Implicit CSV headers name;cpu;ram;os;disk
Valid headers name;cpu;ram;constant;os;disk;latency;optimized;type;internet;maxVariableCost;ephemeral;maxQuantity;minQuantity;location;usage;budget;optimizer;tags;processor;physical
name Unique name within the quote
description Optional description of the resource
cpu Minimum CPU, decimal
ram Minimum memory, Number, in Mega Byte
constant [(null),true,false] for CPU behavior
os [(LINUX),WINDOWS,RHEL,SUSE,CENTOS,DEBIAN]
disk Optional disk size, in Gibibyte (GiB)
latency [WORST, LOW, MEDIUM, GOOD, BEST], default is GOOD
optimized [IOPS,THROUGHPUT], default is auto
type Instance type name, case insensitive, default is auto
internet [NAT,PUBLIC,(PRIVATE)]
ephemeral When true, ephemeral contract is accepted. False by default
minQuantity Minimal quantity, default is 1
maxQuantity Maximal quantity, default is 1
location Location name override of quote's one
usage Usage name override of quote's one
budget Budget name override of quote's one
optimizer Optimizer name override of quote's one
maxVariableCost Maximum variable instance cost you are willing to pay
tags Tags key:value, with ',', ' ' as separator. Tag value is optional
processor Processor: Intel Xeon, AMD,... 'startsWith' is used for the match
physical [(null),true,false] for physical/virtual
{{instance-import-sample}}
name;cpu;ram;os;disk;latency;optimized;constant;instance;usage
mysql;1.5:2048;RHEL;20;GOOD;IOPS;true
nginx;0.5:1000;LINUX
jetty;2:2048;LINUX;;;;m1.large;production2

Trailing columns can be omitted if empty, default value or "false".
For now, only ';' separator is supported".
The headers do not need to match exactly. There is a scoring mechanism to get the best mapping.