Consider a system where one component is scheduling tasks that yield
streams, and passing them to (unknown) clients for consumption.

It would be useful for the scheduler to know that the query
underlying the stream is completed (so it can continue on to it's
next task) without having to wait for the consumer to finish reading
all results.

closes #2

Emit 'close' events when query completes

pg-cursor no longer returns the empty array 'done' signal to the callback
until the cursor recieves a readyForQuery message.  This means pg-query-stream
will not emit 'close' or 'end' events until the server is __truly__ ready for
the next query.  This fixes some race-conditions where some queries
are triggered off of the `end` event of the query-stream

closes #3

- appears that timestamp queries emit a lot of `rows` with length == 0
- `self.once('end')` is added each of these times
- assertion on listener count shows that more than 10 listeners are applied

- moves 'end' event listener to constructor, only listen once
- ensures all existing tests still green

maxListeners on timestamp queries

fix tests, clean up a few things

Is it possible to destroy the stream?

Add travis badge

One of the tests was failing because it was testing that when a stream became readable it never returned a `null` datum on call to `stream.read()`.

In fact, when a readable stream drains it should & does return `null` for calls to `stream.read()` as described [here](https://nodejs.org/api/stream.html#stream_event_readable)

I updated the test to account for this, and they pass now.

Conform to readable stream spec

Avoid race when stream closed while fetching

Add more versions of node to the travis matrix

Remove outdated information about pg.js

Add node LTS and current versions to travis matrix

Upgrade to newest version of pg-cursor

Move eslint to dev dependencies

* Fix for [email protected]

Don't return anything from `stream.submit`

* Add node@10 to travis version

* Relax version of [email protected]

* Upgrade dependencies

- There was a security vuln with mocha, so upgraded mocha.
- Upgraded versions of node we're going to check in travis
- Switched to yarn from npm
- Removed --no-exit as that's standard mocha behavior now

* Enable postgres on newer version of travis

Bumps [js-yaml](https://github.com/nodeca/js-yaml) from 3.9.1 to 3.13.1.
- [Release notes](https://github.com/nodeca/js-yaml/releases)
- [Changelog](https://github.com/nodeca/js-yaml/blob/master/CHANGELOG.md)
- [Commits](nodeca/js-yaml@3.9.1...3.13.1)

Signed-off-by: dependabot[bot] <[email protected]>

Bumps [debug](https://github.com/visionmedia/debug) from 2.6.8 to 2.6.9.
- [Release notes](https://github.com/visionmedia/debug/releases)
- [Changelog](https://github.com/visionmedia/debug/blob/2.6.9/CHANGELOG.md)
- [Commits](debug-js/debug@2.6.8...2.6.9)

Signed-off-by: dependabot[bot] <[email protected]>

This includes fixes in [email protected].  I've relaxed semver a touch so I don't have to release a new version here just for patch changes to pg-cursor.

Bumps [eslint](https://github.com/eslint/eslint) from 4.4.0 to 4.18.2.
- [Release notes](https://github.com/eslint/eslint/releases)
- [Changelog](https://github.com/eslint/eslint/blob/master/CHANGELOG.md)
- [Commits](eslint/eslint@v4.4.0...v4.18.2)

Signed-off-by: dependabot[bot] <[email protected]>

* Bump version of pg-cursor

This includes fixes in [email protected].  I've relaxed semver a touch so I don't have to release a new version here just for patch changes to pg-cursor.

* Pass options to pg-cursor

fixes #55

…d87bfc9035e487e8'

git-subtree-dir: packages/pg-query-stream
git-subtree-mainline: cccf84e
git-subtree-split: 9ced05e

Delete accidental addition