Skip to content
Tools to read data out of couch efficiently
Find file

A tool to extract key/value pairs from couchdb. This code is directly derived from couchdb's file reading code, simplified and pruned down a bit to remove things I'm not using. I've removed gen_servers and tweaked the code to speed things up a bit.


This is licenced under the Apache License, Version 2.0, as was the original couchdb 1.0.3 code this is derived from.

You may obtain a copy of the License at

Notes on optimization:

Couchdb reads the last 4k of the file first, and generally tends to move from the back to the front of the file. This can subvert the readahead optimizations in the os; especially when you know you are going to read the whole file. Prereading the file in advance can shave a lot off of the walk of the btree, and the open code has been altered to read the whole file in 1MB chunks before walking the btree.

Something went wrong with that request. Please try again.