PrivateCoffee/invidious
4
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 2

Extract label for transcript in YouTube response

Browse source
This commit is contained in:
syeopite 2024-06-13 09:05:47 -07:00
parent 5b519123a7
commit f466116cd7
No known key found for this signature in database
GPG key ID: A73C186DA3955A1A
1 changed files with 25 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

29
src/invidious/videos/transcript.cr
View file

@ -8,11 +8,16 @@ module Invidious::Videos
alias TranscriptLine = HeadingLine | RegularLine alias TranscriptLine = HeadingLine | RegularLine
property lines : Array(TranscriptLine) property lines : Array(TranscriptLine)
property language_code : String property language_code : String
property auto_generated : Bool property auto_generated : Bool
# User friendly label for the current transcript.
# Example: "English (auto-generated)"
property label : String
# Initializes a new Transcript struct with the contents and associated metadata describing it # Initializes a new Transcript struct with the contents and associated metadata describing it
def initialize(@lines : Array(TranscriptLine), @language_code : String, @auto_generated : Bool) def initialize(@lines : Array(TranscriptLine), @language_code : String, @auto_generated : Bool, @label : String)
end end
# Generates a protobuf string to fetch the requested transcript from YouTube # Generates a protobuf string to fetch the requested transcript from YouTube
@ -45,14 +50,29 @@ module Invidious::Videos
# Constructs a Transcripts struct from the initial YouTube response # Constructs a Transcripts struct from the initial YouTube response
def self.from_raw(initial_data : Hash(String, JSON::Any), language_code : String, auto_generated : Bool) def self.from_raw(initial_data : Hash(String, JSON::Any), language_code : String, auto_generated : Bool)
segment_list = initial_data.dig("actions", 0, "updateEngagementPanelAction", "content", "transcriptRenderer", transcript_panel = initial_data.dig("actions", 0, "updateEngagementPanelAction", "content", "transcriptRenderer",
"content", "transcriptSearchPanelRenderer", "body", "transcriptSegmentListRenderer" "content", "transcriptSearchPanelRenderer")
)
segment_list = transcript_panel.dig("body", "transcriptSegmentListRenderer")
if !segment_list["initialSegments"]? if !segment_list["initialSegments"]?
raise NotFoundException.new("Requested transcript does not exist") raise NotFoundException.new("Requested transcript does not exist")
end end
# Extract user-friendly label for the current transcript
footer_language_menu = transcript_panel.dig?(
"footer", "transcriptFooterRenderer", "languageMenu", "sortFilterSubMenuRenderer", "subMenuItems"
)
if footer_language_menu
label = footer_language_menu.as_a.select(&.["selected"].as_bool)[0]["title"].as_s
else
label = language_code
end
# Extract transcript lines
initial_segments = segment_list["initialSegments"].as_a initial_segments = segment_list["initialSegments"].as_a
lines = [] of TranscriptLine lines = [] of TranscriptLine
@ -76,6 +96,7 @@ module Invidious::Videos
lines: lines, lines: lines,
language_code: language_code, language_code: language_code,
auto_generated: auto_generated, auto_generated: auto_generated,
label: label
) )
end end