A couchdb helper thing. Instead of overwriting docs (Hulk smash!), use two calls to extend existing docs with passed in hash of docs with new or updated things.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
lib
test
.codeclimate.yml
.eslintignore
.eslintrc.yml
.gitignore
.travis.yml
COPYING
README.md
package-lock.json
package.json
test.config.json

README.md

couchdb_bulkdoc_appender

Greenkeeper badge Code Climate Test Coverage Build Status

This is a utility to help out saving docs to couchdb when those docs might be updating existing docs.

very similar to my couchdb_bulkdoc_saver, except that existing fields in a doc that are not in the "updating" doc are preserved, as I will describe below.

Basically, the need I have is that I have a state document that stores lots of things, and I want to update lots of these documents all at once because I've updated how the geometry stuff (added a few decimal points to the geojson representation). So I have lots of document ids, and just one field from the existing documents to update.

Instead of one by one extracting and updating each document, I am instead wanting to bulk get lots of docs, and then update all of them iwth new information, and then bulk save the lot.

And create any new documents as needed.

Instead of making code inside my application to do this, I am creating a small library with a test.

What is does it perform two calls to CouchDB. The first is to all_docs?include_docs=true which grabs the docs, the revision, and the contents of each doc.

Each doc is visited, and if there isn't an existing doc, the passed in doc will represent the entirety of that document in the db. If there is an existing document, then the passed in data is appended to the existing doc, overwriting anything that already exists in the doc, but everything else is copied as is from that existing document.

Then the second call to CouchDB is to bulk_docs, and the docs are written.

The return value is passed to a callback you supply. If no callback is supplied, then you're on your own.

usage

The program is a function generator. You call it with an options object, and an optional callback function. The saver function is either sent back as the return value, or else sent as the second argument to the passed callback function.

The program expects that the url, port, username, password are in the passed in options object. For example

var make_bulkdoc_appender=require('couchdb_bulkdoc_appender')
var appender = make_bulkdoc_appender({
        "url": "http://127.0.0.1",
        "port":5984,
        "db": "a_test_data",
        "auth":{"username":"shebang",
                "password":"girls rule boys drool"
               })

tests

I have a basic test that checks that this works. It does.