf.get_mut().unwrap();
+
+                buf.set_pts(pts);
+                buf.set_duration(duration);
+            }
+
+            self.srcpad.push(buf)
+        }
+    }
+}
+
+impl ObjectSubclass for TextWrap {
+    const NAME: &'static str = "RsTextWrap";
+    type ParentType = gst::Element;
+    type Instance = gst::subclass::ElementInstanceStruct<Self>;
+    type Class = subclass::simple::ClassStruct<Self>;
+
+    glib_object_subclass!();
+
+    fn new_with_class(klass: &subclass::simple::ClassStruct<Self>) -> Self {
+        let templ = klass.get_pad_template("sink").unwrap();
+        let sinkpad = gst::Pad::new_from_template(&templ, Some("sink"));
+        let templ = klass.get_pad_template("src").unwrap();
+        let srcpad = gst::Pad::new_from_template(&templ, Some("src"));
+
+        srcpad.use_fixed_caps();
+        sinkpad.use_fixed_caps();
+
+        TextWrap::set_pad_functions(&sinkpad);
+
+        let settings = Mutex::new(Settings::default());
+        let state = Mutex::new(State::default());
+
+        Self {
+            srcpad,
+            sinkpad,
+            settings,
+            state,
+        }
+    }
+
+    fn class_init(klass: &mut subclass::simple::ClassStruct<Self>) {
+        klass.set_metadata(
+            "Text Wrapper",
+            "Text/Filter",
+            "Breaks text into fixed-size lines, with optional hyphenationz",
+            "Mathieu Duponchelle <mathieu@centricular.com>",
+        );
+
+        let caps = gst::Caps::builder("text/x-raw")
+            .field("format", &"utf8")
+            .build();
+        let src_pad_template = gst::PadTemplate::new(
+            "src",
+            gst::PadDirection::Src,
+            gst::PadPresence::Always,
+            &caps,
+        )
+        .unwrap();
+        klass.add_pad_template(src_pad_template);
+
+        let sink_pad_template = gst::PadTemplate::new(
+            "sink",
+            gst::PadDirection::Sink,
+            gst::PadPresence::Always,
+            &caps,
+        )
+        .unwrap();
+        klass.add_pad_template(sink_pad_template);
+
+        klass.install_properties(&PROPERTIES);
+    }
+}
+
+impl ObjectImpl for TextWrap {
+    glib_object_impl!();
+
+    fn constructed(&self, obj: &glib::Object) {
+        self.parent_constructed(obj);
+
+        let element = obj.downcast_ref::<gst::Element>().unwrap();
+        element.add_pad(&self.sinkpad).unwrap();
+        element.add_pad(&self.srcpad).unwrap();
+    }
+
+    fn set_property(&self, _obj: &glib::Object, id: usize, value: &glib::Value) {
+        let prop = &PROPERTIES[id];
+
+        match *prop {
+            subclass::Property("dictionary", ..) => {
+                let mut settings = self.settings.lock().unwrap();
+                let mut state = self.state.lock().unwrap();
+                settings.dictionary = value.get().expect("type checked upstream");
+                state.wrapper = None;
+            }
+            subclass::Property("columns", ..) => {
+                let mut settings = self.settings.lock().unwrap();
+                let mut state = self.state.lock().unwrap();
+                settings.columns = value.get_some().expect("type checked upstream");
+                state.wrapper = None;
+            }
+            subclass::Property("lines", ..) => {
+                let mut settings = self.settings.lock().unwrap();
+                settings.lines = value.get_some().expect("type checked upstream");
+            }
+            _ => unimplemented!(),
+        }
+    }
+
+    fn get_property(&self, _obj: &glib::Object, id: usize) -> Result<glib::Value, ()> {
+        let prop = &PROPERTIES[id];
+
+        match *prop {
+            subclass::Property("dictionary", ..) => {
+                let settings = self.settings.lock().unwrap();
+                Ok(settings.dictionary.to_value())
+            }
+            subclass::Property("columns", ..) => {
+                let settings = self.settings.lock().unwrap();
+                Ok(settings.columns.to_value())
+            }
+            subclass::Property("lines", ..) => {
+                let settings = self.settings.lock().unwrap();
+                Ok(settings.lines.to_value())
+            }
+            _ => unimplemented!(),
+        }
+    }
+}
+
+impl ElementImpl for TextWrap {
+    fn change_state(
+        &self,
+        element: &gst::Element,
+        transition: gst::StateChange,
+    ) -> Result<gst::StateChangeSuccess, gst::StateChangeError> {
+        gst_info!(CAT, obj: element, "Changing state {:?}", transition);
+
+        match transition {
+            gst::StateChange::PausedToReady => {
+                let mut state = self.state.lock().unwrap();
+                *state = State::default();
+            }
+            _ => (),
+        }
+
+        let success = self.parent_change_state(element, transition)?;
+
+        Ok(success)
+    }
+}
+
+pub fn register(plugin: &gst::Plugin) -> Result<(), glib::BoolError> {
+    gst::Element::register(
+        Some(plugin),
+        "textwrap",
+        gst::Rank::None,
+        TextWrap::get_type(),
+    )
+}
diff --git a/text/wrap/src/lib.rs b/text/wrap/src/lib.rs
new file mode 100644
index 000000000..c8d09eb98
--- /dev/null
+++ b/text/wrap/src/lib.rs
@@ -0,0 +1,43 @@
+// Copyright (C) 2020 Mathieu Duponchelle <mathieu@centricular.com>
+//
+// This library is free software; you can redistribute it and/or
+// modify it under the terms of the GNU Library General Public
+// License as published by the Free Software Foundation; either
+// version 2 of the License, or (at your option) any later version.
+//
+// This library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+// Library General Public License for more details.
+//
+// You should have received a copy of the GNU Library General Public
+// License along with this library; if not, write to the
+// Free Software Foundation, Inc., 51 Franklin Street, Suite 500,
+// Boston, MA 02110-1335, USA.
+
+#![recursion_limit = "128"]
+
+#[macro_use]
+extern crate glib;
+#[macro_use]
+extern crate gst;
+extern crate hyphenation;
+extern crate textwrap;
+
+mod gsttextwrap;
+
+fn plugin_init(plugin: &gst::Plugin) -> Result<(), glib::BoolError> {
+    gsttextwrap::register(plugin)
+}
+
+gst_plugin_define!(
+    rstextwrap,
+    env!("CARGO_PKG_DESCRIPTION"),
+    plugin_init,
+    concat!(env!("CARGO_PKG_VERSION"), "-", env!("COMMIT_ID")),
+    "LGPL",
+    env!("CARGO_PKG_NAME"),
+    env!("CARGO_PKG_NAME"),
+    env!("CARGO_PKG_REPOSITORY"),
+    env!("BUILD_REL_DATE")
+);
diff --git a/text/wrap/tests/textwrap.rs b/text/wrap/tests/textwrap.rs
new file mode 100644
index 000000000..7c194f857
--- /dev/null
+++ b/text/wrap/tests/textwrap.rs
@@ -0,0 +1,123 @@
+// Copyright (C) 2020 Mathieu Duponchelle <mathieu@centricular.com>
+//
+// This library is free software; you can redistribute it and/or
+// modify it under the terms of the GNU Library General Public
+// License as published by the Free Software Foundation; either
+// version 2 of the License, or (at your option) any later version.
+//
+// This library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+// Library General Public License for more details.
+//
+// You should have received a copy of the GNU Library General Public
+// License along with this library; if not, write to the
+// Free Software Foundation, Inc., 51 Franklin Street, Suite 500,
+// Boston, MA 02110-1335, USA.
+
+use glib::prelude::*;
+
+fn init() {
+    use std::sync::Once;
+    static INIT: Once = Once::new();
+
+    INIT.call_once(|| {
+        gst::init().unwrap();
+        gstrstextwrap::plugin_register_static().expect("textwrap test");
+    });
+}
+
+#[test]
+fn test_columns() {
+    init();
+
+    let input = b"Split this text up";
+
+    let expected_output = "Split\nthis\ntext\nup";
+
+    let mut h = gst_check::Harness::new("textwrap");
+
+    {
+        let wrap = h.get_element().expect("Could not create textwrap");
+        wrap.set_property("columns", &5u32).unwrap();
+    }
+
+    h.set_src_caps_str("text/x-raw, format=utf8");
+
+    let buf = {
+        let mut buf = gst::Buffer::from_mut_slice(Vec::from(&input[..]));
+        let buf_ref = buf.get_mut().unwrap();
+        buf_ref.set_pts(gst::ClockTime::from_seconds(0));
+        buf_ref.set_duration(gst::ClockTime::from_seconds(2));
+        buf
+    };
+
+    assert_eq!(h.push(buf), Ok(gst::FlowSuccess::Ok));
+
+    let buf = h.pull().expect("Couldn't pull buffer");
+
+    assert_eq!(buf.get_pts(), 0.into());
+    assert_eq!(buf.get_duration(), 2 * gst::SECOND);
+
+    let map = buf.map_readable().expect("Couldn't map buffer readable");
+
+    assert_eq!(
+        std::str::from_utf8(map.as_ref()),
+        std::str::from_utf8(expected_output.as_ref())
+    );
+}
+
+#[test]
+fn test_lines() {
+    init();
+
+    let input = b"Split this text up";
+
+    let mut h = gst_check::Harness::new("textwrap");
+
+    {
+        let wrap = h.get_element().expect("Could not create textwrap");
+        wrap.set_property("columns", &5u32).unwrap();
+        wrap.set_property("lines", &2u32).unwrap();
+    }
+
+    h.set_src_caps_str("text/x-raw, format=utf8");
+
+    let buf = {
+        let mut buf = gst::Buffer::from_mut_slice(Vec::from(&input[..]));
+        let buf_ref = buf.get_mut().unwrap();
+        buf_ref.set_pts(gst::ClockTime::from_seconds(0));
+        buf_ref.set_duration(gst::ClockTime::from_seconds(2));
+        buf
+    };
+
+    assert_eq!(h.push(buf), Ok(gst::FlowSuccess::Ok));
+
+    let buf = h.pull().expect("Couldn't pull buffer");
+
+    assert_eq!(buf.get_pts(), 0.into());
+    assert_eq!(buf.get_duration(), gst::SECOND);
+
+    let expected_output = "Split\nthis";
+
+    let map = buf.map_readable().expect("Couldn't map buffer readable");
+
+    assert_eq!(
+        std::str::from_utf8(map.as_ref()),
+        std::str::from_utf8(expected_output.as_ref())
+    );
+
+    let buf = h.pull().expect("Couldn't pull buffer");
+
+    assert_eq!(buf.get_pts(), gst::SECOND);
+    assert_eq!(buf.get_duration(), gst::SECOND);
+
+    let expected_output = "text\nup";
+
+    let map = buf.map_readable().expect("Couldn't map buffer readable");
+
+    assert_eq!(
+        std::str::from_utf8(map.as_ref()),
+        std::str::from_utf8(expected_output.as_ref())
+    );
+}