Login | Register
My pages Projects Community openCollabNet

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Catacomb] Re: A patch for cadaver-dasl



Hello,

I appreciate your comments and attach a new patch for you.

On Fri, 30 Aug 2002, Joe Orton wrote:

> > +++ search.c	Mon Aug 26 20:23:05 2002
> > @@ -0,0 +1,1423 @@
> > +/*
> > +   'search' for cadaver
> > +   Copyright (C) 2000-2001, Joe Orton <joe@manyfish.co.uk>,
>
> I'm still not really happy about this, since I didn't write the code.
> Without a formal copyright assignment I'd prefer you owned the copyright
> (or UCSC, or whatever).

Changed it to Grase Lab, UCSC.

> > +typedef struct dead_prop_t {
>
> POSIX owns the "*_t" namespace, I'd prefer to not invade that... you
> probably don't need the tag name for this typedef anyway, so can you
> just drop the "dead_prop_t" out of this declaration, and similarly with
> resource_t.

Done.

> Declaring a "resource" or a "struct resource_t" is also confusing, since
> cadaver.h already defines a "struct resource", I'm not sure what can be
> done about that though.

Changed it to search_res.

> > +    case ELEM_href : /* href */
> > +	if(sctx->curr==NULL)
> > +	    set_xml_error(sctx, "XML : </%s> is in the wrong place", elm->name);
> ...
> > +	if(sctx->curr==NULL)
> > +	    set_xml_error(sctx, "XML : </%s> is in the wrong place", elm->name);
> ...
> > +	if(sctx->curr==NULL)
> > +	    set_xml_error(sctx, "XML : </%s> is in the wrong place", elm->name);
>
> Way too much duplicated code in there!

Added a new MACRO and replaced them to the macro.

> This is equivalent to ne_accept_207 from ne_207.h.

Done.

Again, I appreciate your help.

--
Sung Kim

Index: Makefile.in
===================================================================
RCS file: /home/cvs/cadaver/Makefile.in,v
retrieving revision 1.45
diff -u -r1.45 Makefile.in
--- Makefile.in	2002/08/03 21:38:19	1.45
+++ Makefile.in	2002/08/31 02:22:55
@@ -36,7 +36,7 @@
 TARGET = $(PACKAGE)
 SUBDIRS = libneon @SUBDIRS@
 OBJECTS = src/cadaver.o src/common.o src/commands.o src/ls.o	\
-	 src/cmdline.o src/options.o src/utils.o src/edit.o
+	 src/cmdline.o src/options.o src/utils.o src/edit.o src/search.o
 LIBOBJS = lib/basename.o lib/dirname.o lib/rpmatch.o lib/yesno.o	\
 	lib/glob.o lib/getpass.o lib/tempname.o lib/mkstemp.o \
 	@LIBOBJS@
@@ -89,6 +89,7 @@
 src/edit.o: src/edit.c src/cadaver.h src/options.h src/common.h
 src/common.o: src/common.c src/common.h config.h
 src/ls.o: src/ls.c src/commands.h src/cadaver.h config.h
+src/search.o: src/search.c src/commands.h src/cadaver.h config.h
 src/utils.o: src/utils.c src/utils.h src/cadaver.h config.h
 lib/netrc.o: lib/netrc.c lib/netrc.h config.h
 lib/getpass.o: lib/getpass.c lib/getpass.h config.h
Index: src/cadaver.c
===================================================================
RCS file: /home/cvs/cadaver/src/cadaver.c,v
retrieving revision 1.142
diff -u -r1.142 cadaver.c
--- src/cadaver.c	2002/08/30 22:49:47	1.142
+++ src/cadaver.c	2002/08/31 02:22:57
@@ -793,8 +793,10 @@
     set_option(opt_namespace, ne_strdup(DEFAULT_NAMESPACE));
     set_bool_option(opt_overwrite, 1);
     set_bool_option(opt_quiet, 1);
+    set_bool_option(opt_searchall, 1);
     lockdepth = NE_DEPTH_INFINITE;
     lockscope = ne_lockscope_exclusive;
+    searchdepth = NE_DEPTH_INFINITE;
 
     /* This is what Markus Kahn says we should do. */
     if ((tmp = getenv("LC_ALL")) ||
Index: src/cadaver.h
===================================================================
RCS file: /home/cvs/cadaver/src/cadaver.h,v
retrieving revision 1.36
diff -u -r1.36 cadaver.h
--- src/cadaver.h	2002/03/07 19:59:10	1.36
+++ src/cadaver.h	2002/08/31 02:22:57
@@ -57,6 +57,7 @@
     cmd_echo, cmd_set, cmd_unset, cmd_rmcol, cmd_lock, cmd_unlock,
     cmd_steal, cmd_discover, cmd_showlocks, cmd_propedit, cmd_propnames,
     cmd_propget, cmd_propset, cmd_propdel, cmd_chexec, cmd_edit, cmd_logout,
+    cmd_search,
     cmd_unknown
 /* DON'T FORGET TO ADD A NEW COMMAND ALIAS WHEN YOU ADD A NEW COMMAND */
 };
@@ -137,6 +138,8 @@
 
 void execute_ls(const char *remote);
 void execute_edit(const char *remote);
+
+void execute_search(int count, const char **args);
 
 void free_resource(struct resource *res);
 void free_resource_list(struct resource *res);
Index: src/commands.c
===================================================================
RCS file: /home/cvs/cadaver/src/commands.c,v
retrieving revision 1.115
diff -u -r1.115 commands.c
--- src/commands.c	2002/08/08 22:29:55	1.115
+++ src/commands.c	2002/08/31 02:23:00
@@ -92,6 +92,7 @@
     C(less), C(cat), C(lpwd), C(lcd), C(lls), C(echo), C(quit),
     C(mget), C(mput), C(rmcol), C(lock), C(unlock), C(discover), C(steal),
     C(chexec), C(showlocks), C(version), C(propget), C(propset), C(propdel),
+    C(search),
 #if 0
 C(propedit), 
 #endif
@@ -1304,6 +1305,23 @@
       N_("propset res propname value"),
       N_("Set property on resource") },
 
+    { cmd_search, true, 1, CMD_VARY, parmscope_remote, TV(execute_search),
+      N_("search query"), 
+      N_("DASL Search resource in current directory" EOL EOL
+	 " Examples : " EOL
+	 "     - search content length is smaller than 100" EOL
+	 "         search getcontentlength < 100" EOL 
+         "     - search author is Sung or Orton" EOL
+         "         search author = Sung or author = Orton" EOL EOL
+	 " Available operations and keywords :" EOL
+	 "     - and, or , (, ), =, <, >, <=, >=, like" EOL EOL
+	 " Set variables (use set command):" EOL
+	 "     - searchdepth : Depth for search" EOL
+         "     - searchorder : Ascending Order property" EOL 
+	 "       ex: set searchorder author" EOL
+         "           set searchorder \"author getcontenttype\"" EOL
+	 "     - searchdorder : Descneding order property" EOL ) },
+    
     { cmd_set, false, 0, 2, parmscope_none, T2(execute_set), 
       N_("set [option] [value]"), N_("Set an option, or display options") },
     { cmd_open, false, 1, 1, parmscope_none, T1(open_connection), 
Index: src/ls.c
===================================================================
RCS file: /home/cvs/cadaver/src/ls.c,v
retrieving revision 1.26
diff -u -r1.26 ls.c
--- src/ls.c	2002/03/07 19:59:10	1.26
+++ src/ls.c	2002/08/31 02:23:01
@@ -41,8 +41,6 @@
 #include "cadaver.h"
 #include "basename.h"
 
-static time_t current_time;
-
 struct fetch_context {
     struct resource **list;
     const char *target; /* Request-URI of the PROPFIND */
@@ -67,42 +65,6 @@
     { NULL }
 };
 
-static char *format_time(time_t when)
-{
-    const char *fmt;
-    static char ret[256];
-    struct tm *local;
-
-    if (when == (time_t)-1) {
-	/* Happens on lock-null resources */
-	return "  (unknown) ";
-    }
-
-    /* from GNU fileutils... this section is 
-     *  Copyright (C) 85, 88, 90, 91, 1995-1999 Free Software Foundation, Inc.
-     */
-    if (current_time > when + 6L * 30L * 24L * 60L * 60L	/* Old. */
-	|| current_time < when - 60L * 60L) {
-	/* The file is fairly old or in the future.
-	   POSIX says the cutoff is 6 months old;
-	   approximate this by 6*30 days.
-	   Allow a 1 hour slop factor for what is considered "the future",
-	   to allow for NFS server/client clock disagreement.
-	   Show the year instead of the time of day.  */
-	fmt = "%b %e  %Y";
-    } else {
-	fmt = "%b %e %H:%M";
-    }
-    /* end FSF copyrighted section. */
-    local = localtime(&when);
-    if (local != NULL) {
-	if (strftime(ret, 256, fmt, local)) {
-	    return ret;
-	}
-    }
-    return "???";
-}
-
 static int compare_resource(const struct resource *r1, 
 			    const struct resource *r2)
 {
@@ -181,7 +143,6 @@
 	    output(o_finish, _("collection is empty.\n"));
 	} else {
 	    out_success();
-	    current_time = time(NULL);
 	    for (current = reslist; current!=NULL; current = next) {
 		next = current->next;
 		if (strlen(current->uri) > strlen(real_remote)) {
Index: src/options.c
===================================================================
RCS file: /home/cvs/cadaver/src/options.c,v
retrieving revision 1.37
diff -u -r1.37 options.c
--- src/options.c	2002/08/25 23:13:39	1.37
+++ src/options.c	2002/08/31 02:23:01
@@ -52,13 +52,20 @@
 static void unset_lockdepth(const char *new);
 static void disp_lockdepth(void);
 
+static void set_searchdepth(const char *new);
+static void unset_searchdepth(const char *new);
+static void disp_searchdepth(void);
+
 /* Option holders */
 
-static int enable_expect, presume_utf8, overwrite, quiet;
+static int enable_expect, presume_utf8, overwrite, quiet, searchall;
 
 enum ne_lock_scope lockscope;
 int lockdepth;
 
+/* search option global */
+int searchdepth;
+
 static struct {
     const char *name;
     enum option_id id;
@@ -84,6 +91,7 @@
     B(expect100, &enable_expect, "Enable use of 'Expect: 100-continue' header"),
     B(utf8, &presume_utf8, "Presume filenames etc are UTF-8 encoded"),
     B(quiet, &quiet, "Whether to display connection status messages"),
+    B(searchall, &searchall, "Whether to search and display all props including dead props"),
 #undef B
 #define S(x,h) { #x, opt_##x, NULL, opt_string, NULL, NULL, NULL, h, NULL }
     S(lockowner, "Lock owner URI"),
@@ -94,6 +102,8 @@
     S(namespace, "Namespace to use for propset/propget commands."),
     S(pager, "Command to run for less/more commands."),
     S(proxy, "Hostname of proxy server"),
+    S(searchorder,  "Search ascending props options"),
+    S(searchdorder, "Search descending props options"),
     { "proxy-port", opt_proxy_port, NULL, opt_string, NULL, NULL, NULL,
       "Port to use on proxy server", NULL },
 #undef S
@@ -110,6 +120,13 @@
       set_lockdepth, unset_lockdepth, disp_lockdepth, "Lock depth options",
       "The lockdepth value must be 0 or infinity."
     },
+
+    /* Several options for search */
+    { "searchdepth", opt_searchdepth, NULL, opt_handled,
+      set_searchdepth, unset_searchdepth, disp_searchdepth, "Search depth options",
+      "The searchdepth value must be 0, 1 or infinity."
+    },
+
     { NULL, 0 }
 };
 
@@ -138,18 +155,18 @@
 	int *val = (int *)options[n].holder;
 	switch (options[n].type) {
 	case opt_bool:
-	    printf(" %10s: %s\n", options[n].name, *val?"on":"off");
+	    printf(" %15s: %s\n", options[n].name, *val?"on":"off");
 	    break;
 	case opt_string:
 	    if (options[n].holder == NULL) {
-		printf(" %10s: unset\n", options[n].name);
+		printf(" %15s: unset\n", options[n].name);
 	    } else {
-		printf(" %10s: %s\n", options[n].name, 
+		printf(" %15s: %s\n", options[n].name, 
 		       (char *)options[n].holder);
 	    }
 	    break;
 	case opt_handled:
-	    printf(" %10s: ", options[n].name);
+	    printf(" %15s: ", options[n].name);
 	    (*options[n].display)();
 	    printf("\n");
 	    break;
@@ -263,6 +280,34 @@
 	printf("infinite");
     else
 	printf("illegal value");
+}
+
+
+static void set_searchdepth(const char *set)
+{
+    if (strcmp(set, "0") == 0 ||
+	strcasecmp(set, "zero") == 0)
+	searchdepth = NE_DEPTH_ZERO;
+    else if (strcasecmp(set, "1") == 0 ||
+	     strcasecmp(set, "one") == 0)
+	searchdepth = NE_DEPTH_ONE;
+    else
+	searchdepth = NE_DEPTH_INFINITE;
+}
+
+static void unset_searchdepth(const char *s)
+{
+    searchdepth = NE_DEPTH_INFINITE;
+}
+
+static void disp_searchdepth(void)
+{
+    if (searchdepth == NE_DEPTH_ZERO)
+	printf("0");
+    else if (searchdepth == NE_DEPTH_ONE)
+	printf("1");
+    else
+	printf("infinity");
 }
 
 void execute_set(const char *opt, const char *newv)
Index: src/options.h
===================================================================
RCS file: /home/cvs/cadaver/src/options.h,v
retrieving revision 1.13
diff -u -r1.13 options.h
--- src/options.h	2002/03/14 22:10:44	1.13
+++ src/options.h	2002/08/31 02:23:01
@@ -39,10 +39,16 @@
     opt_lockstore,
     opt_lockdepth,
     opt_lockscope,
-    opt_pager
+    opt_pager,
+   
+    opt_searchdepth,
+    opt_searchorder,
+    opt_searchdorder,
+    opt_searchall,
 };
 
 extern int lockdepth; /* current lock depth setting. */
+extern int searchdepth; /* current search depth setting. */
 extern enum ne_lock_scope lockscope; /* current lock scope setting. */
 
 void execute_set( const char *opt, const char * );
Index: src/search.c
===================================================================
RCS file: search.c
diff -N search.c
--- /dev/null	Tue May  5 13:32:27 1998
+++ search.c	Fri Aug 30 19:23:05 2002
@@ -0,0 +1,1562 @@
+/* 
+   'search' for cadaver
+   Copyright (C) 2002, GRASE Lab, UCSC <grase@mcse.ucsc.edu>, 
+   except where otherwise indicated.
+                                                                     
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 2 of the License, or
+   (at your option) any later version.
+  
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+  
+   You should have received a copy of the GNU General Public License
+   along with this program; if not, write to the Free Software
+   Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+*/
+
+#include "config.h"
+
+#ifdef HAVE_STDLIB_H
+#include <stdlib.h>
+#endif
+
+#ifdef HAVE_STRING_H
+#include <string.h>
+#endif
+
+#include <time.h>
+
+#include <ne_basic.h>
+#include <ne_request.h>
+#include <ne_props.h>
+#include <ne_uri.h>
+#include <ne_alloc.h>
+#include <ne_dates.h>
+
+#include "i18n.h"
+#include "commands.h"
+#include "cadaver.h"
+#include "basename.h"
+#include "options.h"
+
+/* From ne_props.c
+ * Need to move util.c or something? */
+#define NSPACE(x) ((x) ? (x) : "")
+
+/*
+ * The Macro definations that 'wordtype' might be.
+ */
+#define QUOT            101	/*''' */
+#define COMMA           102	/*',' */
+#define LPAR            103	/*'(' */
+#define RPAR            104	/*')' */
+#define EQ              105	/*'=' */
+#define LE              106	/*'<=' */
+#define LT              107	/*'<' */
+#define GE              108	/*'>=' */
+#define GT              109	/*'>' */
+#define NEQ             110	/*'<>' */
+#define IDEN            111	/*Integer */
+#define INTEGER         112	/*identifier */
+#define UNKNOWN         113	/*unknown charactor */
+#define ENDBUF          114	/*End of the buffer */
+
+#define WORDLEN 256		/*Max length of a identifier(token) in the search command */
+
+/* Dead prop */
+typedef struct dead_prop
+{
+    char *name;
+    char *nspace;
+    char *value;
+    struct dead_prop *next;
+}
+dead_prop;
+
+typedef struct search_res
+{
+    char *href;
+    /* live props */
+    char *creationdate;
+    char *displayname;
+    char *getcontentlanguage;
+    char *getcontentlength;
+    char *getcontenttype;
+    char *getetag;
+    char *getlastmodified;
+    char *lockdiscovery;
+    char *resourcetype;
+    char *source;
+    char *supportedlock;
+    char *collection;
+
+    dead_prop *root;
+    dead_prop *curr;
+    int dead_prop_num;
+
+    struct search_res *next;
+}
+search_res;
+
+/* Search XML parser context */
+typedef struct
+{
+    search_res *root;
+    search_res *curr;
+    int result_num;
+    int start_prop;
+    int err_code;
+}
+search_ctx;
+
+enum
+{
+    ELEM_multistatus = NE_ELM_207_UNUSED,
+    ELEM_response,
+    ELEM_href,
+    ELEM_prop,
+    ELEM_propstat,
+    ELEM_status,
+    ELEM_responsedescription,
+
+    /* props from RFC 2518 , 23 Appendices 23.1 */
+    ELEM_creationdate,
+    ELEM_displayname,
+    ELEM_getcontentlanguage,
+    ELEM_getcontentlength,
+    ELEM_getcontenttype,
+    ELEM_getetag,
+    ELEM_getlastmodified,
+    ELEM_lockdiscovery,
+    ELEM_resourcetype,
+    ELEM_source,
+    ELEM_supportedlock,
+    ELEM_collection,
+
+    ELEM_ignore,
+};
+
+static const struct ne_xml_elm search_elements[] = {
+    {"DAV:", "multistatus", ELEM_multistatus, 0},
+    {"DAV:", "response", ELEM_response, 0},
+    {"DAV:", "responsedescription", ELEM_responsedescription,
+     NE_XML_CDATA},
+    {"DAV:", "href", ELEM_href, NE_XML_CDATA},
+    {"DAV:", "propstat", ELEM_propstat, 0},
+    {"DAV:", "prop", ELEM_prop, 0},
+    {"DAV:", "status", ELEM_status, NE_XML_CDATA},
+
+    /* Live props */
+    {"DAV:", "creationdate", ELEM_creationdate, NE_XML_CDATA},
+    {"DAV:", "displayname", ELEM_displayname, NE_XML_CDATA},
+    {"DAV:", "getcontentlanguage", ELEM_getcontentlanguage, NE_XML_CDATA},
+    {"DAV:", "getcontentlength", ELEM_getcontentlength, NE_XML_CDATA},
+    {"DAV:", "getcontenttype", ELEM_getcontenttype, NE_XML_CDATA},
+    {"DAV:", "getetag", ELEM_getetag, NE_XML_CDATA},
+    {"DAV:", "getlastmodified", ELEM_getlastmodified, NE_XML_CDATA},
+    {"DAV:", "lockdiscovery", ELEM_lockdiscovery, NE_XML_CDATA},
+    {"DAV:", "resourcetype", ELEM_resourcetype, NE_XML_CDATA},
+    {"DAV:", "source", ELEM_source, NE_XML_CDATA},
+    {"DAV:", "supportedlock", ELEM_supportedlock, NE_XML_CDATA},
+    {"DAV:", "collection", ELEM_collection, NE_XML_CDATA},
+
+    /* Ignore it for now */
+    {"DAV:", "lockentry", ELEM_ignore, NE_XML_CDATA},
+    {"DAV:", "lockscope", ELEM_ignore, NE_XML_CDATA},
+    {"DAV:", "locktype", ELEM_ignore, NE_XML_CDATA},
+    {"DAV:", "exclusive", ELEM_ignore, NE_XML_CDATA},
+    {"DAV:", "shared", ELEM_ignore, NE_XML_CDATA},
+    {"DAV:", "read", ELEM_ignore, NE_XML_CDATA},
+    {"DAV:", "write", ELEM_ignore, NE_XML_CDATA},
+
+    /* It deals all unknown elements */
+    {"", "", NE_ELM_unknown, NE_XML_COLLECT},
+    {NULL}
+};
+
+/* 
+ * Basic search parser functions
+ * return NE_OK or error_no
+ * basic_search must be allcated before the function call.
+ */
+static int search_select_gen(const ne_propname * props,
+			     ne_buffer * basic_search);
+static int search_from_gen(const char *href, const int depth,
+			   ne_buffer * basic_search);
+static int search_where_gen(const char *query, ne_buffer * basic_search);
+static int search_orderby_gen(const ne_propname * asc,
+			      const ne_propname * des,
+			      ne_buffer * basic_search);
+static int search_limit_gen(const char *limit, ne_buffer * basic_search);
+
+/*
+ * Static functions for where condition parser
+ */
+static int read_aword(char **string_parsed, char *word_fetched);
+static int first_word_equal(const char *string_parsed,
+			    const char *word_to_compare);
+static int match_fetch(char **string_parsed, const char *str_expected);
+static int search_condition(char **string_parsed, ne_buffer * result_buf);
+static int boolean_term(char **strparsed, ne_buffer * result_buf);
+static int boolean_factor(char **string_parsed, ne_buffer * result_buf);
+static int boolean_primary(char **string_parsed, ne_buffer * result_buf);
+static int operator_translate(const char *operator, char *XML_operator);
+static int predicate(char **string_parsed, ne_buffer * result_buf);
+static int contains_predicate(char **string_parsed, ne_buffer * result_buf);
+static int quoted_string(char **string_parsed, ne_buffer * result_buf);
+static int comparison_value(char **string_parsed, ne_buffer * result_buf);
+static int word_string(char **string_parsed, ne_buffer * result_buf);
+
+/* We do not validate at this piint */
+static int validate_search_elements(void *userdata,
+				    ne_xml_elmid parent, ne_xml_elmid child)
+{
+    return NE_XML_VALID;
+}
+
+/* Set xml parser error */
+static int set_xml_error(search_ctx * sctx, const char *format, ...)
+{
+    sctx->err_code = NE_ERROR;
+    ne_set_error(session, format);
+}
+
+static int start_element(void *userdata, const struct ne_xml_elm *elm,
+			 const char **atts)
+{
+    search_ctx *sctx = (search_ctx *) userdata;
+
+    /* Error occured, ignore remain part */
+    if (sctx->err_code != NE_OK)
+	return sctx->err_code;
+
+    switch (elm->id) {
+    case ELEM_ignore:
+	break;
+
+    case ELEM_response:	/* Start of new response */
+	sctx->curr = ne_calloc(sizeof(search_res));
+	sctx->result_num++;
+	break;
+
+    case ELEM_href:		/* href */
+	break;
+
+    case ELEM_propstat:	/* expecting props */
+	break;
+
+    case ELEM_prop:		/* Start of prop */
+	if (sctx->curr == NULL) {
+	    set_xml_error(sctx, "XML : <%s> is in the wrong place",
+			  elm->name);
+	    break;
+	}
+	sctx->start_prop = 1;
+	break;
+    default:
+	if (get_bool_option(opt_searchall) &&	/* check searchall option */
+	    sctx->curr && sctx->start_prop == 1) {	/* It's prop */
+	    search_res *res = sctx->curr;
+	    res->dead_prop_num++;
+	    res->curr = (dead_prop *) ne_calloc(sizeof(dead_prop));
+	    res->curr->name = ne_strdup(elm->name);
+	    res->curr->nspace = ne_strdup(elm->nspace);
+	}
+	break;
+    }
+
+    return NE_XML_VALID;
+}
+
+#define SEARCH_CP_ELEM(sctx, curr, name, desc, src) \
+do { \
+      if ((curr) == NULL) \
+         set_xml_error((sctx),  "XML : </%s> is in the wrong place", (name));\
+      else \
+         (desc) = ne_strdup(src);\
+} while (0)
+
+
+static int end_element(void *userdata, const struct ne_xml_elm *elm,
+		       const char *cdata)
+{
+    search_ctx *sctx = (search_ctx *) userdata;
+
+    /* Error occured, ignore remain part */
+    if (sctx->err_code != NE_OK)
+	return sctx->err_code;
+
+    switch (elm->id) {
+    case ELEM_ignore:
+	break;
+
+    case ELEM_response:	/* End of new response */
+	/* Nothing to add */
+	if (sctx->curr == NULL) {
+	    set_xml_error(sctx, "XML : </%s> is in the wrong place",
+			  elm->name);
+	    break;
+	}
+
+	/* No HREF */
+	if (sctx->curr->href == NULL) {
+	    set_xml_error(sctx, "XML : No href info in the <%s>...</%s>",
+			  elm->name, elm->name);
+	    break;
+	}
+	/* make link */
+	sctx->curr->next = sctx->root;
+	sctx->root = sctx->curr;
+	sctx->curr = NULL;
+	break;
+
+    case ELEM_href:		/* href */
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name, sctx->curr->href, cdata);
+	break;
+
+	/* live props */
+    case ELEM_creationdate:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->creationdate, cdata);
+	break;
+
+    case ELEM_displayname:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->displayname, cdata);
+	break;
+
+    case ELEM_getcontentlanguage:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->getcontentlanguage, cdata);
+	break;
+
+    case ELEM_getcontentlength:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->getcontentlength, cdata);
+	break;
+
+    case ELEM_getcontenttype:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->getcontenttype, cdata);
+	break;
+
+    case ELEM_getetag:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->getetag, cdata);
+	break;
+
+    case ELEM_getlastmodified:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->getlastmodified, cdata);
+	break;
+
+    case ELEM_lockdiscovery:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->lockdiscovery, cdata);
+	break;
+
+    case ELEM_resourcetype:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->resourcetype, cdata);
+	break;
+
+    case ELEM_source:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->source, cdata);
+	break;
+
+    case ELEM_supportedlock:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->supportedlock, cdata);
+	break;
+
+    case ELEM_collection:
+	SEARCH_CP_ELEM(sctx, sctx->curr, elm->name,
+		       sctx->curr->collection, cdata);
+	break;
+
+    case ELEM_propstat:	/* expecting props */
+	break;
+
+    case ELEM_prop:		/* Start of prop */
+	if (sctx->curr == NULL)
+	    set_xml_error(sctx, "XML : </%s> is in the wrong place",
+			  elm->name);
+	else			/* stop to props */
+	    sctx->start_prop = 0;
+	break;
+
+    default:
+	if (get_bool_option(opt_searchall) &&	/* check searchall option */
+	    sctx->curr && sctx->start_prop == 1) {	/* It's dead prop */
+	    search_res *res = sctx->curr;
+	    res->curr->value = ne_strdup(cdata);
+	    res->curr->next = res->root;
+	    res->root = res->curr;
+	    res->curr = NULL;
+	}
+	break;
+    }
+
+    return NE_XML_VALID;
+}
+
+/* displays search results */
+static int display_results(search_ctx * sctx)
+{
+    search_res *res;
+    dead_prop *dprop;
+
+    if (sctx->err_code) {
+	return sctx->err_code;
+    }
+
+    printf("Found %d results\n\n", sctx->result_num);
+    for (res = sctx->root; res; res = res->next) {
+	long modtime = res->getlastmodified ?
+	    ne_httpdate_parse(res->getlastmodified) : 0;
+	int size = res->getcontentlength ? atol(res->getcontentlength) : 0;
+	char exec_char = ' ';
+
+	printf("%-40s%c %10d  %s <%.10s>\n", res->href, exec_char,
+	       size, format_time(modtime), res->getcontenttype);
+
+	for (dprop = res->root;
+	     get_bool_option(opt_searchall) && dprop; dprop = dprop->next)
+	    printf("\t-  %s:%s = %s\n",	/* better way to show ? */
+		   dprop->nspace, dprop->name, dprop->value);
+    }
+
+    return sctx->err_code;
+}
+
+static void search_ctx_destroy(search_ctx * sctx)
+{
+    search_res *res, *res_free;
+    dead_prop *dprop, *dprop_free;
+
+    for (res = sctx->root; res;) {
+	NE_FREE(res->href);
+	/* live props */
+	NE_FREE(res->creationdate);
+	NE_FREE(res->displayname);
+	NE_FREE(res->getcontentlanguage);
+	NE_FREE(res->getcontentlength);
+	NE_FREE(res->getcontenttype);
+	NE_FREE(res->getetag);
+	NE_FREE(res->getlastmodified);
+	NE_FREE(res->lockdiscovery);
+	NE_FREE(res->resourcetype);
+	NE_FREE(res->source);
+	NE_FREE(res->supportedlock);
+	NE_FREE(res->collection);
+
+	for (dprop = res->root; dprop;) {
+	    NE_FREE(dprop->nspace);
+	    NE_FREE(dprop->name);
+	    NE_FREE(dprop->value);
+
+	    dprop_free = dprop;
+	    dprop = dprop->next;
+	    NE_FREE(dprop_free);
+	}
+	res_free = res;
+	res = res->next;
+	NE_FREE(res_free);
+    }
+}
+
+/* create propname struct from searchorder setting */
+static ne_propname *order_props_create(const char *str)
+{
+    int n;
+    char *buf;
+    char *tok;
+    char *delm = " \t\n\r";
+    int num_props = 0;
+    ne_propname *props;
+
+    /* No props */
+    if (str == NULL)
+	return NULL;
+
+    buf = ne_strdup(str);
+    if (strtok(buf, delm)) {
+	num_props = 1;
+	while (strtok(NULL, delm))
+	    num_props++;
+    }
+
+    /* No props */
+    if (num_props == 0)
+	return NULL;
+
+    /* One more for last NULL */
+    props = ne_calloc(sizeof(ne_propname) * (num_props + 1));
+
+    /* Set first token */
+    NE_FREE(buf);
+    buf = ne_strdup(str);
+    tok = strtok(buf, delm);
+
+    /* Other idea? */
+    props[0].nspace = ne_strdup("DAV:");
+    props[0].name = ne_strdup(tok);
+
+    for (n = 1; (tok = strtok(NULL, delm)); n++) {
+	props[n].nspace = ne_strdup("DAV:");
+	props[n].name = ne_strdup(tok);
+    }
+
+    NE_FREE(buf);
+    return props;
+}
+
+static void order_props_destroy(ne_propname * props)
+{
+    int n;
+
+    if (props == NULL)
+	return;
+
+    for (n = 0; props[n].name != NULL; n++) {
+	NE_FREE((char *) props[n].name);
+	NE_FREE((char *) props[n].nspace);
+    }
+
+    NE_FREE(props);
+}
+
+/* Run search and get the data to sctx */
+static int run_search(ne_session * sess, const char *uri,
+		      int depth, ne_buffer * query, search_ctx * sctx)
+{
+    int ret;
+    ne_request *req;
+    ne_buffer *basic_search = ne_buffer_create();
+    ne_xml_parser *search_parser;
+    const char *searchorder = (const char *) get_option(opt_searchorder);
+    const char *searchdorder = (const char *) get_option(opt_searchdorder);
+    ne_propname *asc = order_props_create(searchorder);
+    ne_propname *des = order_props_create(searchdorder);
+
+    /* create/prep the request */
+    if ((req = ne_request_create(sess, "SEARCH", uri)) == NULL)
+	return NE_ERROR;
+
+    /* Create the request body */
+    ne_buffer_zappend(basic_search,
+		      "<?xml version=\"1.0\" encoding=\"utf-8\" ?>" EOL
+		      "<D:searchrequest xmlns:D=\"DAV:\"><D:basicsearch>"
+		      EOL);
+
+    if (search_select_gen(NULL, basic_search) != NE_OK)
+	return NE_ERROR;
+
+    if (search_from_gen(uri, depth, basic_search) != NE_OK)
+	return NE_ERROR;
+
+    if (search_where_gen(query->data, basic_search) != NE_OK)
+	return NE_ERROR;
+
+    if (search_orderby_gen(asc, des, basic_search) != NE_OK)
+	return NE_ERROR;
+
+    ne_buffer_zappend(basic_search, "</D:basicsearch></D:searchrequest>" EOL);
+
+    ne_set_request_body_buffer(req, basic_search->data,
+			       ne_buffer_size(basic_search));
+
+    /* Plug our XML parser */
+    search_parser = ne_xml_create();
+    ne_xml_push_handler(search_parser, search_elements,
+			validate_search_elements, start_element, end_element,
+			sctx);
+
+    ne_add_request_header(req, "Content-Type", NE_XML_MEDIA_TYPE);
+    ne_add_depth_header(req, depth);
+
+    ne_add_response_body_reader(req, ne_accept_207, ne_xml_parse_v,
+				search_parser);
+
+    /* run the request, see what comes back. */
+    if ((ret = ne_request_dispatch(req)) != NE_OK)
+	return ret;
+
+    /* Check Errors from XML parser */
+    if (sctx->err_code != NE_OK)
+	return NE_ERROR;
+
+    /* Get response code */
+    if (ne_get_status(req)->code != 207)
+	return NE_ERROR;
+
+    /* destroy request, parse, and etc */
+    order_props_destroy(asc);
+    order_props_destroy(des);
+
+    ne_buffer_destroy(basic_search);
+    ne_request_destroy(req);
+    ne_xml_destroy(search_parser);
+
+    return NE_OK;
+}
+
+/* Main execute routine for search */
+void execute_search(int count, const char **args)
+{
+    int ret;
+    const char **pnt;
+    search_ctx *sctx = ne_calloc(sizeof(search_ctx));
+    ne_buffer *query = ne_buffer_create();
+
+    /* default is success */
+    sctx->err_code = NE_OK;
+
+    for (pnt = args; *pnt != NULL; pnt++) {
+	/* Need quota */
+	if (strchr(*pnt, ' ') || strchr(*pnt, '\t'))
+	    ne_buffer_concat(query, "'", *pnt, "' ", NULL);
+	else
+	    ne_buffer_concat(query, *pnt, " ", NULL);
+    }
+
+    printf(_("Using query: "));
+    printf("%s, ", query->data);
+
+    /* Run search and get data to sctx */
+    ret = run_search(session, path, searchdepth, query, sctx);
+    if (ret == NE_OK) {
+	display_results(sctx);
+    }
+
+    out_result(ret);
+
+    search_ctx_destroy(sctx);
+    ne_buffer_destroy(query);
+}
+
+/* Generate select part of the query. allprop if the props arg is NULL. */
+static int search_select_gen(const ne_propname * props,
+			     ne_buffer * basic_search)
+{
+    int n;
+
+    if (!basic_search) {
+	ne_set_error(session, "select_gen: no buffer");
+	return NE_ERROR;
+    }
+
+    if (props == NULL) {
+	ne_buffer_zappend(basic_search,
+			  "<D:select><D:allprop/></D:select>" EOL);
+	return NE_OK;
+    }
+
+    ne_buffer_zappend(basic_search, "<D:select><D:prop>" EOL);
+
+    for (n = 0; props[n].name != NULL; n++) {
+	ne_buffer_concat(basic_search, "<", props[n].name, " xmlns=\"",
+			 NSPACE(props[n].nspace), "\"/>" EOL, NULL);
+    }
+
+    ne_buffer_zappend(basic_search, "</D:prop></D:select>" EOL);
+
+    return NE_OK;
+}
+
+static int search_from_gen(const char *href, const int depth,
+			   ne_buffer * basic_search)
+{
+    const char *depth_str;
+
+    if (!basic_search || !href) {
+	ne_set_error(session, "from_gen: no buffer or no href");
+	return NE_ERROR;
+    }
+
+    switch (depth) {
+    case NE_DEPTH_ONE:
+	depth_str = "1";
+	break;
+    case NE_DEPTH_ZERO:
+	depth_str = "0";
+	break;
+    default:
+	depth_str = "infinity";
+	break;
+    }
+
+    ne_buffer_concat(basic_search, "<D:from><D:scope><D:href>",
+		     href, "</D:href><D:depth>", depth_str,
+		     "</D:depth></D:scope></D:from>" EOL, NULL);
+
+    return NE_OK;
+}
+
+/* Parse a searchquery. It will call the search_condition() function and
+ * check the ending status. If the ending status is not ENDBUF, then there will
+ * be a syntax error.
+ * The parsing result will be saved in 'result_buf'.
+ *
+ * Returns:
+ * NE_OK: success
+ * NE_ERROR: syntax error
+ * */
+static int search_where_gen(const char *condition_str,
+			    ne_buffer * basic_search)
+{
+    char identifier[WORDLEN + 1] = "";
+    char *string_parsed = ne_strdup(condition_str);
+    char *ptr_backup = string_parsed;
+    /* The buffer storing the parsing result of search condition */
+    ne_buffer *result_buf;
+
+    if (!basic_search || !condition_str) {
+	ne_set_error(session, "where_gen: no buffer or no query");
+	return NE_ERROR;
+    }
+
+    result_buf = ne_buffer_create();
+
+    /* Fill boby from <D:where> to </D:where> */
+    ne_buffer_zappend(basic_search, "<D:where>" EOL);
+
+    if (search_condition(&string_parsed, result_buf) == NE_ERROR) {
+	NE_FREE(ptr_backup);
+	ne_buffer_destroy(result_buf);
+	return NE_ERROR;	/*Parsing error */
+    }
+
+    /*The ending of a condition must be an ENDBUF */
+    if (read_aword(&string_parsed, identifier) != ENDBUF) {
+	ne_set_error(session, "Syntax error in the search condition.");
+	NE_FREE(ptr_backup);
+	ne_buffer_destroy(result_buf);
+	return NE_ERROR;
+    }
+
+    /*Append the parsing result of the search condition */
+    ne_buffer_zappend(basic_search, result_buf->data);
+
+    ne_buffer_zappend(basic_search, "</D:where>" EOL);
+
+    NE_FREE(ptr_backup);
+    ne_buffer_destroy(result_buf);
+
+    return NE_OK;
+}				/*End of ne_search_where_gen */
+
+static int search_orderby_gen(const ne_propname * asc,
+			      const ne_propname * des,
+			      ne_buffer * basic_search)
+{
+    int n;
+
+    if (!basic_search) {
+	ne_set_error(session, "orderby_gen: no buffer or no query");
+	return NE_ERROR;
+    }
+
+    /* No order information */
+    if (asc == NULL && des == NULL)
+	return NE_OK;
+
+    ne_buffer_zappend(basic_search, "<D:orderby>" EOL);
+
+    for (n = 0; asc && asc[n].name != NULL; n++) {
+	ne_buffer_zappend(basic_search, "<D:order><D:prop>" EOL);
+	ne_buffer_concat(basic_search, "<", asc[n].name, " xmlns=\"",
+			 NSPACE(asc[n].nspace), "\"/>" EOL, NULL);
+	ne_buffer_zappend(basic_search,
+			  "</D:prop><D:ascending/></D:order>" EOL);
+    }
+
+    for (n = 0; des && des[n].name != NULL; n++) {
+	ne_buffer_zappend(basic_search, "<D:order><D:prop>" EOL);
+	ne_buffer_concat(basic_search, "<", des[n].name, " xmlns=\"",
+			 NSPACE(des[n].nspace), "\"/>" EOL, NULL);
+	ne_buffer_zappend(basic_search,
+			  "</D:prop><D:descending/></D:order>" EOL);
+    }
+
+    ne_buffer_zappend(basic_search, "</D:orderby>" EOL);
+
+    return NE_OK;
+}
+
+
+/* Search command parser
+ *
+ * This is the Search command parser implementation. 
+ * The goal of this code is to translate
+ * dasl search command into XML format.
+ *
+ * The user interface of search command is:
+ *  search resource_URI display_fields condition orderby
+ *
+ *  This code will parse the three parts of the search command, 
+ *  dispolay_fields,
+ *   
+ *  condition and orderby, and translate them into XML, which is a part of a 
+ *  dasl request.
+ * 
+ * The BNF of display_fields, condition and orderby is as below:
+ *
+ *<display_fields> ::= identifier {,identifier}
+ *<orderby clause> ::= identifier ["asc"|"desc"] {,identifier ["asc"|"desc"]}
+ *
+ *<search condition> ::= <boolean term> | <search condition> " or " <boolean term>
+ *<boolean term> ::= <boolean factor> | <boolean term> " and " <boolean factor>
+ *<boolean factor> ::= [not] <boolean primary>
+ *<boolean primary> ::= <predicate>|"("<search condition>")"
+ *<predicate> ::= <comparison predicate>|<like predicate>|<isdefined predicate>
+ *<comparison predicate> ::=  <column_name> <comparaison_op> 
+ *                            ( <number> | <quoted_string> | <wordstring>) 
+ *<like predicate> ::= <column_name> like <quoted_string> -----Formally, 
+ *                      [not] like match_string
+ *<contains predicate> ::= contains (<quoted_string> | <wordstring>)
+ *<column_name>
+ *      ::= identifier
+ *<comparaison_op>
+ *      ::= "=" | "<" | ">" | "<>" | "!=" | "<=" | ">="
+ *<quoted_string> ::= "'" {<any_character>} "'"
+ *<wordstring> ::= <any char except for blank, tab and ')'>{<any_character except for blank, tab and ')'>} 
+ *<match_string>
+ *      ::= "'" { <any_character> | "_" | "%" } "'"
+ *<identifier> ::= <letter> { <letter>|<digital>|<underline>}
+ *<number> ::= [+|-]<digital>{<digital>}
+ */
+
+/* Read a word from the buffer. The word is fetched into 'word_fetched'. 
+ * Every time a word is read, '*string_parsed' is changed to the pointer
+ * pointing to the char just after 'word_fetched'.
+ * 
+ * Parameters:
+ *     string_parsed:  -- *string_parsed points to the buffer. '*string_parsed'
+ *                           will be changed every time a word is fetched.
+ *     word_fetched:   -- The word that is fetched.
+ * Returns:
+ *     ENDBUF , if the end of buffer, '\0', has been reached.
+ *     UNKNOWN, if the word is invalid.
+ *     According word type, if the word is valid.
+ * */
+static int read_aword(char **string_parsed, char *word_fetched)
+{
+    int i = 0;
+    int ifetch = 0;
+    char *stringptr = *string_parsed;
+
+    word_fetched[0] = '\0';
+
+    /*Find the left bound of a word */
+    while (isspace(stringptr[i]))
+	i++;
+
+    if (stringptr[i] == '\0')
+	return ENDBUF;
+
+    /*
+     * Check whether the next word is a identifier. An identifier may
+     * be a keyword or a column name.
+     */
+    if (isalpha(stringptr[i])) {
+	do {
+	    word_fetched[ifetch++] = stringptr[i++];
+	} while (isalnum(stringptr[i]) || (stringptr[i] == '_'));
+
+	*string_parsed = *string_parsed + i;
+	word_fetched[ifetch] = '\0';
+	return IDEN;		/*Identifier */
+    }
+
+    /* Check whether the next word is an integer */
+    /* A problem here: + , or - !!!!! PK */
+    if (stringptr[i] == '+' || stringptr[i] == '-' || isdigit(stringptr[i])) {
+	do {
+	    word_fetched[ifetch++] = stringptr[i++];
+	} while (isdigit(stringptr[i]));
+
+	*string_parsed = *string_parsed + i;
+	word_fetched[ifetch] = '\0';
+	return INTEGER;		/*Identifier */
+    }
+
+    /* Check whether the next word is "<=", ">=", or "<>" */
+    if ((stringptr[i] == '>') && (stringptr[i + 1] == '=')) {
+	strcpy(word_fetched, ">=");
+	*string_parsed = *string_parsed + i + 2;
+	return GE;		/*Indicates >= */
+    }
+
+    if ((stringptr[i] == '<') && (stringptr[i + 1] == '=')) {
+	strcpy(word_fetched, "<=");
+	*string_parsed = *string_parsed + i + 2;
+	return LE;		/*Indicates <= */
+    }
+
+    if ((stringptr[i] == '<') && (stringptr[i + 1] == '>')) {
+	strcpy(word_fetched, "<>");
+	*string_parsed = *string_parsed + i + 2;
+	return NEQ;		/*Indicates <> */
+    }
+
+    /*Check whether the next word is "'", ",", "(", ")", ">",
+     * "<", "=", */
+    switch (stringptr[i]) {
+    case '\'':
+	strcpy(word_fetched, "'");
+	*string_parsed = *string_parsed + i + 1;
+	return QUOT;
+    case ',':
+	strcpy(word_fetched, ",");
+	*string_parsed = *string_parsed + i + 1;
+	return COMMA;
+    case '(':
+	strcpy(word_fetched, "(");
+	*string_parsed = *string_parsed + i + 1;
+	return LPAR;
+    case ')':
+	strcpy(word_fetched, ")");
+	*string_parsed = *string_parsed + i + 1;
+	return RPAR;
+    case '>':
+	strcpy(word_fetched, ">");
+	*string_parsed = *string_parsed + i + 1;
+	return GT;
+    case '<':
+	strcpy(word_fetched, "<");
+	*string_parsed = *string_parsed + i + 1;
+	return LT;
+    case '=':
+	strcpy(word_fetched, "=");
+	*string_parsed = *string_parsed + i + 1;
+	return EQ;
+    default:
+	word_fetched[0] = stringptr[i];
+	word_fetched[1] = '\0';
+	*string_parsed = *string_parsed + i + 1;
+	return UNKNOWN;
+    }				/*End of switch */
+}				/*End of read_aword */
+
+/*Read the first word from the string buffer, string_parsed. If the
+ * first is equal to the 'word_to_compare', return 1, else return 0.
+ * The first word will be saved to 'the_first_word', but the pointer
+ * to the buffer, string_parsed, will not be changed.
+ * Returns:
+ * 1: The first word is equal to the word expected.
+ * 0: otherwise
+ * */
+static int first_word_equal(const char *string_parsed,
+			    const char *word_to_compare)
+{
+    char *string_buffer = NULL;
+    char *ptr_backup = NULL;
+    char first_word[WORDLEN + 1] = "";
+
+    string_buffer = ne_strdup(string_parsed);
+    ptr_backup = string_buffer;
+
+    read_aword(&string_buffer, first_word);
+
+    NE_FREE(ptr_backup);
+
+    if (strcasecmp(first_word, word_to_compare) == 0)
+	return 1;		/*equal */
+    else
+	return 0;		/*Not equal */
+}				/*End of first_word_compare */
+
+/*Read the first word from the string buffer, string_parsed. If the
+ * first is an interger, return 1, else return 0.
+ * The first word will be saved to 'the_first_word', but the pointer
+ * to the buffer, string_parsed, will not be changed.
+ * Returns:
+ * 1: The first word is equal to the word expected.
+ * 0: otherwise
+ * */
+static int first_word_integer(const char *string_parsed)
+{
+    char *string_buffer = NULL;
+    char *ptr_backup = NULL;
+    char first_word[WORDLEN + 1] = "";
+    int ret;
+
+    string_buffer = ne_strdup(string_parsed);
+    ptr_backup = string_buffer;
+
+    if (read_aword(&string_buffer, first_word) == INTEGER)
+	ret = 1;
+    else
+	ret = 0;
+
+    NE_FREE(ptr_backup);
+
+    return ret;
+}				/*End of first_word_compare */
+
+/* The function reads the first word in the string buffer, *string_parsed, and
+ * changes *string_parsed. 
+ * Returns:
+ * NE_ERROR: If the word read is equal to 'str_expected'.
+ * NE_OK: Otherwise
+ * */
+static int match_fetch(char **string_parsed, const char *str_expected)
+{
+    char first_word[WORDLEN + 1] = "";
+
+    read_aword(string_parsed, first_word);
+
+    if (strcmp(first_word, str_expected) == 0)
+	return NE_OK;
+    else
+	return NE_ERROR;
+
+}				/*End of match_fetch */
+
+/*
+ * Parse the search condition
+ * <search condition> ::= <boolean term> | <search condition> or <boolean term>
+ *
+ * The parsing result is saved into result_buf.
+ * '*string_parsed' is changed to the pointer pointing to the position after
+ * <search condition>.
+ * Returns:
+ * NE_OK: success
+ * NE_ERROR: syntax error
+ * */
+static int search_condition(char **string_parsed, ne_buffer * result_buf)
+{
+    char identifier[WORDLEN + 1] = "";
+    /* Indicates whether there is any 'or' in the search condition */
+    int added_or = 0;
+    ne_buffer *term_result;
+
+    ne_buffer_clear(result_buf);
+
+    term_result = ne_buffer_create();
+
+    if (boolean_term(string_parsed, term_result) == NE_ERROR) {
+	ne_buffer_destroy(term_result);	/*Free the buffer */
+	return NE_ERROR;	/*parsing error */
+    }
+
+    if (first_word_equal(*string_parsed, "or") == 1) {
+	added_or = 1;
+	ne_buffer_concat(result_buf, "<D:or>" EOL, term_result->data, NULL);
+    }
+    else
+	ne_buffer_zappend(result_buf, term_result->data);
+
+    /*For or <boolean term> or <boolean term> ... */
+    while (first_word_equal(*string_parsed, "or") == 1) {
+	read_aword(string_parsed, identifier);	/*Read 'or' */
+
+	if (boolean_term(string_parsed, term_result) == NE_ERROR) {
+	    ne_buffer_destroy(term_result);	/*Free the buffer */
+	    return NE_ERROR;	/*Parsing error */
+	}
+	ne_buffer_zappend(result_buf, term_result->data);
+    }				/*End of while */
+
+    if (added_or == 1)
+	ne_buffer_zappend(result_buf, "</D:or>" EOL);
+
+    ne_buffer_destroy(term_result);	/*Free the buffer */
+
+    return NE_OK;		/*success */
+}				/*End of search_condition */
+
+/*
+ * Parse a boolean term.
+ * <boolean term> ::= <boolean factor> | <boolean term> and <boolean factor>
+ *
+ * The parsing result is saved into result_buf.
+ * '*string_parsed' is changed to the pointer pointing to the position after
+ * <boolean term>.
+ *
+ * Returns:
+ * NE_OK: success
+ * NE_ERROR: syntax error
+ */
+static int boolean_term(char **string_parsed, ne_buffer * result_buf)
+{
+    char identifier[WORDLEN + 1] = "";
+    /*Indicates whether there is any 'and' in the boolean term. */
+    int added_and = 0;
+    ne_buffer *factor_result;
+
+    ne_buffer_clear(result_buf);
+    factor_result = ne_buffer_create();
+
+    if (boolean_factor(string_parsed, factor_result) == NE_ERROR) {
+	ne_buffer_destroy(factor_result);
+	return NE_ERROR;	/*parsing error */
+    }
+
+    if (first_word_equal(*string_parsed, "and") == 1) {
+	added_and = 1;
+	ne_buffer_concat(result_buf, "<D:and>" EOL, factor_result->data,
+			 NULL);
+    }
+    else
+	ne_buffer_zappend(result_buf, factor_result->data);
+
+    while (first_word_equal(*string_parsed, "and") == 1) {
+	read_aword(string_parsed, identifier);	/*Read 'and' */
+
+	if (boolean_factor(string_parsed, factor_result) == NE_ERROR) {
+	    ne_buffer_destroy(factor_result);
+	    return NE_ERROR;	/*Parsing error */
+	}
+	ne_buffer_zappend(result_buf, factor_result->data);
+    }				/*End of while */
+
+    if (added_and == 1)
+	ne_buffer_zappend(result_buf, "</D:and>" EOL);
+
+    ne_buffer_destroy(factor_result);
+
+    return NE_OK;		/*success */
+}				/*End of boolean_term */
+
+/*
+ * Parse a boolean factor.
+ * <boolean factor> ::= [not] <boolean primary>
+ *
+ * The parsing result is saved into result_buf.
+ * '*string_parsed' is changed to the pointer pointing to the position after
+ * <boolean factor>.
+ *
+ * Returns:
+ * NE_OK: success
+ * NE_ERROR: syntax error
+ */
+static int boolean_factor(char **string_parsed, ne_buffer * result_buf)
+{
+    char identifier[WORDLEN + 1] = "";
+    int added_not = 0;		/*Indicates whether there is any 'not'
+				   in the search condition */
+    ne_buffer *boolean_primary_result;
+
+    ne_buffer_clear(result_buf);
+    boolean_primary_result = ne_buffer_create();
+
+    if (first_word_equal(*string_parsed, "not") == 1) {
+	read_aword(string_parsed, identifier);	/*Read "not" */
+	ne_buffer_zappend(result_buf, "<D:not>" EOL);
+	added_not = 1;
+    }
+
+    if (boolean_primary(string_parsed, boolean_primary_result) == NE_ERROR) {
+	ne_buffer_destroy(boolean_primary_result);
+	return NE_ERROR;	/*parsing error */
+    }
+
+    ne_buffer_zappend(result_buf, boolean_primary_result->data);
+
+    if (added_not == 1)
+	ne_buffer_zappend(result_buf, "</D:not>" EOL);
+
+    ne_buffer_destroy(boolean_primary_result);
+
+    return NE_OK;		/*success */
+}				/*End of boolean_factor */
+
+/*
+ * Parse a boolean primary.
+ * <boolean primary> ::= <predicate> | "("<search condition>")"
+ *
+ * The parsing result is saved into result_buf.
+ * '*string_parsed' is changed to the pointer pointing to the position after
+ * <boolean primary>.
+ *
+ * Returns:
+ * NE_OK: success
+ * NE_ERROR: syntax error
+ */
+static int boolean_primary(char **string_parsed, ne_buffer * result_buf)
+{
+    char identifier[WORDLEN + 1] = "";
+    /*Indicates whether there is any 'not' in the search condition */
+    int added_not = 0;
+    ne_buffer *sub_result;
+
+    ne_buffer_clear(result_buf);
+    sub_result = ne_buffer_create();
+
+    if (first_word_equal(*string_parsed, "(") == 1) {
+	/*It is the case of "("<search condition>")" */
+
+	read_aword(string_parsed, identifier);	/*Read "(" */
+
+	if (search_condition(string_parsed, sub_result) == NE_ERROR) {
+	    ne_buffer_destroy(sub_result);
+	    return NE_ERROR;	/*Parsing error */
+	}
+
+	/*Read and match ")" */
+	if (match_fetch(string_parsed, ")") == NE_ERROR) {
+	    ne_set_error(session,
+			 "Syntax error: A ')' is expected in the search condition.");
+	    ne_buffer_destroy(sub_result);
+	    return NE_ERROR;	/*parsing error */
+	}
+	ne_buffer_zappend(result_buf, sub_result->data);
+    }
+    else {			/*It is the case of <predicate> */
+	if (predicate(string_parsed, sub_result) == NE_ERROR) {
+	    ne_buffer_destroy(sub_result);
+	    return NE_ERROR;	/*parsing error */
+	}
+	ne_buffer_zappend(result_buf, sub_result->data);
+    }
+
+    ne_buffer_destroy(sub_result);
+
+    return NE_OK;		/*success */
+}				/*End of boolean_primary */
+
+/*
+ *  Translate comparison operator in the search condition to XML form.
+ *  Returns:
+ *    1: if the operator is valid.
+ *    0: if the operator is not valid.
+ */
+static int operator_translate(const char *operator, char *XML_operator)
+{
+    int operator_valid = 1;
+
+    XML_operator[0] = '\0';
+
+    if (strcmp(operator, "=") == 0)
+	strcpy(XML_operator, "eq");
+    else if (strcmp(operator, ">=") == 0)
+	strcpy(XML_operator, "gte");
+    else if (strcmp(operator, "<=") == 0)
+	strcpy(XML_operator, "lte");
+    else if (strcmp(operator, ">") == 0)
+	strcpy(XML_operator, "gt");
+    else if (strcmp(operator, "<") == 0)
+	strcpy(XML_operator, "lt");
+    else if (strcmp(operator, "!=") == 0)
+	strcpy(XML_operator, "not");
+    else if (strcmp(operator, "<>") == 0)
+	strcpy(XML_operator, "not");
+    else if (strcmp(operator, "like") == 0)
+	strcpy(XML_operator, "like");
+    else
+	operator_valid = 0;	/*Invalid operator */
+
+    return operator_valid;
+}				/*End of operator_translate */
+
+/*
+ * Parse a predicate.
+ * <predicate> ::= <comparison predicate> | <like predicate> | <contains predicate>
+ * <comparison predicate> ::=  <column_name> <comparaison_op> ( <number> | <quoted_string> | <wordstring>) 
+ * <like predicate> ::= <column_name> like <match_string>
+ * <contains predicate> ::= contains (<quoted_string> | <wordstring>)
+ * <comparaison_op>
+ *         ::= "=" | "<" | ">" | "<>" | "!=" | "<=" | ">="
+ *   
+ * The parsing result is saved into result_buf.
+ * '*string_parsed' is changed to the pointer pointing to the position after
+ * the predicate.
+ *
+ * Returns:
+ * NE_OK: success
+ * NE_ERROR: syntax error
+ */
+static int predicate(char **string_parsed, ne_buffer * result_buf)
+{
+    char column_name[WORDLEN + 1] = "";
+    char operator[WORDLEN + 1] = "";
+    char XML_operator[WORDLEN + 1] = "";
+    ne_buffer *comparing_value;
+
+    ne_buffer_clear(result_buf);
+
+    if (first_word_equal(*string_parsed, "contains") == 1) {
+	/* It is the case of <contains predicate> */
+	ne_buffer *contains_result;
+	contains_result = ne_buffer_create();
+	if (contains_predicate(string_parsed, contains_result)
+	    == NE_ERROR) {
+	    ne_buffer_destroy(contains_result);
+	    return NE_ERROR;	/*Parsing error */
+	}
+	ne_buffer_zappend(result_buf, contains_result->data);
+	return NE_OK;
+    }
+
+    comparing_value = ne_buffer_create();
+
+    /*Read the column name */
+    if (read_aword(string_parsed, column_name) != IDEN) {
+	ne_set_error(session,
+		     "A column name is expected in the search condition.");
+	ne_buffer_destroy(comparing_value);
+	return NE_ERROR;	/*Parsing error */
+    }
+
+    /*Read the 'operator' */
+    read_aword(string_parsed, operator);
+
+    /*Translate the operator to XML form */
+    if (operator_translate(operator, XML_operator) == 0) {
+	ne_set_error(session,
+		     "Syntax error: Invalid operator in the search condition.");
+	ne_buffer_destroy(comparing_value);
+	return NE_ERROR;
+    }
+
+    if (strcasecmp(operator, "like") == 0) {
+	/*It is the case of like predicate and then parse the match string */
+	if (first_word_equal(*string_parsed, "'")) {
+	    /* For the case of a quoted string */
+	    if (quoted_string(string_parsed, comparing_value) == NE_ERROR) {
+		ne_buffer_destroy(comparing_value);
+		return NE_ERROR;	/*Parsing error */
+	    }
+	}
+	else {
+	    /* For the case of word string. */
+	    if (word_string(string_parsed, comparing_value) == NE_ERROR) {
+		ne_set_error(session,
+			     "Syntax error: A quoted string or a word string is expected in the search condition.");
+		ne_buffer_destroy(comparing_value);
+		return NE_ERROR;	/*Parsing error */
+	    }
+	}
+    }
+    else {
+	/* It is the case of comparison predicate and then 
+	 * parse the comparison value */
+	if (comparison_value(string_parsed, comparing_value) == NE_ERROR) {
+	    ne_buffer_destroy(comparing_value);
+	    return NE_ERROR;	/*parsing error */
+	}
+    }
+
+    ne_buffer_concat(result_buf, "<D:",
+		     XML_operator,
+		     ">" EOL
+		     "<D:prop><D:",
+		     column_name,
+		     "/></D:prop>" EOL
+		     "<D:literal>",
+		     comparing_value->data,
+		     "</D:literal>" EOL "</D:", XML_operator, ">" EOL, NULL);
+
+    ne_buffer_destroy(comparing_value);
+
+    return NE_OK;		/*success */
+}				/*End of predicate */
+
+/*
+ * Parse a contains predicate.
+ * <contains predicate> ::= contains <quoted_string> 
+ *
+ * The parsing result is saved into result_str.
+ * '*string_parsed' is changed to the pointer pointing to the position after
+ * the predicate.
+ *
+ * Returns:
+ * 1: success
+ * 0: syntax error
+ */
+int contains_predicate(char **string_parsed, ne_buffer * result_buf)
+{
+    ne_buffer *contain_string;
+
+    ne_buffer_clear(result_buf);
+    contain_string = ne_buffer_create();
+
+    /*Read 'contains' */
+    if (match_fetch(string_parsed, "contains") == NE_ERROR) {
+	/*The case of <contains predicate> */
+	ne_set_error(session,
+		     "Syntax error: A 'contains' is expected in the search condition.");
+	ne_buffer_destroy(contain_string);
+	return NE_ERROR;
+    }
+
+    /*Now parse the containing string */
+    if (first_word_equal(*string_parsed, "'") == 1) {
+	/*For the case of <quoted string> */
+	if (quoted_string(string_parsed, contain_string) == NE_ERROR) {
+	    ne_buffer_destroy(contain_string);
+	    return NE_ERROR;	/*Parsing error */
+	}
+    }
+    else {
+	/* For the case of <wordstring> */
+	if (word_string(string_parsed, contain_string) == NE_ERROR) {
+	    ne_set_error(session,
+			 "Syntax error: A quoted string or a word string is expected in the search condition.");
+	    ne_buffer_destroy(contain_string);
+	    return NE_ERROR;	/*Parsing error */
+	}
+    }
+
+    ne_buffer_concat(result_buf, "<D:contains> EOL",
+		     contain_string->data, "</D:contains>" EOL, NULL);
+
+    return NE_OK;		/*success */
+}				/*End of contains_predicate */
+
+/*
+ * Parse a quoted string.
+ * <quoted_string> ::= "'" {<any_character>} "'"
+ *
+ * The parsing result is saved into result_buf.
+ * '*string_parsed' is changed to the pointer pointing to the position after
+ * the predicate.
+ *
+ * Returns:
+ * 1: success
+ * 0: syntax error
+ */
+static int quoted_string(char **string_parsed, ne_buffer * result_buf)
+{
+    char identifier[WORDLEN + 1] = "";
+    char previous_char;
+    char current_char;
+    char tmp_str[2] = "";
+    int i = 0;
+
+    ne_buffer_clear(result_buf);
+
+    /*Read a quotation mark */
+    if (match_fetch(string_parsed, "'") == NE_ERROR) {
+	ne_set_error(session,
+		     "Syntax error: A ' is expected in the search condition.");
+	return NE_ERROR;	/*Parsing error */
+    }
+
+    /*To parse {"any_character"}. Considering '\'' case when parsing */
+    current_char = previous_char = (*string_parsed)[0];
+    while ((current_char != '\'') &&
+	   (current_char != '\0') ||
+	   ((current_char == '\'') && (previous_char == '\\'))) {
+	tmp_str[0] = current_char;
+	tmp_str[1] = '\0';
+	ne_buffer_zappend(result_buf, tmp_str);
+	previous_char = current_char;
+	i++;
+	current_char = (*string_parsed)[i];
+    }
+
+    if (current_char != '\'') {	/*There should be a ending ' */
+	ne_set_error(session,
+		     "An ending ' is expected in the search condition.");
+	return NE_ERROR;	/*parsing error */
+    }
+
+    *string_parsed = *string_parsed + i + 1;	/*1 for ending ' */
+
+    return NE_OK;		/*success */
+}				/*End of auoted_string */
+
+/*
+ * Parse a word string.
+ * <wordstring> ::= <any char except for blank, tab and ')'>{<any_character except for blank, tab and ')'>} 
+ *
+ * The parsing result is saved into result_buf.
+ * '*string_parsed' is changed to the pointer pointing to the position after
+ * the predicate.
+ *
+ * Returns:
+ * 1: success
+ * 0: syntax error
+ */
+static int word_string(char **string_parsed, ne_buffer * result_buf)
+{
+    char identifier[WORDLEN + 1] = "";
+    char previous_char;
+    char current_char;
+    char tmp_str[2] = "";
+    int i = 0;
+
+    ne_buffer_clear(result_buf);
+
+    /*Find the left bound of a word */
+    while (isspace((*string_parsed)[i]))
+	i++;
+
+    if ((*string_parsed)[i] == '\0' || (*string_parsed)[i] == ')')	/* Empty string */
+	return NE_ERROR;
+
+    current_char = previous_char = (*string_parsed)[i];
+    while ((current_char != ' ') &&
+	   (current_char != '\t') &&
+	   (current_char != ')') && (current_char != '\0')) {
+	tmp_str[0] = current_char;
+	tmp_str[1] = '\0';
+	ne_buffer_zappend(result_buf, tmp_str);
+	previous_char = current_char;
+	i++;
+	current_char = (*string_parsed)[i];
+    }
+
+    *string_parsed = *string_parsed + i;
+
+    return NE_OK;		/*success */
+}				/*End of auoted_string */
+
+/*
+ * Parse a comparison_value.
+ *<comparison_value>
+ *      ::=  <number> | <quoted_string> | <wordstring>
+ *
+ * The parsing result is saved into result_buf.
+ * '*string_parsed' is changed to the pointer pointing to the position after
+ * the predicate.
+ *
+ * Returns:
+ * NE_OK: success
+ * NE_ERROR: syntax error
+ */
+static int comparison_value(char **string_parsed, ne_buffer * result_buf)
+{
+    char identifier[WORDLEN + 1] = "";
+    ne_buffer *comparing_value;
+
+    ne_buffer_clear(result_buf);
+    comparing_value = ne_buffer_create();
+
+    if (first_word_equal(*string_parsed, "'") == 1) {
+	/*It is the case of quoted string */
+	if (quoted_string(string_parsed, comparing_value) == NE_ERROR) {
+	    ne_buffer_destroy(comparing_value);
+	    return NE_ERROR;	/*Parsing error */
+	}
+	ne_buffer_zappend(result_buf, comparing_value->data);
+    }
+    else {			/* It is the case of <number> or <wordstring> */
+	/*An integer or a word string is expected */
+	if (first_word_integer(*string_parsed) == 1) {
+	    read_aword(string_parsed, identifier);
+	    ne_buffer_zappend(result_buf, identifier);
+	}
+	else {			/* It is the case of a word string */
+	    if (word_string(string_parsed, comparing_value) == NE_ERROR) {
+		ne_set_error(session,
+			     "Syntax error: An integer, quoted string or word string is expected in the search condition.");
+		ne_buffer_destroy(comparing_value);
+		return NE_ERROR;
+	    }
+	}
+	ne_buffer_zappend(result_buf, comparing_value->data);
+    }
+
+    ne_buffer_destroy(comparing_value);
+
+    return NE_OK;		/*success */
+}				/*End of comparison_value */
Index: src/utils.c
===================================================================
RCS file: /home/cvs/cadaver/src/utils.c,v
retrieving revision 1.2
diff -u -r1.2 utils.c
--- src/utils.c	2002/03/07 19:59:10	1.2
+++ src/utils.c	2002/08/31 02:23:05
@@ -21,6 +21,7 @@
 
 #include "config.h"
 
+#include <time.h>
 #include <sys/types.h>
 
 #include <ne_basic.h>
@@ -52,5 +53,44 @@
     }
     free_resource_list(res);
     return ret;
+}
+
+
+char *format_time(time_t when)
+{
+    const char *fmt;
+    static char ret[256];
+    struct tm *local;
+    time_t current_time;
+    
+    if (when == (time_t)-1) {
+	/* Happens on lock-null resources */
+	return "  (unknown) ";
+    }
+
+    /* from GNU fileutils... this section is 
+     *  Copyright (C) 85, 88, 90, 91, 1995-1999 Free Software Foundation, Inc.
+     */
+    current_time = time(NULL);
+    if (current_time > when + 6L * 30L * 24L * 60L * 60L	/* Old. */
+	|| current_time < when - 60L * 60L) {
+	/* The file is fairly old or in the future.
+	   POSIX says the cutoff is 6 months old;
+	   approximate this by 6*30 days.
+	   Allow a 1 hour slop factor for what is considered "the future",
+	   to allow for NFS server/client clock disagreement.
+	   Show the year instead of the time of day.  */
+	fmt = "%b %e  %Y";
+    } else {
+	fmt = "%b %e %H:%M";
+    }
+    /* end FSF copyrighted section. */
+    local = localtime(&when);
+    if (local != NULL) {
+	if (strftime(ret, 256, fmt, local)) {
+	    return ret;
+	}
+    }
+    return "???";
 }
 
Index: src/utils.h
===================================================================
RCS file: /home/cvs/cadaver/src/utils.h,v
retrieving revision 1.1
diff -u -r1.1 utils.h
--- src/utils.h	2002/01/14 20:00:33	1.1
+++ src/utils.h	2002/08/31 02:23:05
@@ -24,3 +24,6 @@
 /* Returns resource type of resource with given URI; where resr_error
  * means "resource not found." */
 enum resource_type getrestype(const char *uri);
+
+/* Returns time to display */
+char *format_time(time_t when);