Skip to content

E
edx-platform
Commit de06dda3 by Mike Chen
Options

added capaxmlconverter app to convert xml into json format

parent 7f6f04df
Showing with 462 additions and 121 deletions
cms/djangoapps/capaconverter/__init__.py 0 → 100644
from lxml import etree
import logging
import json
from StringIO import StringIO
import re
class CapaXMLConverter(object):
class CapaXMLConverterError(BaseException):
def __init__(self, msg):
self.msg = msg
super(CapaXMLConverter.CapaXMLConverterError, self).__init__()
def __str__(self):
return self.msg
class TagNotConsumedError(CapaXMLConverterError):
pass
class InvalidNestingError(CapaXMLConverterError):
pass
def __init__(self, logger=None):
if logger:
self.logger = logger
else:
self.logger = logging.getLogger(__name__)
self.end_of_text_tags = ['text', 'endouttext']
self.group_tags = ['multiplechoiceresponse', 'choiceresponse',
'truefalseresponse']
self.response_tags = ['choice', 'numericalresponse', 'formularesponse', 'stringresponse',
'schematicresponse', 'symbolicresponse', 'customresponse']
self.grouped_tags = ['choice']
self.script_tags = ['script']
self.text_included_tags = ['choice', 'schematicresponse', 'customresponse']
self.attr_copy_tags = ['responseparam']
self.type_map = {
'script': 'script',
'choice': 'choice',
'multiplechoiceresponse': 'multiple_choice',
'choiceresponse': 'multiple_choice',
'truefalseresponse': 'true_false',
'numericalresponse': 'numerical',
'stringresponse': 'string',
'formularesponse': 'formula',
'schematicresponse': 'schematic',
'symbolicresponse': 'symbolic',
'customresponse': 'custom',
'img': 'image',
'answer': 'script',
}
self.group_children_field = {
'multiplechoiceresponse': 'choices',
'choiceresponse': 'choices',
'truefalseresponse': 'choices',
}
self.element_post_hooks = {
'schematicresponse': self.process_schematicresponse,
'formularesponse': self.process_formularesponse,
'customresponse': self.process_customresponse,
}
self.all_tags = self.group_tags + self.response_tags + self.script_tags
self.attribute_rules = {
'*': [
],
'script': [
{
'src': 'type',
'dest': 'language',
'wrapper': lambda x: x.split('/')[1].strip(),
}
],
'customresponse': [
],
'choice': [
{
'src': 'correct',
'dest': 'correct',
'wrapper': lambda x: x == 'true',
'required': True
},
{
'src': 'location',
'dest': 'location',
},
{
'dest': 'text',
'default': '',
},
],
'answer': [
{
'src': 'type',
'dest': 'language',
'wrapper': lambda x: x.split('/')[1].strip(),
}
],
'img': [
{
'src': 'src',
'dest': 'url',
}
],
'numericalresponse': [
{
'src': 'answer',
'dest': 'answer',
}
],
'stringresponse': [
{
'src': 'answer',
'dest': 'answer',
},
],
'schematicresponse': [
{
'src': 'answer',
'dest': 'answer',
}
],
'responseparam': [
{
'src': 'default',
'dest': 'tolerance',
}
],
'formularesponse': [
{
'src': 'samples',
'dest': '_samples_',
}
],
}
super(CapaXMLConverter, self).__init__()
def build_from_element(self, element):
return self.copy_attribute(element, {'type': self.type_map[element.tag], '_tag_': element.tag})
def copy_attribute(self, element, out):
rules = self.attribute_rules['*'] + self.attribute_rules.get(element.tag, [])
for rule in rules:
if 'src' in rule:
if rule.get('required', False) or rule['src'] in element.keys():
val = element.get(rule['src'])
if 'wrapper' in rule:
val = rule['wrapper'](val)
out[rule['dest']] = val
elif 'default' in rule:
out[rule['dest']] = rule['default']
elif 'dest' in rule and 'default' in rule:
out[rule['dest']] = rule['default']
return out
def pretty_print(self, data):
print json.dumps(data, indent=2)
def dict_del_key(self, d, k):
if isinstance(d, dict) and k in d:
del d[k]
values = d.values() if isinstance(d, dict) else d
for v in values:
if isinstance(v, dict) or isinstance(v, list):
self.dict_del_key(v, k)
return d
def feed_post_process_hook(self, element, out):
if element.tag in self.element_post_hooks:
out = self.element_post_hooks[element.tag](element, out)
return out
def merge_adjacent_string(self, l):
if l:
ret = [l[0], ]
for x in l[1:]:
if isinstance(ret[-1], basestring) and isinstance(x, basestring):
ret[-1] += x
else:
ret.append(x)
return ret
return l
def iterate_element(self, element, ):
if element.text:
yield element.text
for el in element:
yield el
if el.tail:
yield el.tail
def split_element_on_tag(self, element, tag):
t = []
for part in self.iterate_element(element):
if isinstance(part, basestring) or part.tag == tag:
t.append(part)
else:
t.append(etree.tostring(part, with_tail=False))
return self.merge_adjacent_string(t)
# look for only (img, str) or (img, ) in <center> tag
def picky_center_element_format(self, center):
img_el_idx, img_el = -1, None
ret = []
for el in self.iterate_element(center):
if not isinstance(el, basestring) and not el.tag == "img":
return None
elif isinstance(el, basestring):
if not el.strip():
continue
if img_el != None:
img = self.build_from_element(img_el)
img['title'] = el
ret.append(img)
img_el = None
else:
return None
elif el.tag == "img":
if img_el != None:
ret.append(self.build_from_element(img_el))
else:
img_el = el
if img_el != None:
ret.append(self.build_from_element(img_el))
return ret
def convert_xml_file(self, filename):
out = {'scripts': [], 'contents': []}
temp = {'text': '', 'section': {}, 'in_text': False, 'embedded_text': False}
# replace <startouttext /> and <endouttext /> first
with open(filename, 'r') as f:
problem_text = f.read()
problem_text = re.sub("startouttext\s*/", "text", problem_text)
problem_text = re.sub("endouttext\s*/", "/text", problem_text)
io = StringIO(problem_text)
for event, element in etree.iterparse(io, events=('start', 'end')):
self.logger.debug("%s %s" % ("entering" if event == "start" else "leaving", element))
if event == 'start':
if element.tag in self.text_included_tags:
temp['embedded_text'] = True
if element.tag == 'text':
temp['in_text'] = True
if temp['embedded_text']:
# the text is a part of other element, say <choice><text>Choice A</text></choice>
t = etree.tostring(element).replace('<text>', '').replace('</text>', '').replace('\n', '')
temp['text'] = t
else:
# or it's individual text. find <center> elements and see if there's image inside
# e.g. [x] = ['See the diagram below. \n', <center><img src="" /></center>, 'What's .. ?']
for x in self.split_element_on_tag(element, 'center'):
if isinstance(x, basestring):
out["contents"].append({'type': 'text', 'text': x})
else:
# self.picky_center_element_format returns a list of {'type': 'image', ...} if
#
ret = self.picky_center_element_format(x)
if ret is None:
out["contents"].append({'type': 'text', 'text': etree.tostring(x)})
else:
for el in ret:
out["contents"].append(el)
element.clear()
# find response groups and create a group
elif element.tag in self.group_tags:
temp['group'] = self.build_from_element(element)
temp['group'][self.group_children_field[element.tag]] = []
# or it's indivial element
elif element.tag in self.response_tags + self.script_tags:
temp['section'] = self.build_from_element(element)
elif event == "end":
if element.tag in self.text_included_tags:
temp['embedded_text'] = False
temp['section']['text'] = temp['text']
temp['text'] = ''
# if we want to copy the attributes from this element
if element.tag in self.attr_copy_tags:
self.copy_attribute(element, temp['section'])
# if it's a script, put text as code and add it into scripts part of output
if element.tag in self.script_tags:
temp['section']['code'] = element.text
out['scripts'].append(self.feed_post_process_hook(element, temp['section']))
temp['section'] = None
elif element.tag in self.response_tags:
if temp.get('group', None):
temp['group'][self.group_children_field[temp['group']['_tag_']]].append(
self.feed_post_process_hook(element, temp['section']))
else:
out['contents'].append(self.feed_post_process_hook(element, temp['section']))
temp['section'] = None
elif element.tag in self.group_tags:
out['contents'].append(self.feed_post_process_hook(element, temp['group']))
temp['group'] = None
return (self.dict_del_key(out, '_tag_'))
def process_customresponse(self, element, out):
answer_el = element.find('answer')
answer = self.build_from_element(answer_el)
answer['code'] = answer_el.text
out['answer'] = answer
return out
def process_schematicresponse(self, element, out):
answer_el = element.find('answer')
answer = self.build_from_element(answer_el)
answer['code'] = answer_el.text
out['answer'] = answer
return out
def process_formularesponse(self, element, out):
return out
cms/djangoapps/capaconverter/management/commands/convert.py 0 → 100644
###
### One-off script for importing courseware form XML format
###
from django.core.management.base import BaseCommand, CommandError
import json
from capaconverter import CapaXMLConverter
class Command(BaseCommand):
help = \
'''Import the specified data directory into the default ModuleStore'''
def handle(self, *args, **options):
self.converter = CapaXMLConverter()
# print json.dumps(self.converter.convert_xml_file("/Users/ccp/code/mitx_all/mitx/1.xml"), indent=2)
print json.dumps(self.converter.convert_xml_file("/Users/ccp/code/mitx_all/data/6.002x/problems/HW3ID1.xml"), indent=2)
# print json.dumps(self.converter.convert_xml_file("/Users/ccp/code/mitx_all/data/6.002x/problems/multichoice.xml"), indent=2)
cms/djangoapps/capaconverter/tests.py 0 → 100755
#!/usr/bin/env python
from django.utils import unittest
import logging
import os.path
from __init__ import CapaXMLConverter
import json
import sys
from lxml import etree
class CapaXMLConverterTestCase(unittest.TestCase):
def setUp(self):
self.converter = CapaXMLConverter()
self.converter.logger.setLevel(logging.DEBUG)
self.problems_folder = os.path.join(os.path.dirname(os.path.realpath(__file__)), "problems")
logging.info("Testing problems from folder %s" % self.problems_folder)
self.problem_files = map(lambda filename: os.path.join(self.problems_folder, filename),
filter(lambda filename: filename.endswith(".xml"),
os.listdir(self.problems_folder)))
logging.info("Found %d lon-CAPA XML files. " % len(self.problem_files))
def test_center(self):
xml = '<center><img src="/aa" /></center>'
elements = self.converter.picky_center_element_format(etree.fromstring(xml))
self.assertEqual(elements, [{'url': '/aa', '_tag_': 'img', 'type': 'image'}])
xml = '<center><img src="/aa" />title</center>'
elements = self.converter.picky_center_element_format(etree.fromstring(xml))
self.assertEqual(elements, [{'url': '/aa', '_tag_': 'img', 'type': 'image', 'title': 'title'}])
xml = '<center><img src="/aa" />title<input /></center>'
elements = self.converter.picky_center_element_format(etree.fromstring(xml))
self.assertEqual(elements, None)
def test_iterator(self):
xml = """<text>In this problem we will investigate a fun idea called "duality."
<br />
Consider the series circuit in the diagram shown.
<center>
<img src="/static/images/circuits/duality.gif" />
</center>
We are given device parameters \(V=$V\)V, \(R_1=$R1\Omega\), and \(R_2=$R2\Omega\).
All of the unknown voltages and currents are labeled in associated reference
directions. Solve this circuit for the unknowns and enter them into
the boxes given.
<br />
The value (in Volts) of \(v_1\) is: </text>"""
elements = list(self.converter.iterate_element(etree.fromstring(xml)))
self.assertEqual(7, len(elements))
def test_xmls(self):
for filepath in self.problem_files:
try:
out = self.converter.convert_xml_file(filepath)
except:
print "Failed to convert file %s" % filepath
raise
f = open(filepath.replace(".xml", ".json"), "w")
json.dump(out, f, indent=2)
f.close()
if __name__ == '__main__':
unittest.main()
\ No newline at end of file
cms/djangoapps/contentstore/management/commands/convert.py deleted 100644 → 0
###
### One-off script for importing courseware form XML format
###
from django.core.management.base import BaseCommand, CommandError
import json
from lxml import etree
class CapaXMLConverter(object):
def convert_from_xml(self, filename):
out = {'scripts': [], 'contents': []}
temp = {'text':'', 'response':None}
with open(filename, "r") as f:
for event, element in etree.iterparse(f, events=("start", "end")):
if event == "start" and element.tag == "br":
temp['text'] += '\n\n'
elif event == "start" and element.text:
temp['text'] += element.text
elif event == "end" and element.tail:
temp['text'] += element.tail
if event == "start":
if element.tag == 'multiplechoiceresponse':
temp['group'] = {'type': 'multiple_choice', 'choices': []}
elif element.tag == 'truefalseresponse':
temp['group'] = {'type': 'true_false', 'statements': []}
elif element.tag == "choice":
if temp['group']['type'] == 'multiple_choice':
temp['response'] = {'type':'choice', 'text': '', 'correct': element.get('correct') == "true"}
elif temp['group']['type'] == 'true_false':
temp['response'] = {'type':'statement', 'text': '', 'correct': element.get('correct') == "true"}
elif event == "end":
if element.tag == "endouttext":
if temp['response']:
temp['response']['text'] += temp['text'].strip()
else:
out['contents'].append({'type':'paragraph', 'text': temp['text'].strip()})
temp['text'] = ''
elif element.tag in ["multiplechoiceresponse", "truefalseresponse"]:
out['contents'].append(temp['group'])
temp['group'] = None
elif element.tag == "choice":
if temp['group']['type'] == 'true_false':
temp['group']['statements'].append(temp['response'])
elif temp['group']['type'] == 'multiple_choice':
temp['group']['choices'].append(temp['response'])
temp['response'] = None
# self.parse_tree(tree, out)
return out
class Command(BaseCommand):
help = \
'''Import the specified data directory into the default ModuleStore'''
def handle(self, *args, **options):
self.converter = CapaXMLConverter()
# print json.dumps(self.converter.convert_from_xml("/Users/ccp/code/mitx_all/data/6.002x/problems/HW3ID1.xml"), indent=2)
print json.dumps(self.converter.convert_from_xml("/Users/ccp/code/mitx_all/data/6.002x/problems/multichoice.xml"), indent=2)
cms/envs/common.py
......@@ -311,6 +311,7 @@ INSTALLED_APPS = (
'contentstore',
'github_sync',
'student', # misleading name due to sharing with lms
'capaconverter',
# For asset pipelining
'pipeline',
......
format.ref → cms/static/grammars/format.ref
- scripts:
scripts:
- type: 'script'
language: 'python'
code: 'print "Hello world!"'
contents:
- type: 'text'
text: 'This is a sample paragraph. The linebreaks here matter.\n\n'
- type: 'multiple_choice'
randomize: true
choices:
- type: 'choice'
text: 'Choice A'
correct: true
- type: 'choice'
text: 'Choice B'
correct: false
- type: 'choice'
text: 'Choice C'
correct: true
- type: 'true_false'
statements:
- type: 'statement'
text: 'Sun revolves around Earth. '
correct: false
- type: 'statement'
text: 'This is a true statement. '
correct: true
- type: 'string'
answer: 'banana'
- type: 'numerical'
tolerance: '5%'
answer: 6
- type: 'formula'
answer: '-A*(RF/RS)'
samples: 10
tolerance: '5%'
variables:
- type: 'variable'
symbol: 'A'
range: '1-3'
- type: 'variable'
symbol: 'RF'
range: '1-3'
- type: 'variable'
symbol: 'RS'
range: '1-3'
- type: 'variable'
symbol: 'L'
range: '1-3'
- type: 'variable'
symbol: 'T'
range: '1-3'
- type: 'custom'
script:
- type: 'script'
language: 'python'
code: 'print "Hello world!"'
- contents:
- type: 'text'
text: 'This is a sample paragraph. The linebreaks here should matter..?'
- type: 'linebreaks'
count: 2
- type: 'multiple_choice'
randomize: true
choices:
- type: 'choice'
text: 'Choice A'
correct: true
- type: 'choice'
text: 'Choice B'
correct: false
- type: 'choice'
text: 'Choice C'
correct: true
- type: 'true_false'
statements:
- type: 'statement'
text: 'Sun revolves around Earth. '
correct: false
- type: 'statement'
text: 'This is a true statement. '
correct: true
- type: 'string'
answer: 'banana'
- type: 'numerical'
tolerance: '5%'
answer: 6
- type: 'formula'
answer: '-A*(RF/RS)'
samples: 10
tolerance: '5%'
variables:
- type: 'variable'
symbol: 'A'
range: '1-3'
- type: 'variable'
symbol: 'RF'
range: '1-3'
- type: 'variable'
symbol: 'RS'
range: '1-3'
- type: 'variable'
symbol: 'L'
range: '1-3'
- type: 'variable'
symbol: 'T'
range: '1-3'
- type: 'custom'
script:
- type: 'script'
language: 'python'
code: 'correct = true'
code: 'correct = true'
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment