On-wire algorithm in pseudo code

[avtobiff]

Starting point for discussion about pseudo code for the on-wire algorithm.

// collect analysis rules
rules = common rules + encoding rules  // encoding may override common
verbosity = verbosity level from user
skips = skip-analysis from user

// the subset of the schema to be analyzed
filters = yang artifact filters from user
paths-old = instance-identifiers for all nodes in old yang artifact - filters
paths-new = instance-identifiers for all nodes in new yang artifact - filters
yang-nodes-old = get old yang-artifact for all paths-old
yang-nodes-new = get new yang-artifact for all paths-new

// initialize empty analysis-result
analysis-result = []

// perform analysis
for each p-new in paths-new:

    // node removed in new model
    if p-new is not in paths-old:
        append p-new as removed to analysis-result

    // node exists in both models, p-old == p-new
    else:
        yang-node-old = yang-nodes-old[p-new]
        yang-node-new = yang-nodes-new[p-new]
        for each rule in rules:
            analyze(rule, yang-node-old, yang-node-new) and \
                append result to analysis-result

// report
for result in analysis-result:

    if result[rule] is not in skips AND result[rule] is above verbosity:

        print result

[avtobiff]

Should we capture additions in the new YANG artifact?