y2t added

2022-12-29 12:22:48 +01:00 · 2022-12-29 12:22:48 +01:00 · 5e3d902471
parent 3b93c9ffbf
commit 5e3d902471
3 changed files with 90 additions and 0 deletions
--- a/proj/y2t/handler.php
+++ b/proj/y2t/handler.php
--- a/proj/y2t/index.php
+++ b/proj/y2t/index.php
@ -0,0 +1,27 @@
+<!DOCTYPE HTML>
+<html>
+<head>
+<title>Y2T</title>
+  <meta charset="utf-8">
+  <meta name="viewport" content="width=device-width, initial-scale=1.0">
+  <link rel="stylesheet" href="style.css">
+</head>
+<body>
+
+<div id="root">
+
+    <form method="post" action="handler.php">
+        <label for="url">URL:</label><br>
+        <input type="text" id="url" name="url"><br>
+        <label for="lang">Language:</label><br>
+        <input type="text" id="lang" name="lang"><br><br>
+        <input type="submit" value="Submit">
+      </form>
+      
+  
+</div>
+<script src="assets/js/vue.min.js"></script>
+<script src="assets/js/app.js"></script>
+
+</body>
+</html>
--- a/proj/y2t/y2t.py
+++ b/proj/y2t/y2t.py
@ -0,0 +1,63 @@
+import youtube_dl
+import sys
+
+url = sys.argv[1]
+lang = sys.argv[2]
+
+ydl_opts = {
+    'writesubtitles': True,
+    'writeautomaticsub': True,
+    'subtitlesformat': 'vtt',
+    'subtitleslangs': [lang],
+    'lkjsdlfkjsdf': 'sdflkj',
+    'outtmpl': '%(id)s',
+    'skip_download': True,
+    'writethumbnail': True,
+    'quiet': True
+}
+
+ydl = youtube_dl.YoutubeDL(ydl_opts)
+
+ydl.download([url])
+
+info = ydl.extract_info(url, download=False)
+
+creator = info['uploader']
+title = info['title']
+id = info['id']
+
+subFn = id + "." + lang + ".vtt"
+
+txt = ""
+
+with open(subFn, "r") as f:
+    lines = f.readlines()
+    
+    for i in range(len(lines)):
+        if i < len(lines) - 1 and ">" not in lines[i] and ":" not in lines[i] and lines[i].strip() != "" and lines[i + 1].strip() == "":
+            txt += lines[i]
+            
+          
+txt = txt.replace("\n"," ")  
+#print(txt)
+words = txt.split()
+
+sList = []
+wCount = 0
+maxWordCount = 1000
+currentString = ""
+
+for w in words:
+    wCount +=1
+    currentString += w + " "
+    if wCount == maxWordCount:
+        sList.append(currentString)
+        wCount = 0
+        currentString = ""
+if currentString:
+    sList.append(currentString)
+
+blocks = len(sList)
+
+for i in range(blocks):
+    print('This is part '+ str(i+1) +'/'+ str(blocks) + "of the transcript of a video named: "+ title)