apachesolr_clean_text() should strip extra spaces

Currently apachesolr_clean_text() strips out all html tags from the $node->body and replaces them with spaces. This is fine (prevents running words together) but if there a lot of html tags in the content, then a lot of extra space gets added.

This causes problems, especially in apachesolr_index_node_solr_document(). When building the document teaser, truncate_utf8() is used, which could result in empty teasers if there is a lot of space in the cleaned node body.

Attached patch strips extra spaces from cleaned text.