Re: [Patch v7 04/14] notmuch-tag: factor out double quoting routine

[Jani Nikula <jani@nikula.org>]

On Fri, 14 Dec 2012, david@tethera.net wrote:
> From: David Bremner <bremner@debian.org>
>
> This could live in tag-util as well, but it is really nothing specific
> to tags (although the conventions are arguable specific to Xapian).
>
> The API is changed from "caller-allocates" to "readline-like". The scan for
> max tag length is pushed down into the double quoting routine.
> ---
>  notmuch-tag.c      |   50 ++++++++++++++++----------------------------------
>  util/string-util.c |   34 ++++++++++++++++++++++++++++++++++
>  util/string-util.h |    8 ++++++++
>  3 files changed, 58 insertions(+), 34 deletions(-)
>
> diff --git a/notmuch-tag.c b/notmuch-tag.c
> index 0965ee7..13f2268 100644
> --- a/notmuch-tag.c
> +++ b/notmuch-tag.c
> @@ -20,6 +20,7 @@
>  
>  #include "notmuch-client.h"
>  #include "tag-util.h"
> +#include "string-util.h"
>  
>  static volatile sig_atomic_t interrupted;
>  
> @@ -37,25 +38,6 @@ handle_sigint (unused (int sig))
>  }
>  
>  static char *
> -_escape_tag (char *buf, const char *tag)
> -{
> -    const char *in = tag;
> -    char *out = buf;
> -
> -    /* Boolean terms surrounded by double quotes can contain any
> -     * character.  Double quotes are quoted by doubling them. */
> -    *out++ = '"';
> -    while (*in) {
> -	if (*in == '"')
> -	    *out++ = '"';
> -	*out++ = *in++;
> -    }
> -    *out++ = '"';
> -    *out = 0;
> -    return buf;
> -}
> -
> -static char *
>  _optimize_tag_query (void *ctx, const char *orig_query_string,
>  		     const tag_op_list_t *list)
>  {
> @@ -67,44 +49,44 @@ _optimize_tag_query (void *ctx, const char *orig_query_string,
>       * parenthesize and the exclusion part of the query must not use
>       * the '-' operator (though the NOT operator is fine). */
>  
> -    char *escaped, *query_string;
> +    char *escaped = NULL;
> +    size_t escaped_len = 0;
> +    char *query_string;
>      const char *join = "";
>      size_t i;
> -    unsigned int max_tag_len = 0;
>  
>      /* Don't optimize if there are no tag changes. */
>      if (tag_op_list_size (list) == 0)
>  	return talloc_strdup (ctx, orig_query_string);
>  
> -    /* Allocate a buffer for escaping tags.  This is large enough to
> -     * hold a fully escaped tag with every character doubled plus
> -     * enclosing quotes and a NUL. */
> -    for (i = 0; i < tag_op_list_size (list); i++)
> -	if (strlen (tag_op_list_tag (list, i)) > max_tag_len)
> -	    max_tag_len = strlen (tag_op_list_tag (list, i));
> -
> -    escaped = talloc_array (ctx, char, max_tag_len * 2 + 3);
> -    if (! escaped)
> -	return NULL;
> -
>      /* Build the new query string */
>      if (strcmp (orig_query_string, "*") == 0)
>  	query_string = talloc_strdup (ctx, "(");
>      else
>  	query_string = talloc_asprintf (ctx, "( %s ) and (", orig_query_string);
>  
> +
> +    /* Boolean terms surrounded by double quotes can contain any
> +     * character.  Double quotes are quoted by doubling them. */
> +
>      for (i = 0; i < tag_op_list_size (list) && query_string; i++) {
> +	double_quote_str (ctx,
> +			  tag_op_list_tag (list, i),
> +			  &escaped, &escaped_len);

Check return value?

> +
>  	query_string = talloc_asprintf_append_buffer (
>  	    query_string, "%s%stag:%s", join,
>  	    tag_op_list_isremove (list, i) ? "" : "not ",
> -	    _escape_tag (escaped, tag_op_list_tag (list, i)));
> +	    escaped);
>  	join = " or ";
>      }
>  
>      if (query_string)
>  	query_string = talloc_strdup_append_buffer (query_string, ")");
>  
> -    talloc_free (escaped);
> +    if (escaped)
> +	talloc_free (escaped);
> +
>      return query_string;
>  }
>  
> diff --git a/util/string-util.c b/util/string-util.c
> index 44f8cd3..ea7c25b 100644
> --- a/util/string-util.c
> +++ b/util/string-util.c
> @@ -20,6 +20,7 @@
>  
>  
>  #include "string-util.h"
> +#include "talloc.h"
>  
>  char *
>  strtok_len (char *s, const char *delim, size_t *len)
> @@ -32,3 +33,36 @@ strtok_len (char *s, const char *delim, size_t *len)
>  
>      return *len ? s : NULL;
>  }
> +
> +
> +int
> +double_quote_str (void *ctx, const char *str,
> +		  char **buf, size_t *len)
> +{
> +    const char *in;
> +    char *out;
> +    size_t needed = 3;
> +
> +    for (in = str; *in; in++)
> +	needed += (*in == '"') ? 2 : 1;
> +
> +    if (needed > *len)
> +	*buf = talloc_realloc (ctx, *buf, char, 2*needed);

You fail to set *len to 2*needed, leading to doing realloc every time.

Also, I think you should follow the getline pattern like you did in
hex_encode: if *buf == NULL, the input value of *len is ignored.

BR,
Jani.

> +
> +    if (! *buf)
> +	return 1;
> +
> +    out = *buf;
> +
> +    *out++ = '"';
> +    in = str;
> +    while (*in) {
> +	if (*in == '"')
> +	    *out++ = '"';
> +	*out++ = *in++;
> +    }
> +    *out++ = '"';
> +    *out = 0;
> +
> +    return 0;
> +}
> diff --git a/util/string-util.h b/util/string-util.h
> index ac7676c..b593bc7 100644
> --- a/util/string-util.h
> +++ b/util/string-util.h
> @@ -19,4 +19,12 @@
>  
>  char *strtok_len (char *s, const char *delim, size_t *len);
>  
> +/* Copy str to dest, surrounding with double quotes.
> + * Any internal double-quotes are doubled, i.e. a"b -> "a""b"
> + *
> + * Output is into buf; it may be talloc_realloced
> + * return 0 on success, non-zero on failure.
> + */
> +int double_quote_str (void *talloc_ctx, const char *str,
> +		      char **buf, size_t *len);
>  #endif
> -- 
> 1.7.10.4
>
> _______________________________________________
> notmuch mailing list
> notmuch@notmuchmail.org
> http://notmuchmail.org/mailman/listinfo/notmuch