notmuch/test, branch 0.18.1

lib: Separate all phrases indexed by _notmuch_message_gen_terms

2014-06-18T21:03:18Z

This adds a 100 termpos gap between all phrases indexed by _notmuch_message_gen_terms. This fixes a bug where terms from the end of one header and the beginning of another header could match together in a single phrase and a separate bug where term positions of un-prefixed terms overlapped. This fix only affects newly indexed messages. Messages that are already indexed won't benefit from this fix without re-indexing, but the fix won't make things any worse for existing messages.

test: Known-broken test for overlapping/adjacent termpos

2014-06-18T20:56:52Z

This adds two known-broken tests and one working test related to the term positions assigned to terms from different headers or MIME parts. The first test fails because we don't create a termpos gap between different headers. The second test fails because we don't adjust termpos at all when indexing multiple parts.

lib: Index name and address of from/to headers as a phrase

2014-06-18T20:55:14Z

Previously, we indexed the name and address parts of from/to headers with two calls to _notmuch_message_gen_terms. In general, this indicates that these parts are separate phrases. However, because of an implementation quirk, the two calls to _notmuch_message_gen_terms generated adjacent term positions for the prefixed terms, which happens to be the right thing to do in this case, but the wrong thing to do for all other calls. Furthermore, _notmuch_message_gen_terms produced potentially overlapping term positions for the un-prefixed copies of the terms, which is simply wrong. This change indexes both the name and address in a single call to _notmuch_message_gen_terms, indicating that they should be part of a single phrase. This masks the problem with the un-prefixed terms (fixing the two known-broken tests) and puts us in a position to fix the unintentionally phrases generated by other calls to _notmuch_message_gen_terms.

test: Add search tests for combined name/address queries

2014-06-18T20:54:05Z

Two of these are currently known-broken. We index the name and address parts in two separate calls to _notmuch_message_gen_terms. Currently this has the effect of placing the term positions of the prefixed terms from the second call right after those of the first call, but screws up the term positions of the non-prefixed terms.

test: Fix from/to search test queries

2014-06-18T20:53:29Z

Two of the search tests for "from" and "to" queries were clearly trying to search for prefixed phrases, but forgot to shell quote the phrases. Fix this by quoting them correctly.

lib: resurrect support for single-message mbox files

2014-06-14T01:59:04Z

This is effectively a revert of commit 6812136bf576d894591606d9e10096719054d1f9 Author: Jani Nikula Date: Mon Mar 31 00:21:48 2014 +0300 lib: drop support for single-message mbox files The intention was to drop support for indexing new single-message mbox files (and whether that was a good idea in the first place is arguable). However this inadvertently broke support for reading headers from previously indexed single-message mbox files, which is far worse. Distinguishing between the two cases would require more code than simply bringing back support for single-message mbox files.

test: use --quick when starting emacs.

2014-06-14T01:55:14Z

At least in emacs24, this removes the "site-lisp" directories from the load path in addition to enforcing --no-site-lisp --no-init-file. This works around a slightly mysterious bug on Debian that causes test-lib.el not to load when there is cl-lib.el(c) in some site-lisp directory. It should be harmless in general since we really don't want to load any files from addon packages to emacs.

test: allow pending break points in atomicity script.

2014-05-17T21:32:14Z

This seems to fix problems with the symbol rename not being defined at startup on at least OS/X and some Debian Linux architectures.

test/Makefile.local: Added configured TALLOC_LDFLAGS.

2014-05-17T21:31:55Z

The linking to talloc is hard-coded in the testing Makefile. This patch causes the linking to talloc to be done according to how TALLOC_LDFLAGS was configured. Signed-off-by: Charles Celerier

test: use test_expect_equal for PATH test, update message

2014-04-17T22:53:57Z

- The old test was quite impossible to debug; the new one shows the difference between the two directories, if any. - "repository" doesn't make sense for out of tree builds. Or tarball builds, for that matter.