I won't guarantee this is the 'best algorithm', but here's what we use. (This
is in a final class with only static helper methods):
// Set of characters / Strings SOLR treats as having special meaning in a
query, and the corresponding Escaped versions.
// Note that the actual operators '&&' and '||' don't show up here - we'll
just escape the characters '&' and '|' wherever they occur.
private static final String[] SOLR_SPECIAL_CHARACTERS = new String[] {"+",
"-", "&", "|", "!", "(", ")", "{", "}", "[", "]", "^", "\"", "~", "*", "?",
":", "\\"};
private static final String[] SOLR_REPLACEMENT_CHARACTERS = new String[]
{"\\+", "\\-", "\\&", "\\|", "\\!", "\\(", "\\)", "\\{", "\\}", "\\[", "\\]",
"\\^", "\\\"", "\\~", "\\*", "\\?", "\\:", "\\\\"};
/**
* Escapes all special characters from the Search Terms, so they don't get
confused with
* the Solr query language special characters.
* @param value - Search Term to escape
* @return - escaped Search value, suitable for a Solr "q" parameter
*/
public static String escapeSolrCharacters(String value)
{
return StringUtils.replaceEach(value, SOLR_SPECIAL_CHARACTERS,
SOLR_REPLACEMENT_CHARACTERS);
}
Bob Sandiford | Lead Software Engineer | SirsiDynix
P: 800.288.8020 X6943 | bob.sandif...@sirsidynix.com
www.sirsidynix.com
> -----Original Message-----
> From: Bill Bell [mailto:billnb...@gmail.com]
> Sent: Sunday, September 25, 2011 12:22 AM
> To: solr-user@lucene.apache.org
> Subject: Best Solr escaping?
>
> What is the best algorithm for escaping strings before sending to Solr?
> Does
> someone have some code?
>
> A few things I have witnessed in "q" using DIH handler
> * Double quotes - " that are not balanced can cause several issues from
> an
> error (strip the double quote?), to no results.
> * Should we use + or %20 and what cases make sense:
> > * "Dr. Phil Smith" or "Dr.+Phil+Smith" or "Dr.%20Phil%20Smith" - also
> what is
> > the impact of double quotes?
> * Unmatched parenthesis I.e. Opening ( and not closing.
> > * (Dr. Holstein
> > * Cardiologist+(Dr. Holstein
> Regular encoding of strings does not always work for the whole string
> due to
> several issues like white space:
> * White space works better when we use back quote "Bill\ Bell"
> especially
> when using facets.
>
> Thoughts? Code? Ideas? Better Wikis?
>
>
