extract_classes.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import sys
import xml.etree.ElementTree as ET

def dump_one_entry(name, kind, path, fragment):
    if fragment != '':
        path = "%s#%s" % (path, fragment)
    print ("INSERT INTO searchIndex (name, type, path) "
           "VALUES ('%s', '%s', '%s');") % (name, kind, path)

def dump_entries(entries):
    for (name, kind, path, fragment) in entries:
        dump_one_entry(name, kind, path, fragment)

def node_to_entry(node, kind, path_ending, prefix):
    name = join(prefix, node.find('name').text)
    refid = node.attrib['refid']
    if prefix == '':
        path = "%s%s" % (refid, path_ending)
        fragment = ''
    else:
        pieces = refid.split('_')
        path = "%s%s" % ('_'.join(pieces[:-1]), path_ending)
        fragment = pieces[-1][1:]
    return (name, kind, path, fragment)

def join(ns, name):
    if ns == '':
        return name
    return ns + '::' + name

def dump_sqlite_index_script(node, entities, prefix=''):
    for child in node:
        if 'kind' not in child.attrib:
            continue

        childPrefix = prefix
        if child.attrib['kind'] in entities:
            docset_kind = entities[child.attrib['kind']]
            dump_entries([node_to_entry(child, docset_kind, '.html', prefix)])

            nameNode = child.find('name')
            if nameNode is not None:
                childPrefix = join(prefix, nameNode.text)
            dump_sqlite_index_script(child, entities, childPrefix)

def print_usage():
    print "Usage: extract_class_from_xml.py <path-to-index.xml>"

if __name__ == "__main__":
    if len(sys.argv) < 2:
        print "Please provide path to the index.xml file"
        print_usage()
        exit(1)

    entities = {
        'class': 'cl',
        'struct': 'Struct',
        'function': 'func'
    }

    index_xml_file = sys.argv[1]
    xml_root = ET.parse(index_xml_file).getroot()

    print 'CREATE TABLE searchIndex('
    print '  id INTEGER PRIMARY KEY,'
    print '  name TEXT,'
    print '  type TEXT,'
    print '  path TEXT'
    print ');'
    print 'CREATE UNIQUE INDEX anchor ON searchIndex (name, type, path);'

    print "BEGIN TRANSACTION;"
    dump_sqlite_index_script(xml_root, entities)
    print "COMMIT TRANSACTION;"