I have a valid JSON string stored in a cassandra column that is about 1MB but could grow to (worst case) 50MB and it could be retrieved by over a 1000 people at anyone time. When the string is retrieved JSON.parse is called to convert the string to a JSON object before doing any operation in node
most of the time, the operation is to pull out a single key:value.
It seems inefficient to have to convert a the "whole" string to the object first, only to pull out a single key:value, so i thought it would be better to just for-loop through the string, till i find the required key.
This however presents its challenges e,g nested objects, escaped characters etc
1) is there a simple JSON string parser that can iterate through a JSON string?
2) an alternative is to store the JSON object in memcache, it does mean duplicate data (cassandra and memcache) but in different formats (string and object respectively), is this more efficient or a bit much?
3) I saw clarinet, but the authors tests shows that JSON.parse still out performed clarinet on a big file
4) any other ideas?