notmuch-restore: add support for input format 'batch-tag'

This can be enabled with the new --format=batch-tag command line
option to "notmuch restore". The input must consist of lines of the
format:

    +<tag>|-<tag> [...] [--] id:<msg-id>

Each line is interpreted similarly to "notmuch tag" command line
arguments. The delimiter is one or more spaces ' '. Any characters in
<tag> and <search-terms> MAY be hex encoded with %NN where NN is the
hexadecimal value of the character. Any ' ' and '%' characters in
<tag> and <msg-id> MUST be hex encoded (using %20 and %25,
respectively). Any characters that are not part of <tag> or
<search-terms> MUST NOT be hex encoded.

Leading and trailing space ' ' is ignored. Empty lines and lines
beginning with '#' are ignored.

Commit message mainly stolen from Jani's batch tagging commit, to
follow.

notmuch-restore.c

@@ -19,18 +19,21 @@
  */
 
 #include "notmuch-client.h"
+#include "dump-restore-private.h"
+#include "tag-util.h"
+#include "string-util.h"
+
+static regex_t regex;
 
 static int
-tag_message (notmuch_database_t *notmuch, const char *message_id,
+tag_message (unused (void *ctx),
-	     char *file_tags, notmuch_bool_t remove_all,
+	     notmuch_database_t *notmuch,
-	     notmuch_bool_t synchronize_flags)
+	     const char *message_id,
+	     tag_op_list_t *tag_ops,
+	     tag_op_flag_t flags)
 {
     notmuch_status_t status;
-    notmuch_tags_t *db_tags;
-    char *db_tags_str;
     notmuch_message_t *message = NULL;
-    const char *tag;
-    char *next;
     int ret = 0;
 
     status = notmuch_database_find_message (notmuch, message_id, &message);
@@ -44,72 +47,87 @@ tag_message (notmuch_database_t *notmuch, const char *message_id,
 
     /* In order to detect missing messages, this check/optimization is
      * intentionally done *after* first finding the message. */
-    if (! remove_all && (file_tags == NULL || *file_tags == '\0'))
+    if ((flags & TAG_FLAG_REMOVE_ALL) || tag_op_list_size (tag_ops))
-	goto DONE;
+	tag_op_list_apply (message, tag_ops, flags);
 
-    db_tags_str = NULL;
-    for (db_tags = notmuch_message_get_tags (message);
-	 notmuch_tags_valid (db_tags);
-	 notmuch_tags_move_to_next (db_tags)) {
-	tag = notmuch_tags_get (db_tags);
-
-	if (db_tags_str)
-	    db_tags_str = talloc_asprintf_append (db_tags_str, " %s", tag);
-	else
-	    db_tags_str = talloc_strdup (message, tag);
-    }
-
-    if (((file_tags == NULL || *file_tags == '\0') &&
-	 (db_tags_str == NULL || *db_tags_str == '\0')) ||
-	(file_tags && db_tags_str && strcmp (file_tags, db_tags_str) == 0))
-	goto DONE;
-
-    notmuch_message_freeze (message);
-
-    if (remove_all)
-	notmuch_message_remove_all_tags (message);
-
-    next = file_tags;
-    while (next) {
-	tag = strsep (&next, " ");
-	if (*tag == '\0')
-	    continue;
-	status = notmuch_message_add_tag (message, tag);
-	if (status) {
-	    fprintf (stderr, "Error applying tag %s to message %s:\n",
-		     tag, message_id);
-	    fprintf (stderr, "%s\n", notmuch_status_to_string (status));
-	    ret = 1;
-	}
-    }
-
-    notmuch_message_thaw (message);
-
-    if (synchronize_flags)
-	notmuch_message_tags_to_maildir_flags (message);
-
-  DONE:
-    if (message)
     notmuch_message_destroy (message);
 
     return ret;
 }
 
+/* Sup dump output is one line per message. We match a sequence of
+ * non-space characters for the message-id, then one or more
+ * spaces, then a list of space-separated tags as a sequence of
+ * characters within literal '(' and ')'. */
+
+static int
+parse_sup_line (void *ctx, char *line,
+		char **query_str, tag_op_list_t *tag_ops)
+{
+
+    regmatch_t match[3];
+    char *file_tags;
+    int rerr;
+
+    tag_op_list_reset (tag_ops);
+
+    chomp_newline (line);
+
+    /* Silently ignore blank lines */
+    if (line[0] == '\0') {
+	return 1;
+    }
+
+    rerr = xregexec (&regex, line, 3, match, 0);
+    if (rerr == REG_NOMATCH) {
+	fprintf (stderr, "Warning: Ignoring invalid sup format line: %s\n",
+		 line);
+	return 1;
+    }
+
+    *query_str = talloc_strndup (ctx, line + match[1].rm_so,
+				 match[1].rm_eo - match[1].rm_so);
+    file_tags = talloc_strndup (ctx, line + match[2].rm_so,
+				match[2].rm_eo - match[2].rm_so);
+
+    char *tok = file_tags;
+    size_t tok_len = 0;
+
+    tag_op_list_reset (tag_ops);
+
+    while ((tok = strtok_len (tok + tok_len, " ", &tok_len)) != NULL) {
+
+	if (*(tok + tok_len) != '\0') {
+	    *(tok + tok_len) = '\0';
+	    tok_len++;
+	}
+
+	if (tag_op_list_append (ctx, tag_ops, tok, FALSE))
+	    return -1;
+    }
+
+    return 0;
+
+}
+
 int
 notmuch_restore_command (unused (void *ctx), int argc, char *argv[])
 {
     notmuch_config_t *config;
     notmuch_database_t *notmuch;
-    notmuch_bool_t synchronize_flags;
     notmuch_bool_t accumulate = FALSE;
+    tag_op_flag_t flags = 0;
+    tag_op_list_t *tag_ops;
+
     char *input_file_name = NULL;
     FILE *input = stdin;
     char *line = NULL;
     size_t line_size;
     ssize_t line_len;
-    regex_t regex;
+
-    int rerr;
+    int ret = 0;
     int opt_index;
+    int input_format = DUMP_FORMAT_AUTO;
 
     config = notmuch_config_open (ctx, NULL, NULL);
     if (config == NULL)
@@ -119,9 +137,15 @@ notmuch_restore_command (unused (void *ctx), int argc, char *argv[])
 			       NOTMUCH_DATABASE_MODE_READ_WRITE, &notmuch))
 	return 1;
 
-    synchronize_flags = notmuch_config_get_maildir_synchronize_flags (config);
+    if (notmuch_config_get_maildir_synchronize_flags (config))
+	flags |= TAG_FLAG_MAILDIR_SYNC;
 
     notmuch_opt_desc_t options[] = {
+	{ NOTMUCH_OPT_KEYWORD, &input_format, "format", 'f',
+	  (notmuch_keyword_t []){ { "auto", DUMP_FORMAT_AUTO },
+				  { "batch-tag", DUMP_FORMAT_BATCH_TAG },
+				  { "sup", DUMP_FORMAT_SUP },
+				  { 0, 0 } } },
 	{ NOTMUCH_OPT_STRING, &input_file_name, "input", 'i', 0 },
 	{ NOTMUCH_OPT_BOOLEAN,  &accumulate, "accumulate", 'a', 0 },
 	{ 0, 0, 0, 0, 0 }
@@ -134,6 +158,9 @@ notmuch_restore_command (unused (void *ctx), int argc, char *argv[])
 	return 1;
     }
 
+    if (! accumulate)
+	flags |= TAG_FLAG_REMOVE_ALL;
+
     if (input_file_name) {
 	input = fopen (input_file_name, "r");
 	if (input == NULL) {
@@ -149,49 +176,77 @@ notmuch_restore_command (unused (void *ctx), int argc, char *argv[])
 		 argv[opt_index]);
 	return 1;
     }
+    char *p;
 
-    /* Dump output is one line per message. We match a sequence of
+    line_len = getline (&line, &line_size, input);
-     * non-space characters for the message-id, then one or more
+    if (line_len == 0)
-     * spaces, then a list of space-separated tags as a sequence of
+	return 0;
-     * characters within literal '(' and ')'. */
+
+    tag_ops = tag_op_list_create (ctx);
+    if (tag_ops == NULL) {
+	fprintf (stderr, "Out of memory.\n");
+	return 1;
+    }
+
+    for (p = line; (input_format == DUMP_FORMAT_AUTO) && *p; p++) {
+	if (*p == '(')
+	    input_format = DUMP_FORMAT_SUP;
+    }
+
+    if (input_format == DUMP_FORMAT_AUTO)
+	input_format = DUMP_FORMAT_BATCH_TAG;
+
+    if (input_format == DUMP_FORMAT_SUP)
 	if ( xregcomp (&regex,
 		       "^([^ ]+) \\(([^)]*)\\)$",
 		       REG_EXTENDED) )
 	    INTERNAL_ERROR ("compile time constant regex failed.");
 
-    while ((line_len = getline (&line, &line_size, input)) != -1) {
+    do {
-	regmatch_t match[3];
+	char *query_string;
-	char *message_id, *file_tags;
 
-	chomp_newline (line);
+	if (input_format == DUMP_FORMAT_SUP) {
+	    ret = parse_sup_line (ctx, line, &query_string, tag_ops);
+	} else {
+	    ret = parse_tag_line (ctx, line, TAG_FLAG_BE_GENEROUS,
+				  &query_string, tag_ops);
 
-	rerr = xregexec (&regex, line, 3, match, 0);
+	    if (ret == 0) {
-	if (rerr == REG_NOMATCH) {
+		if (strncmp ("id:", query_string, 3) != 0) {
-	    fprintf (stderr, "Warning: Ignoring invalid input line: %s\n",
+		    fprintf (stderr, "Unsupported query: %s\n", query_string);
-		     line);
 		    continue;
 		}
-
+		/* delete id: from front of string; tag_message
-	message_id = xstrndup (line + match[1].rm_so,
+		 * expects a raw message-id.
-			       match[1].rm_eo - match[1].rm_so);
+		 *
-	file_tags = xstrndup (line + match[2].rm_so,
+		 * XXX: Note that query string id:foo and bar will be
-			      match[2].rm_eo - match[2].rm_so);
+		 * interpreted as a message id "foo and bar". This
-
+		 * should eventually be fixed to give a better error
-	tag_message (notmuch, message_id, file_tags, ! accumulate,
+		 * message.
-		     synchronize_flags);
+		 */
-
+		query_string = query_string + 3;
-	free (message_id);
+	    }
-	free (file_tags);
 	}
 
+	if (ret > 0)
+	    continue;
+
+	if (ret < 0 || tag_message (ctx, notmuch, query_string,
+				    tag_ops, flags))
+	    break;
+
+    }  while ((line_len = getline (&line, &line_size, input)) != -1);
+
+    if (input_format == DUMP_FORMAT_SUP)
 	regfree (&regex);
 
     if (line)
 	free (line);
 
     notmuch_database_destroy (notmuch);
+
     if (input != stdin)
 	fclose (input);
 
-    return 0;
+    return ret;
 }