From e877371a0b86775f3793f7eb4e9512436f8ed6f0 Mon Sep 17 00:00:00 2001 From: j Date: Fri, 18 Mar 2016 18:35:41 +0100 Subject: [PATCH] avoid space titles --- oml/media/epub.py | 3 +++ oml/media/pdf.py | 3 +++ 2 files changed, 6 insertions(+) diff --git a/oml/media/epub.py b/oml/media/epub.py index b848200..81ccb17 100644 --- a/oml/media/epub.py +++ b/oml/media/epub.py @@ -178,6 +178,9 @@ def info(epub): data['date'] = data['date'].split('T')[0] if 'language' in data and isinstance(data['language'], str): data['language'] = get_language(data['language']) + for key in list(data): + if isinstance(data[key], str) and not data[key].strip(): + del data[key] return data def extract_text(path): diff --git a/oml/media/pdf.py b/oml/media/pdf.py index eb7502c..fdf5a01 100644 --- a/oml/media/pdf.py +++ b/oml/media/pdf.py @@ -224,6 +224,9 @@ def info(pdf): for key in ('title', 'author', 'producer', 'creator'): if key in data: del data[key] + for key in list(data): + if isinstance(data[key], str) and not data[key].strip(): + del data[key] return data '''