From 9d3be321e7758027b8726ca056e1ea9277b24572 Mon Sep 17 00:00:00 2001 From: Kovid Goyal Date: Wed, 9 May 2007 21:47:06 +0000 Subject: [PATCH] Handle comments in tags correctly --- src/libprs500/lrf/html/convert_from.py | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/src/libprs500/lrf/html/convert_from.py b/src/libprs500/lrf/html/convert_from.py index 7fd671452a..51126e69d1 100644 --- a/src/libprs500/lrf/html/convert_from.py +++ b/src/libprs500/lrf/html/convert_from.py @@ -27,7 +27,10 @@ from urlparse import urlparse from tempfile import mkdtemp from operator import itemgetter from math import ceil, floor -from PIL import Image as PILImage +try: + from PIL import Image as PILImage +except ImportError: + import Image as PILImage from libprs500.lrf.html.BeautifulSoup import BeautifulSoup, Comment, Tag, \ NavigableString, Declaration, ProcessingInstruction @@ -424,10 +427,10 @@ class HTMLConverter(object): return '' text = '' for c in tag.contents: - if isinstance(c, NavigableString): - text += str(c) - elif isinstance(c, Comment): + if isinstance(c, HTMLConverter.IGNORED_TAGS): return '' + if isinstance(c, NavigableString): + text += str(c) elif isinstance(c, Tag): text += self.get_text(c) return text