From ba66f938e978e5472d31606fd46b1081ae6fcbc9 Mon Sep 17 00:00:00 2001 From: Tucker Evans Date: Mon, 11 Dec 2017 00:46:11 -0500 Subject: CSC2621/assignments/search: Commiting to change computers. See next commit for changes --- search/index.go | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) (limited to 'search') diff --git a/search/index.go b/search/index.go index a0efb6c..43d086b 100644 --- a/search/index.go +++ b/search/index.go @@ -28,6 +28,7 @@ type document struct { } var r *regexp.Regexp; +var nonAN *regexp.Regexp; func newDocument() *document { return &document{nil, nil}; @@ -97,17 +98,27 @@ func parseDoc(fd *os.File) (*document, error) { text = r.ReplaceAllString(text, "> <"); t_text = r.ReplaceAllString(text, "> <"); - r_doc = newDocument(); + text = sanitize.HTML(text); + t_text = sanitize.HTML(t_text); + + text = nonAN.ReplaceAllString(text, " "); + t_text = nonAN.ReplaceAllString(t_text, " "); + r_doc = newDocument(); r_doc.text = strings.Fields(sanitize.HTML(text)); r_doc.title = strings.Fields(sanitize.HTML(t_text)); return r_doc, nil; } +func printIndex(w map[string]*wordList) string { +return ""; +} + func init() { log.SetOutput(os.Stderr); r, _ = regexp.Compile("><"); + nonAN, _ = regexp.Compile("[<>!@#&()–[{}]:;,?/*'\"]|( and)|( a)|( an)|( and)|( are)|( as)|( at)|( be)|( by)|( for)|( from)|( has)|( he)|( in)|( is)|( it)|( its)|( of)|( on)|( that)|( the)|( to)|( was)|( were)|( will)|( with)") } func main() { -- cgit v1.1