Summary: Ref T8788. See D17702. This allows `bin/search index` to index stuff which only implements `Ngrams`, not `Fulltext`.
Test Plan: Kinda poked around `bin/search index` a bit, yell if you hit more issues deeper down the stack?
Reviewers: amckinley
Reviewed By: amckinley
Maniphest Tasks: T8788
Differential Revision: https://secure.phabricator.com/D17704
Summary: Ref T10054. Take specialization off the objects and put it on Engine subclasses instead. One reason for this is that certain objects (like users) might have multiple different sets of panels in the future (e.g., their user profile and their home page).
Test Plan:
- No user-visible changes.
- PanelEngine no longer has any hardcoded "project" stuff.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T10054
Differential Revision: https://secure.phabricator.com/D15018
Summary: Ref T10054. This has no product impact, but prepares us for customizable side nav on "profiles" (today, projects; probably users some day; and maybe other stuff down the road).
Test Plan: Clicked all links on a profile, everything was exactly the same as before.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T10054
Differential Revision: https://secure.phabricator.com/D15007
Summary:
Ref T9979. This uses ngrams (specifically, trigrams) to build a reasonably efficient index for substring matching. Specifically, for a package like "Example", with ID 123, we store rows like this:
```
< ex, 123>
<exa, 123>
<xam, 123>
<amp, 123>
<mpl, 123>
<ple, 123>
<le , 123>
```
When the user searches for `exam`, we join this table for packages with tokens `exa` and `xam`. MySQL can do this a lot more efficiently than it can process a `LIKE "%exam%"` query against a huge table.
When the user searches for a one-letter or two-letter string, we only search the beginnings of words. This is probably what they want, the only thing we can do quickly, and a reasonable/expected behavior for typeaheads.
Test Plan:
- Ran storage upgrades and search indexer.
- Searched for stuff with "name contains".
- Used typehaead and got sensible results.
- Searched for `aabbccddeeffgghhiijjkkllmmnnooppqqrrssttuuvvwwxxyyzz` and saw only 16 joins.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T9979
Differential Revision: https://secure.phabricator.com/D14846
Summary: Ref T9979. This simplifies/standardizes the code a bit, but mostly gives us more consistent class names and structure.
Test Plan:
- Used `bin/search index --type ...` to index documents of every indexable type.
- Searched for documents by unique text, found them.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T9979
Differential Revision: https://secure.phabricator.com/D14842
Summary:
This reverts commit 1583738842.
See T8646 for discussion. This version of the feature feels terrible on real data.
Test Plan: Strict revert.
Reviewers: chad
Reviewed By: chad
Differential Revision: https://secure.phabricator.com/D14097
Summary:
Ref T8646. This is fairly rough:
This interface is very niche, and not really flexible enough to accommodate other result customization (but I don't think we have any plans here)?
I'm just //summarizing// the content of documents, basically showing the first paragraph of their content, summary, etc. This isn't what Google does: it shows snippets surrounding the actual search terms. However, this is more involved and might be less useful in structured data: for example, I'd imagine that the first line of most phriciton documents, maniphest tasks and Differential revisions really might be the best machine-generatable summary of them. The actual contextual snippeting in Google doesn't often seem hugely useful to me. But this might also not be very useful.
There's not much design, not sure if you had any ideas.
I only implemented this for tasks, revisions and the wiki since those seem most useful.
I'm generally on the fence about this, but it's not a ton of work to swap out for something else later. Maybe we can see how it feels? But happy to toss it or rethink the approach.
Test Plan: {F788026}
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T8646
Differential Revision: https://secure.phabricator.com/D14095
Summary: Releeph does some really old sketchy stuff in result rendering. Modernize it, remove the holdover/oldschool interface (these were the last two implementors) and make errors a little softer.
Test Plan: Viewed Releeph products and branches.
Reviewers: btrahan, chad
Reviewed By: chad
Subscribers: epriestley
Differential Revision: https://secure.phabricator.com/D13604
Summary: This allows the SavedQuery to modify what the result list looks like (e.g., include display flags and similar).
Test Plan: Looked at some ApplicationSearch apps.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2625
Differential Revision: https://secure.phabricator.com/D6346
Summary:
Ref T2625. Lifts almost all of the search logic out of Paste controllers and into Search.
This uses controller delegation for generalization. We use this in a few places, but don't use it very much yet. I think it's pretty reasonable as-is, but I might be able to make even more stuff free.
There are some slightly rough edges around routes, still, but I want to hit Phame and Differential (which both have multiple application search engines) before trying to generalize that.
Test Plan: Executed, browsed and managed Paste searches.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2625
Differential Revision: https://secure.phabricator.com/D6073
