From 9b2ddce115c970343c42d3ba353c36184f27ec95 Mon Sep 17 00:00:00 2001
From: Rob Hallam <0504004h@student.gla.ac.uk>
Date: Fri, 19 Jul 2024 14:49:01 +0100
Subject: [PATCH] feat: add duration-related methods to SourceMedia

---
 pipeline/utils.py | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

diff --git a/pipeline/utils.py b/pipeline/utils.py
index 9c8555a..0e90f0a 100644
--- a/pipeline/utils.py
+++ b/pipeline/utils.py
@@ -45,12 +45,35 @@ class Source():
         if not provider:
             raise ValueError("Provider must be provided") # TODO: #API -- decide if this is necessary
         self.provider = provider
+
     def __str__(self):
         """See: 'accessing the object should return the path to the media file'"""
         return self.path
 
     def __repr__(self):
         return f"Source({self.source}, {self.path}, {self.provider})"
+
+    def duration(self):
+        """Return the duration of the media file at self.path (result is cached)"""
+        return self._duration or self._get_duration(self.path)
+
+    def _get_duration(self, file):
+        """Use ffprobe to get the duration of the media file at self.path and cache result (_duration)
+
+        usage: ffprobe -v error -show_entries format=duration -of default=noprint_wrappers=1:nokey=1 <file>
+        """
+        # test if file exists
+        try:
+            with open(file) as _:
+                pass
+        except FileNotFoundError:
+            raise FileNotFoundError(f"File not found: {file}")
+        # cache the result
+        self._duration = 0.0 or float(subprocess.check_output(["ffprobe", "-v", "error", "-show_entries", "format=duration", "-of", "default=noprint_wrappers=1:nokey=1", file]))
+        return self._duration
+
+
+
 class Interval():
     """An interval of time in a media file