Created
December 28, 2011 19:44
-
-
Save gregglind/1529358 to your computer and use it in GitHub Desktop.
Search and Mangle json-like data structures
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
""" | |
Tasks | |
------- | |
Search and transform jsonable structures, specifically to make it 'easy' to make tabular/csv output for other consumers. | |
Example | |
~~~~~~~~~~~~~ | |
*give me a list of all the fields called 'id' in this stupid, gnarly | |
thing* | |
>>> Q('id',gnarly_data) | |
['id1','id2','id3'] | |
Observations: | |
--------------------- | |
1) 'simple data structures' exist and are common. They are tedious | |
to search. | |
2) The DOM is another nested / treeish structure, and jQuery selector is | |
a good tool for that. | |
3a) R, Numpy, Excel and other analysis tools want 'tabular' data. These | |
analyses are valuable and worth doing. | |
3b) Dot/Graphviz, NetworkX, and some other analyses *like* treeish/dicty | |
things, and those analyses are also worth doing! | |
3c) Some analyses are best done using 'one-off' and custom code in C, Python, | |
or another 'real' programming language. | |
4) Arbitrary transforms are tedious and error prone. SQL is one solution, | |
XSLT is another, | |
5) the XPATH/XML/XSLT family is.... not universally loved :) They are | |
very complete, and the completeness can make simple cases... gross. | |
6) For really complicated data structures, we can write one-off code. Getting | |
80% of the way is mostly okay. There will always have to be programmers | |
in the loop. | |
7) Re-inventing SQL is probably a failure mode. So is reinventing XPATH, XSLT | |
and the like. Be wary of mission creep! Re-use when possible (e.g., can | |
we put the thing into a DOM using | |
8) If the interface is good, people can improve performance later. | |
Simplifying | |
--------------- | |
1) Assuming 'jsonable' structures | |
2) keys are strings or stringlike. Python allows any hashable to be a key. | |
for now, we pretend that doesn't happen. | |
3) assumes most dicts are 'well behaved'. DAG, no cycles! | |
4) assume that if people want really specialized transforms, they can do it | |
themselves. | |
""" | |
from collections import Counter, namedtuple | |
import csv | |
import itertools | |
from itertools import product | |
from operator import attrgetter as aget, itemgetter as iget | |
import operator | |
import sys | |
## note 'url' appears multiple places and not all extensions have same struct | |
ex1 = { | |
'name': 'Gregg', | |
'extensions': [ | |
{'id':'hello', | |
'url':'url1'}, | |
{'id':'gbye', | |
'url':'url2', | |
'more': dict(url='url3')}, | |
] | |
} | |
## much longer example | |
ex2 = {u'metadata': {u'accessibilities': [{u'name': u'accessibility.tabfocus', | |
u'value': 7}, | |
{u'name': u'accessibility.mouse_focuses_formcontrol', u'value': False}, | |
{u'name': u'accessibility.browsewithcaret', u'value': False}, | |
{u'name': u'accessibility.win32.force_disabled', u'value': False}, | |
{u'name': u'accessibility.typeaheadfind.startlinksonly', u'value': False}, | |
{u'name': u'accessibility.usebrailledisplay', u'value': u''}, | |
{u'name': u'accessibility.typeaheadfind.timeout', u'value': 5000}, | |
{u'name': u'accessibility.typeaheadfind.enabletimeout', u'value': True}, | |
{u'name': u'accessibility.tabfocus_applies_to_xul', u'value': False}, | |
{u'name': u'accessibility.typeaheadfind.flashBar', u'value': 1}, | |
{u'name': u'accessibility.typeaheadfind.autostart', u'value': True}, | |
{u'name': u'accessibility.blockautorefresh', u'value': False}, | |
{u'name': u'accessibility.browsewithcaret_shortcut.enabled', | |
u'value': True}, | |
{u'name': u'accessibility.typeaheadfind.enablesound', u'value': True}, | |
{u'name': u'accessibility.typeaheadfind.prefillwithselection', | |
u'value': True}, | |
{u'name': u'accessibility.typeaheadfind.soundURL', u'value': u'beep'}, | |
{u'name': u'accessibility.typeaheadfind', u'value': False}, | |
{u'name': u'accessibility.typeaheadfind.casesensitive', u'value': 0}, | |
{u'name': u'accessibility.warn_on_browsewithcaret', u'value': True}, | |
{u'name': u'accessibility.usetexttospeech', u'value': u''}, | |
{u'name': u'accessibility.accesskeycausesactivation', u'value': True}, | |
{u'name': u'accessibility.typeaheadfind.linksonly', u'value': False}, | |
{u'name': u'isInstantiated', u'value': True}], | |
u'extensions': [{u'id': u'216ee7f7f4a5b8175374cd62150664efe2433a31', | |
u'isEnabled': True}, | |
{u'id': u'1aa53d3b720800c43c4ced5740a6e82bb0b3813e', u'isEnabled': False}, | |
{u'id': u'01ecfac5a7bd8c9e27b7c5499e71c2d285084b37', u'isEnabled': True}, | |
{u'id': u'1c01f5b22371b70b312ace94785f7b0b87c3dfb2', u'isEnabled': True}, | |
{u'id': u'fb723781a2385055f7d024788b75e959ad8ea8c3', u'isEnabled': True}], | |
u'fxVersion': u'9.0', | |
u'location': u'zh-CN', | |
u'operatingSystem': u'WINNT Windows NT 5.1', | |
u'surveyAnswers': u'', | |
u'task_guid': u'd69fbd15-2517-45b5-8a17-bb7354122a75', | |
u'tpVersion': u'1.2', | |
u'updateChannel': u'beta'}, | |
u'survey_data': { | |
u'extensions': [{u'appDisabled': False, | |
u'id': u'testpilot?labs.mozilla.com', | |
u'isCompatible': True, | |
u'isEnabled': True, | |
u'isPlatformCompatible': True, | |
u'name': u'Test Pilot'}, | |
{u'appDisabled': True, | |
u'id': u'dict?www.youdao.com', | |
u'isCompatible': False, | |
u'isEnabled': False, | |
u'isPlatformCompatible': True, | |
u'name': u'Youdao Word Capturer'}, | |
{u'appDisabled': False, | |
u'id': u'jqs?sun.com', | |
u'isCompatible': True, | |
u'isEnabled': True, | |
u'isPlatformCompatible': True, | |
u'name': u'Java Quick Starter'}, | |
{u'appDisabled': False, | |
u'id': u'?20a82645-c095-46ed-80e3-08825760534b?', | |
u'isCompatible': True, | |
u'isEnabled': True, | |
u'isPlatformCompatible': True, | |
u'name': u'Microsoft .NET Framework Assistant'}, | |
{u'appDisabled': False, | |
u'id': u'?a0d7ccb3-214d-498b-b4aa-0e8fda9a7bf7?', | |
u'isCompatible': True, | |
u'isEnabled': True, | |
u'isPlatformCompatible': True, | |
u'name': u'WOT'}], | |
u'version_number': 1}} | |
def denorm(queries,iterable_of_things,default=None): | |
""" | |
'repeat', or 'stutter' to 'tableize' for downstream. | |
(I have no idea what a good word for this is!) | |
Think ``kronecker`` products, or: | |
``SELECT single,multiple FROM table;`` | |
single multiple | |
------- --------- | |
id1 val1 | |
id1 val2 | |
Args: | |
queries: iterable of ``Q`` queries. | |
iterable_of_things: to be queried. | |
Returns: | |
list of 'stuttered' output, where if a query returns | |
a 'single', it gets repeated appropriately. | |
""" | |
def _denorm(queries,thing): | |
fields = [] | |
results = [] | |
for q in queries: | |
#print q | |
r = Ql(q,thing) | |
#print "-- result: ", r | |
if not r: | |
r = [default] | |
if type(r[0]) is type({}): | |
fields.append(sorted(r[0].keys())) # dicty answers | |
else: | |
fields.append([q]) # stringy answer | |
results.append(r) | |
#print results | |
#print fields | |
flist = list(flatten(*map(iter,fields))) | |
prod = itertools.product(*results) | |
for p in prod: | |
U = dict() | |
for (ii,thing) in enumerate(p): | |
#print ii,thing | |
if type(thing) is type({}): | |
U.update(thing) | |
else: | |
U[fields[ii][0]] = thing | |
yield U | |
return list(flatten(*[_denorm(queries,thing) for thing in iterable_of_things])) | |
def default_iget(fields,default=None,): | |
""" itemgetter with 'default' handling, that *always* returns lists | |
API CHANGES from ``operator.itemgetter`` | |
Note: Sorry to break the iget api... (fields vs *fields) | |
Note: *always* returns a list... unlike itemgetter, | |
which can return tuples or 'singles' | |
""" | |
myiget = operator.itemgetter(*fields) | |
L = len(fields) | |
def f(thing): | |
try: | |
ans = list(myiget(thing)) | |
if L < 2: | |
ans = [ans,] | |
return ans | |
except KeyError: | |
# slower! | |
return [thing.get(x,default) for x in fields] | |
f.__doc__ = "itemgetter with default %r for fields %r" %(default,fields) | |
f.__name__ = "default_itemgetter" | |
return f | |
def flatten(*stack): | |
""" | |
helper function for flattening iterables of generators in a | |
sensible way. | |
""" | |
stack = list(stack) | |
while stack: | |
try: x = stack[0].next() | |
except StopIteration: | |
stack.pop(0) | |
continue | |
if hasattr(x,'next') and callable(getattr(x,'next')): | |
stack.insert(0, x) | |
#if isinstance(x, (GeneratorType,listerator)): | |
else: yield x | |
def _Q(filter_, thing): | |
""" underlying machinery for Q function recursion """ | |
T = type(thing) | |
if T is type({}): | |
for k,v in thing.iteritems(): | |
#print k,v | |
if filter_ == k: | |
if type(v) is type([]): | |
yield iter(v) | |
else: | |
yield v | |
if type(v) in (type({}),type([])): | |
yield Q(filter_,v) | |
elif T is type([]): | |
for k in thing: | |
#print k | |
yield Q(filter_,k) | |
else: | |
# no recursion. | |
pass | |
def Q(filter_,thing): | |
""" | |
type(filter): | |
- list: a flattened list of all searches (one list) | |
- dict: dict with vals each of which is that search | |
Notes: | |
[1] 'parent thing', with space, will do a descendent | |
[2] this will come back 'flattened' jQuery style | |
[3] returns a generator. Use ``Ql`` if you want a list. | |
""" | |
if type(filter_) is type([]): | |
return flatten(*[_Q(x,thing) for x in filter_]) | |
elif type(filter_) is type({}): | |
d = dict.fromkeys(filter_.keys()) | |
#print d | |
for k in d: | |
#print flatten(Q(k,thing)) | |
d[k] = Q(k,thing) | |
return d | |
else: | |
if " " in filter_: # i.e. "antecendent post" | |
parts = filter_.strip().split() | |
r = None | |
for p in parts: | |
r = Ql(p,thing) | |
thing = r | |
return r | |
else: # simple. | |
return flatten(_Q(filter_,thing)) | |
def Ql(filter_,thing): | |
""" same as Q, but returns a list, not a generator """ | |
res = Q(filter_,thing) | |
if type(filter_) is type({}): | |
for k in res: | |
res[k] = list(res[k]) | |
return res | |
else: | |
return list(res) | |
def countit(fields,iter_of_iter,default=None): | |
""" | |
note: robust to fields not being in i_of_i, using ``default`` | |
""" | |
C = Counter() # needs hashables | |
T = namedtuple("Thing",fields) | |
get = default_iget(*fields,default=default) | |
return Counter( | |
(T(*get(thing)) for thing in iter_of_iter) | |
) | |
## right now this works for one row... | |
def printout(queries,things,default=None, f=sys.stdout, **kwargs): | |
""" will print header and objects | |
**kwargs go to csv.DictWriter | |
help(csv.DictWriter) for more. | |
""" | |
results = denorm(queries,things,default=None) | |
fields = set(itertools.chain(*(x.keys() for x in results))) | |
W = csv.DictWriter(f=f,fieldnames=fields,**kwargs) | |
#print "---prod---" | |
#print list(prod) | |
W.writeheader() | |
for r in results: | |
W.writerow(r) | |
def test_run(): | |
print "\n>>> print list(Q('url',ex1))" | |
print list(Q('url',ex1)) | |
assert list(Q('url',ex1)) == ['url1','url2','url3'] | |
assert Ql('url',ex1) == ['url1','url2','url3'] | |
print "\n>>> print list(Q(['name','id'],ex1))" | |
print list(Q(['name','id'],ex1)) | |
assert Ql(['name','id'],ex1) == ['Gregg','hello','gbye'] | |
print "\n>>> print Ql('more url',ex1)" | |
print Ql('more url',ex1) | |
print "\n>>> list(Q('extensions',ex1))" | |
print list(Q('extensions',ex1)) | |
print "\n>>> print Ql('extensions',ex1)" | |
print Ql('extensions',ex1) | |
print "\n>>> printout(['name','extensions'],[ex1,], extrasaction='ignore')" | |
printout(['name','extensions'],[ex1,], extrasaction='ignore') | |
print "\n\n" | |
from pprint import pprint as pp | |
print "-- note that the extension fields are also flattened! (and N/A) -- " | |
pp(denorm(['location','fxVersion','notthere','survey_data extensions'],[ex2,], default="N/A")[:2]) | |
if __name__ == "__main__": | |
test_run() |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment