subber 0.1.3 → 0.1.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: ad658615874be25c00f5f43e4a2900cf87e6d281
-  data.tar.gz: 1da66b5b728cfa6f1e5f763cda3145e599fd5c5f
+  metadata.gz: 9c082a74b89529051c409e05338515568c17dc26
+  data.tar.gz: aa6dc20d6082c75894abb08261e55baca483e708
 SHA512:
-  metadata.gz: 815c21c2147784312e22aeb391653e00e0de96daef0f070d9b3859f755b9673dcd9cbb8b1c037e3ced42eaa6e4e19bda6f380fc328e7f7d509d2f3d4df17ffb9
-  data.tar.gz: 29567db3a1f1d948215976250b1f19f714f8b7d7142ef25a28dea36e6f4a04077d608b5117bb73791b9c1a59de0ab5c836aeb0b273006c2a8c0b68e07eecebee
+  metadata.gz: e83d6e85c03d1cc6f4e9fa3e980d034361495d05134f93a48836ead18b882b55e9e225800fdde5745b7f446871f6d2ef6774100a0badf5fb30d41ec0ed444f3b
+  data.tar.gz: 7c3d78acb225b3d14ec74c487c1a0e150a73e7f799a8cfa08c48d40770dee57142d05d8af39d936ba590c8e029d2c91118250b71dd2273661c3cc0846a70e26f

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    subber (0.1.2)
+    subber (0.1.4)
 
 GEM
   remote: https://rubygems.org/

data/lib/subber/formatter/vtt.rb

@@ -8,12 +8,12 @@ module Subber::Formatter
       # @return [String]
       #
       def format(subtitles)
-        subtitle_texts =
+        cues =
           subtitles.map do |subtitle|
-            convert_subtitle_to_text(subtitle)
+            convert_subtitle_to_cue(subtitle)
           end
 
-        file_content = subtitle_texts.join
+        file_content = cues.join
         file_content = add_webvtt_header(file_content)
         file_content = add_window_line_break(file_content)
         file_content
@@ -38,14 +38,14 @@ module Subber::Formatter
       # @param subtitle [Subber::Subtitle]
       # @return [String]
       #
-      def convert_subtitle_to_text(subtitle)
-        subtitle_text = [
+      def convert_subtitle_to_cue(subtitle)
+        cue = [
           build_counter(subtitle),
           build_time_range(subtitle),
           build_content(subtitle),
         ].join("\n")
 
-        "#{subtitle_text}\n\n"
+        "#{cue}\n\n"
       end
 
       # @param subtitle [Subber::Subtitle]

data/lib/subber/parser/vtt.rb

@@ -1,14 +1,14 @@
 module Subber::Parser
   class Vtt < Base
-    SUBTITLE_REGEX = /([^\n]*)\n([^\n]*)(\n(.*))?/m
+    SUBTITLE_REGEX = /(\d*)\n?(^\d{0,2}:?\d{2}:\d{2}\.\d{3}\s-->\s\d{0,2}:?\d{2}:\d{2}\.\d{3}$)\n?(.*)/m
     COUNTER_REGEX = /\d+/
-    TIME_RANGE_REGEX = /(\d{2}:\d{2}:\d{2}\.\d{3})\s*-->\s*(\d{2}:\d{2}:\d{2}\.\d{3})/
-    TIMECODE_REGEX = /(\d{2}):(\d{2}):(\d{2})\.(\d{3})/
+    TIME_RANGE_REGEX = /(^\d{0,2}:?\d{2}:\d{2}\.\d{3})\s-->\s(\d{0,2}:?\d{2}:\d{2}\.\d{3}$)/
+    TIMECODE_REGEX = /(^\d{0,2}):?(\d{2}):(\d{2})\.(\d{3})/
 
-    DELIMITER_REGEX = /\n?\n\n/
+    CUE_DELIMITER_REGEX = /\n\n/
     WINDOW_LINE_BREAK_REGEX = /\r/
-    WEBVTT_HEADER_REGEX = /WEBVTT\n\n/
     BYTE_ORDER_MARK_STRING = "\xEF\xBB\xBF"
+    INVALID_CUE_START_STRINGS = %w(WEBVTT NOTE STYLE REGION)
 
     class << self
       # @param file_content [String]
@@ -16,38 +16,36 @@ module Subber::Parser
       #
       def parse(file_content)
         file_content = remove_window_line_break(file_content)
-        file_content = remove_webvtt_header(file_content)
+        cues = extract_cues(file_content)
 
-        subtitle_texts = file_content.split(DELIMITER_REGEX)
-        subtitle_texts.map do |subtitle_text|
-          convert_text_to_subtitle(subtitle_text)
+        cues.map.with_index do |cue, index|
+          convert_cue_to_subtitle(cue, index)
         end
       end
 
       private
 
       # @param file_content [String]
-      # @return [String]
+      # @return [Array<String>]
       #
-      def remove_webvtt_header(file_content)
-        file_content.sub(WEBVTT_HEADER_REGEX, '')
+      def extract_cues(file_content)
+        cues = file_content.split(CUE_DELIMITER_REGEX)
+        cues.reject do |cue|
+          cue.start_with?(*INVALID_CUE_START_STRINGS)
+        end
       end
 
-      # @param file_content [String]
-      # @return [String]
+      # @param cue [String]
+      # @param index [Integer]
+      # @return [Array<Subber::Subtitle>]
       #
-      def remove_window_line_break(file_content)
-        file_content.gsub(WINDOW_LINE_BREAK_REGEX, '')
-      end
+      def convert_cue_to_subtitle(cue, index)
+        matches = cue.match(SUBTITLE_REGEX).to_a
+        raise(Subber::Errors::InvalidVttFormat, cue) if matches.empty?
 
-      # @param subtitle_text [String]
-      # @return [Subber::Subtitle]
-      #
-      def convert_text_to_subtitle(subtitle_text)
-        matches = subtitle_text.match(SUBTITLE_REGEX).to_a
-        raise(Subber::Errors::InvalidSrtFormat, subtitle_text) if matches.empty?
+        _cue, counter, time_range_string, content = matches
 
-        _subtitle_text, counter, time_range_string, _new_line, content = matches
+        counter = (index + 1).to_s if counter.empty?
 
         counter = extract_counter(counter)
         from, to = extract_time_range(time_range_string)
@@ -56,14 +54,21 @@ module Subber::Parser
           counter: counter,
           start_time: convert_time_to_ms(from),
           end_time: convert_time_to_ms(to),
-          content: content
+          content: content.strip
         )
       rescue Subber::Errors::InvalidCounter
-        raise(Subber::Errors::InvalidCounter, subtitle_text)
+        raise(Subber::Errors::InvalidCounter, cue)
       rescue Subber::Errors::InvalidTimeRange
-        raise(Subber::Errors::InvalidTimeRange, subtitle_text)
+        raise(Subber::Errors::InvalidTimeRange, cue)
       rescue Subber::Errors::InvalidTimestamp
-        raise(Subber::Errors::InvalidTimestamp, subtitle_text)
+        raise(Subber::Errors::InvalidTimestamp, cue)
+      end
+
+      # @param file_content [String]
+      # @return [String]
+      #
+      def remove_window_line_break(file_content)
+        file_content.gsub(WINDOW_LINE_BREAK_REGEX, '')
       end
 
       # @param  counter_string [String]

data/lib/subber/version.rb

@@ -1,3 +1,3 @@
 module Subber
-  VERSION = "0.1.3"
+  VERSION = "0.1.4"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: subber
 version: !ruby/object:Gem::Version
-  version: 0.1.3
+  version: 0.1.4
 platform: ruby
 authors:
 - Viki
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2018-05-04 00:00:00.000000000 Z
+date: 2018-05-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler