From: Jameson Rollins <jrollins@finestructure.net>
To: Olly Betts <olly@survex.com>, notmuch@notmuchmail.org
In-Reply-To: <slrnhmpbhk.4kj.olly@msgid.survex.com>
References: <87636bws2o.fsf@servo.finestructure.net>
	<slrnhmpbhk.4kj.olly@msgid.survex.com>
Date: Mon, 08 Feb 2010 12:24:06 -0500
Message-ID: <87hbpr8wuh.fsf@servo.finestructure.net>
MIME-Version: 1.0
Content-Type: multipart/signed; boundary="=-=-=";
	micalg=pgp-sha256; protocol="application/pgp-signature"
Subject: Re: [notmuch] strange behavior of indexing of and searching for
 strings containing '[]'
Precedence: list

--=-=-=
Content-Transfer-Encoding: quoted-printable

On Fri, 5 Feb 2010 23:48:03 +0000 (UTC), Olly Betts <olly@survex.com> wrote:
> On 2010-02-05, Jameson Rollins wrote:
> > Hey, folks.  I've been noticing some strange behavior of notmuch search
> > results for strings containing '[]'.  Here are some searches for some
> > exact strings in messages subjects:
>=20
> The '[]' is a red herring.  Xapian's TermGenerator and QueryParser classes
> treat these two characters pretty much as if they were spaces.

Ah.  Thanks for the response, Olly.  This clears things up a lot.

> > servo:~ 0$ notmuch search subject:'emacs paned UI'
>=20
> Note that the '' is quoting for the shell only here.  So Xapian sees:
>=20
>     subject:emacs paned UI
>=20
> Assuming you are defaulting to an AND search, that's `emacs in the subjec=
t'
> AND `paned anywhere in the indexed text' AND `UI anywhere in the indexed =
text'.

I don't think that this is exactly correct.  The quoting is interpreted
by the shell in order to construct a single string that is then passed
as an argument to the program.  Notmuch should then be seeing the single
string argument as the search parameter, and not breaking it up further.

Here's an example of what I mean:

servo:~/tmp/cdtemp.AYroUf 0$ cat parse=20
#!/bin/bash
for arg; do echo "$arg"; done
servo:~/tmp/cdtemp.AYroUf 0$ ./parse subject:foo bar baz
subject:foo
bar
baz
servo:~/tmp/cdtemp.AYroUf 0$ ./parse subject:'foo bar' baz
subject:foo bar
baz
servo:~/tmp/cdtemp.AYroUf 0$ ./parse subject:"foo bar" baz
subject:foo bar
baz
servo:~/tmp/cdtemp.AYroUf 0$=20

As you can see in the last command, the argument "subject'foo bar'" is
passed as a single string by the shell, and should therefore be
interpreted as such by notmuch.

jamie.

--=-=-=
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
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=6eE7
-----END PGP SIGNATURE-----
--=-=-=--