a

7 files changed, 440 insertions, 0 deletions

diff --git a/config/mpv/scripts/subs2srs/utils/base64.lua b/config/mpv/scripts/subs2srs/utils/base64.lua
new file mode 100644
index 0000000..0fe2d06
--- /dev/null
+++ b/config/mpv/scripts/subs2srs/utils/base64.lua
@@ -0,0 +1,46 @@
+--[[
+Copyright: Ren Tatsumoto and contributors
+License: GNU GPL, version 3 or later; http://www.gnu.org/licenses/gpl.html
+
+Encoding and decoding in base64
+]]
+
+-- http://lua-users.org/wiki/BaseSixtyFour
+
+-- character table string
+local b = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/'
+
+-- encoding
+local function enc(data)
+    return ((data:gsub('.', function(x)
+        local r,b='',x:byte()
+        for i=8,1,-1 do r=r..(b%2^i-b%2^(i-1)>0 and '1' or '0') end
+        return r;
+    end)..'0000'):gsub('%d%d%d?%d?%d?%d?', function(x)
+        if (#x < 6) then return '' end
+        local c=0
+        for i=1,6 do c=c+(x:sub(i,i)=='1' and 2^(6-i) or 0) end
+        return b:sub(c+1,c+1)
+    end)..({ '', '==', '=' })[#data%3+1])
+end
+
+-- decoding
+local function dec(data)
+    data = string.gsub(data, '[^'..b..'=]', '')
+    return (data:gsub('.', function(x)
+        if (x == '=') then return '' end
+        local r,f='',(b:find(x)-1)
+        for i=6,1,-1 do r=r..(f%2^i-f%2^(i-1)>0 and '1' or '0') end
+        return r;
+    end):gsub('%d%d%d?%d?%d?%d?%d?%d?', function(x)
+        if (#x ~= 8) then return '' end
+        local c=0
+        for i=1,8 do c=c+(x:sub(i,i)=='1' and 2^(8-i) or 0) end
+        return string.char(c)
+    end))
+end
+
+return {
+    enc = enc,
+    dec = dec,
+}
diff --git a/config/mpv/scripts/subs2srs/utils/filename_factory.lua b/config/mpv/scripts/subs2srs/utils/filename_factory.lua
new file mode 100644
index 0000000..a794fc7
--- /dev/null
+++ b/config/mpv/scripts/subs2srs/utils/filename_factory.lua
@@ -0,0 +1,89 @@
+--[[
+Copyright: Ren Tatsumoto and contributors
+License: GNU GPL, version 3 or later; http://www.gnu.org/licenses/gpl.html
+
+Creates image and audio filenames compatible with Anki.
+]]
+
+local mp = require('mp')
+local h = require('helpers')
+
+local filename
+
+local anki_compatible_length = (function()
+    -- Anki forcibly mutilates all filenames longer than 119 bytes when you run `Tools->Check Media...`.
+    local allowed_bytes = 119
+    local timestamp_bytes = #'_99h99m99s999ms-99h99m99s999ms.webp'
+
+    return function(str, timestamp)
+        -- if timestamp provided, recalculate limit_bytes
+        local limit_bytes = allowed_bytes - (timestamp and #timestamp or timestamp_bytes)
+
+        if #str <= limit_bytes then
+            return str
+        end
+
+        local bytes_per_char = h.contains_non_latin_letters(str) and #'車' or #'z'
+        local limit_chars = math.floor(limit_bytes / bytes_per_char)
+
+        if limit_chars == limit_bytes then
+            return str:sub(1, limit_bytes)
+        end
+
+        local ret = h.subprocess {
+            'awk',
+            '-v', string.format('str=%s', str),
+            '-v', string.format('limit=%d', limit_chars),
+            'BEGIN{print substr(str, 1, limit); exit}'
+        }
+
+        if ret.status == 0 then
+            ret.stdout = h.remove_newlines(ret.stdout)
+            ret.stdout = h.remove_leading_trailing_spaces(ret.stdout)
+            return ret.stdout
+        else
+            return 'subs2srs_' .. os.time()
+        end
+    end
+end)()
+
+local make_media_filename = function()
+    filename = mp.get_property("filename") -- filename without path
+    filename = h.remove_extension(filename)
+    filename = h.remove_filename_text_in_parentheses(filename)
+    filename = h.remove_text_in_brackets(filename)
+    filename = h.remove_special_characters(filename)
+end
+
+local function timestamp_range(start_timestamp, end_timestamp, extension)
+    -- Generates a filename suffix of the form: _00h00m00s000ms-99h99m99s999ms.extension
+    -- Extension must already contain the dot.
+    return string.format(
+            '_%s_%s%s',
+            h.human_readable_time(start_timestamp),
+            h.human_readable_time(end_timestamp),
+            extension
+    )
+end
+
+local function timestamp_static(timestamp, extension)
+    -- Generates a filename suffix of the form: _00h00m00s000ms.extension
+    -- Extension must already contain the dot.
+    return string.format(
+            '_%s%s',
+            h.human_readable_time(timestamp),
+            extension
+    )
+end
+
+local make_filename = function(...)
+    local args = {...}
+    local timestamp = #args < 3 and timestamp_static(...) or timestamp_range(...)
+    return string.lower(anki_compatible_length(filename, timestamp) .. timestamp)
+end
+
+mp.register_event("file-loaded", make_media_filename)
+
+return {
+    make_filename = make_filename,
+}
diff --git a/config/mpv/scripts/subs2srs/utils/forvo.lua b/config/mpv/scripts/subs2srs/utils/forvo.lua
new file mode 100644
index 0000000..09bc596
--- /dev/null
+++ b/config/mpv/scripts/subs2srs/utils/forvo.lua
@@ -0,0 +1,145 @@
+--[[
+Copyright: Ren Tatsumoto and contributors
+License: GNU GPL, version 3 or later; http://www.gnu.org/licenses/gpl.html
+
+Utils for downloading pronunciations from Forvo
+]]
+
+local utils = require('mp.utils')
+local msg = require('mp.msg')
+local h = require('helpers')
+local base64 = require('utils.base64')
+local self = {
+    output_dir_path = nil,
+}
+
+local function url_encode(url)
+    -- https://gist.github.com/liukun/f9ce7d6d14fa45fe9b924a3eed5c3d99
+    local char_to_hex = function(c)
+        return string.format("%%%02X", string.byte(c))
+    end
+    if url == nil then
+        return
+    end
+    url = url:gsub("\n", "\r\n")
+    url = url:gsub("([^%w _%%%-%.~])", char_to_hex)
+    url = url:gsub(" ", "+")
+    return url
+end
+
+local function reencode(source_path, dest_path)
+    local args = {
+        'mpv',
+        source_path,
+        '--loop-file=no',
+        '--keep-open=no',
+        '--video=no',
+        '--no-ocopy-metadata',
+        '--no-sub',
+        '--audio-channels=mono',
+        '--oacopts-add=vbr=on',
+        '--oacopts-add=application=voip',
+        '--oacopts-add=compression_level=10',
+        '--af-append=silenceremove=1:0:-50dB',
+        table.concat { '--oac=', self.config.audio_codec },
+        table.concat { '--of=', self.config.audio_format },
+        table.concat { '--oacopts-add=b=', self.config.audio_bitrate },
+        table.concat { '-o=', dest_path }
+    }
+    return h.subprocess(args)
+end
+
+local function reencode_and_store(source_path, filename)
+    local reencoded_path = utils.join_path(self.output_dir_path, filename)
+    local result = reencode(source_path, reencoded_path)
+    return result.status == 0
+end
+
+local function curl_save(source_url, save_location)
+    local curl_args = { 'curl', source_url, '-s', '-L', '-o', save_location }
+    return h.subprocess(curl_args).status == 0
+end
+
+local function get_pronunciation_url(word)
+    local file_format = self.config.audio_extension:sub(2)
+    local forvo_page = h.subprocess { 'curl', '-s', string.format('https://forvo.com/search/%s/ja', url_encode(word)) }.stdout
+    local play_params = string.match(forvo_page, "Play%((.-)%);")
+
+    if play_params then
+        local iter = string.gmatch(play_params, "'(.-)'")
+        local formats = { mp3 = iter(), ogg = iter() }
+        return string.format('https://audio00.forvo.com/%s/%s', file_format, base64.dec(formats[file_format]))
+    end
+end
+
+local function make_forvo_filename(word)
+    return string.format('forvo_%s%s', self.platform.windows and os.time() or word, self.config.audio_extension)
+end
+
+local function get_forvo_pronunciation(word)
+    local audio_url = get_pronunciation_url(word)
+
+    if h.is_empty(audio_url) then
+        msg.warn(string.format("Seems like Forvo doesn't have audio for word %s.", word))
+        return
+    end
+
+    local filename = make_forvo_filename(word)
+    local tmp_filepath = utils.join_path(self.platform.tmp_dir(), filename)
+
+    local result
+    if curl_save(audio_url, tmp_filepath) and reencode_and_store(tmp_filepath, filename) then
+        result = string.format(self.config.audio_template, filename)
+    else
+        msg.warn(string.format("Couldn't download audio for word %s from Forvo.", word))
+    end
+
+    os.remove(tmp_filepath)
+    return result
+end
+
+local append = function(new_data, stored_data)
+    if self.config.use_forvo == 'no' then
+        -- forvo functionality was disabled in the config file
+        return new_data
+    end
+
+    if type(stored_data[self.config.vocab_audio_field]) ~= 'string' then
+        -- there is no field configured to store forvo pronunciation
+        return new_data
+    end
+
+    if h.is_empty(stored_data[self.config.vocab_field]) then
+        -- target word field is empty. can't continue.
+        return new_data
+    end
+
+    if self.config.use_forvo == 'always' or h.is_empty(stored_data[self.config.vocab_audio_field]) then
+        local forvo_pronunciation = get_forvo_pronunciation(stored_data[self.config.vocab_field])
+        if not h.is_empty(forvo_pronunciation) then
+            if self.config.vocab_audio_field == self.config.audio_field then
+                -- improperly configured fields. don't lose sentence audio
+                new_data[self.config.audio_field] = forvo_pronunciation .. new_data[self.config.audio_field]
+            else
+                new_data[self.config.vocab_audio_field] = forvo_pronunciation
+            end
+        end
+    end
+
+    return new_data
+end
+
+local set_output_dir = function(dir_path)
+    self.output_dir_path = dir_path
+end
+
+local function init(config, platform)
+    self.config = config
+    self.platform = platform
+end
+
+return {
+    append = append,
+    init = init,
+    set_output_dir = set_output_dir,
+}
diff --git a/config/mpv/scripts/subs2srs/utils/pause_timer.lua b/config/mpv/scripts/subs2srs/utils/pause_timer.lua
new file mode 100644
index 0000000..e37b0ea
--- /dev/null
+++ b/config/mpv/scripts/subs2srs/utils/pause_timer.lua
@@ -0,0 +1,33 @@
+--[[
+Copyright: Ren Tatsumoto and contributors
+License: GNU GPL, version 3 or later; http://www.gnu.org/licenses/gpl.html
+
+Pause timer stops playback when reaching a set timing.
+]]
+
+local mp = require('mp')
+local stop_time = -1
+local check_stop
+
+local set_stop_time = function(time)
+    stop_time = time
+    mp.observe_property("time-pos", "number", check_stop)
+end
+
+local stop = function()
+    mp.unobserve_property(check_stop)
+    stop_time = -1
+end
+
+check_stop = function(_, time)
+    if time > stop_time then
+        stop()
+        mp.set_property("pause", "yes")
+    end
+end
+
+return {
+    set_stop_time = set_stop_time,
+    check_stop = check_stop,
+    stop = stop,
+}
diff --git a/config/mpv/scripts/subs2srs/utils/play_control.lua b/config/mpv/scripts/subs2srs/utils/play_control.lua
new file mode 100644
index 0000000..901377d
--- /dev/null
+++ b/config/mpv/scripts/subs2srs/utils/play_control.lua
@@ -0,0 +1,61 @@
+--[[
+Copyright: Ren Tatsumoto and contributors
+License: GNU GPL, version 3 or later; http://www.gnu.org/licenses/gpl.html
+
+Provides additional methods for controlling playback.
+]]
+
+local mp = require('mp')
+local h = require('helpers')
+local pause_timer = require('utils.pause_timer')
+local Subtitle = require('subtitles.subtitle')
+
+local current_sub
+
+local function stop_at_the_end(sub)
+    pause_timer.set_stop_time(sub['end'] - 0.050)
+    h.notify("Playing till the end of the sub...", "info", 3)
+end
+
+local function play_till_sub_end()
+    local sub = Subtitle:now()
+    mp.commandv('seek', sub['start'], 'absolute')
+    mp.set_property("pause", "no")
+    stop_at_the_end(sub)
+end
+
+local function sub_seek(direction, pause)
+    mp.commandv("sub_seek", direction == 'backward' and '-1' or '1')
+    mp.commandv("seek", "0.015", "relative+exact")
+    if pause then
+        mp.set_property("pause", "yes")
+    end
+    pause_timer.stop()
+end
+
+local function sub_rewind()
+    mp.commandv('seek', Subtitle:now()['start'] + 0.015, 'absolute')
+    pause_timer.stop()
+end
+
+local function check_sub()
+    local sub = Subtitle:now()
+    if sub and sub ~= current_sub then
+        mp.unobserve_property(check_sub)
+        stop_at_the_end(sub)
+    end
+end
+
+local function play_till_next_sub_end()
+    current_sub = Subtitle:now()
+    mp.observe_property("sub-text", "string", check_sub)
+    mp.set_property("pause", "no")
+    h.notify("Waiting till next sub...", "info", 10)
+end
+
+return {
+    play_till_sub_end = play_till_sub_end,
+    play_till_next_sub_end = play_till_next_sub_end,
+    sub_seek = sub_seek,
+    sub_rewind = sub_rewind,
+}
diff --git a/config/mpv/scripts/subs2srs/utils/switch.lua b/config/mpv/scripts/subs2srs/utils/switch.lua
new file mode 100644
index 0000000..5dac1c6
--- /dev/null
+++ b/config/mpv/scripts/subs2srs/utils/switch.lua
@@ -0,0 +1,38 @@
+--[[
+Copyright: Ren Tatsumoto and contributors
+License: GNU GPL, version 3 or later; http://www.gnu.org/licenses/gpl.html
+
+Switch cycles between values in a table.
+]]
+
+local make_switch = function(states)
+    local self = {
+        states = states,
+        current_state = 1
+    }
+    local bump = function()
+        self.current_state = self.current_state + 1
+        if self.current_state > #self.states then
+            self.current_state = 1
+        end
+    end
+    local get = function()
+        return self.states[self.current_state]
+    end
+    local set = function(new_state)
+        for idx, value in ipairs(self.states) do
+            if value == new_state then
+                self.current_state = idx
+            end
+        end
+    end
+    return {
+        bump = bump,
+        get = get,
+        set = set,
+    }
+end
+
+return {
+    new = make_switch
+}
diff --git a/config/mpv/scripts/subs2srs/utils/timings.lua b/config/mpv/scripts/subs2srs/utils/timings.lua
new file mode 100644
index 0000000..d2408d6
--- /dev/null
+++ b/config/mpv/scripts/subs2srs/utils/timings.lua
@@ -0,0 +1,28 @@
+--[[
+Copyright: Ren Tatsumoto and contributors
+License: GNU GPL, version 3 or later; http://www.gnu.org/licenses/gpl.html
+
+Object that remembers manually set timings.
+]]
+
+local new_timings = function()
+    local self = { ['start'] = -1, ['end'] = -1, }
+    local is_set = function(position)
+        return self[position] >= 0
+    end
+    local set = function(position, time)
+        self[position] = time
+    end
+    local get = function(position)
+        return self[position]
+    end
+    return {
+        is_set = is_set,
+        set = set,
+        get = get,
+    }
+end
+
+return {
+    new = new_timings,
+}