def _get_code_w_scope(data, view, position, obj_end, opts, element_name):
"""Decode a BSON code_w_scope to bson.code.Code."""
code_end = position + _UNPACK_INT_FROM(data, position)[0]
code, position = _get_string(data, view, position + 4, code_end, opts,
element_name)
scope, position = _get_object(data, view, position, code_end, opts,
element_name)
if position != code_end:
raise InvalidBSON('scope outside of javascript code boundaries')
return Code(code, scope), position
def test_code(self):
a_string = "hello world"
a_code = Code("hello world")
self.assertTrue(a_code.startswith("hello"))
self.assertTrue(a_code.endswith("world"))
self.assertTrue(isinstance(a_code, Code))
self.assertFalse(isinstance(a_string, Code))
self.assertEqual(a_code.scope, {})
a_code.scope["my_var"] = 5
self.assertEqual(a_code.scope, {"my_var": 5})
def analyze_tweets():
if mapcollection not in db.collection_names():
map_function = Code(open('mapReduce/mapFunctionTweets.js', 'r').read())
reduce_function = Code(open('mapReduce/reduceFunctionTweets.js', 'r').read())
collection.map_reduce(map_function, reduce_function, out=mapcollection)
json_file = open(output, 'wb')
json_file.write('[')
first = True
for doc in db[mapcollection].find().sort([('value', -1)]).limit(50):
if first:
first = False
else:
json_file.write(',')
json_file.write(json.dumps(doc, indent=2, default=json_util.default))
json_file.write(']')
json_file.close()
def get_statistics( self, request, **kwargs ):
'''
Get "statistics" aim is to (ultimately) run MapReduce functions on
the set of data in a specific repository.
'''
day_map = Code( '''
function() {
day = Date.UTC( this.timestamp.getFullYear(), this.timestamp.getMonth(), this.timestamp.getDate() );
emit( { day: day }, { count: 1 } )
}''')
day_reduce = Code( '''
function( key, values ) {
var count = 0;
values.forEach( function( v ) {
count += v[ 'count' ];
});
return {count: count};
}''')
# Grab the survey that we're querying survey data for
repo_filter = { 'repo': ObjectId( kwargs.get( 'mongo_id' ) ) }
cursor = db.data.find( repo_filter )
first = db.data.find_one( repo_filter, sort=[( '_id', pymongo.ASCENDING )] )
last = db.data.find_one( repo_filter, sort=[( '_id', pymongo.DESCENDING )] )
count_by_day = []
result = db.data.map_reduce( day_map, day_reduce, "myresults", query=repo_filter )
for doc in result.find():
count_by_day.append( {
'day': doc[ '_id' ][ 'day' ],
'value': doc[ 'value' ][ 'count' ] })
stats = {
'total_count': cursor.count(),
'count_by_day': count_by_day,
'first_submission': first,
'last_submission': last,
}
return self.create_response( request, stats )
def get_phrase_heading_counts(phrase,
speakername=None,
how_many=25,
from_date=None,
to_date=None):
''' A list of headings by number of occurrences for a phrase '''
query = {"phrase": phrase}
if from_date and to_date:
query["date"] = {"$gte": from_date, "$lte": to_date}
if speakername:
query["speakername"] = speakername
map = Code(
"function () {"
" emit(this.headingtitle.substring(0,64) + ' ('+ this.date + ')',1);"
"}")
reduce = Code("function (key, values) {" " return Array.sum(values)" "}")
results = db.phrases.map_reduce(map, reduce, "results", query=query)
for doc in results.find().sort("value", -1).limit(how_many):
yield doc
def mapReduce():
db = client.Corpus
tweets = db.tweets
map = Code("function() {emit(this.via,1);}")
reduce = Code("""function(key, values) {
var res=0;
values.forEach(function(v){res +=1})
return {count: res};
}""")
result = tweets.map_reduce(map,
reduce,
"via_count",
query={"sentiment": 2},
limit=10)
print result
for doc in db.via_count.find():
print(doc)
def get_phrases_containing(fragment,
how_many=25,
from_date=None,
to_date=None,
speakername=None):
''' A list of phrases containing some text '''
#query = {"phrase":re.compile(".*"+fragment+".*", re.IGNORECASE)}
query = {
"phrase": re.compile("(^|\s)(" + fragment + ")($|\s)", re.IGNORECASE)
}
if from_date and to_date:
query["date"] = {"$gte": from_date, "$lte": to_date}
if speakername:
query["speakername"] = speakername
map = Code("function () {" " emit(this.phrase,1);" "}")
reduce = Code("function (key, values) {" " return Array.sum(values)" "}")
results = db.phrases.map_reduce(map, reduce, "results", query=query)
for doc in results.find().sort("value", -1).limit(how_many):
yield doc
def group(self, keys, initial, reduce, condition=None, finalize=None, **kwargs):
body = {
"ns": self._collection_name,
"initial": initial,
"$reduce": Code(reduce),
}
if isinstance(keys, (bytes, unicode)):
body["$keyf"] = Code(keys)
else:
body["key"] = self._normalize_fields_projection(keys)
if condition:
body["cond"] = condition
if finalize:
body["finalize"] = Code(finalize)
result = yield self._database.command("group", body, **kwargs)
defer.returnValue(result)
def snippet(self, _input):
"""Takes in JavaScript string and code to function.
Args:
_input(:obj:`str`): JavaScript string.
Return:
(:obj:`bson.code.Code`)
"""
return Code(_input)
def object_hook(dct, compile_re=True):
if "$oid" in dct:
return ObjectId(str(dct["$oid"]))
if "$numberLong" in dct:
return int(dct["$numberLong"])
if "$decimal" in dct:
v = str(dct["$decimal"])
if "$precision" in dct:
precision = dct["$precision"][0]
scale = dct["$precision"][1]
d = Decimal(v, precision, scale)
else:
d = Decimal(v)
return d
if "$ref" in dct:
return DBRef(dct["$ref"], dct["$id"], dct.get("$db", None))
if "$date" in dct:
try:
secs = float(dct["$date"]) / 1000.0
return EPOCH_AWARE + datetime.timedelta(seconds=secs)
except ValueError:
return datetime.datetime.strptime(dct["$date"], "%Y-%m-%d")
if "$timestamp" in dct:
try:
ms = long_type(dct["$timestamp"])
return Timestamp(ms / 1000, ms % 1000 * 1000)
except ValueError:
dt = datetime.datetime.strptime(dct["$timestamp"],
"%Y-%m-%d-%H.%M.%S.%f")
secs = long_type(time.mktime(dt.timetuple()))
return Timestamp(secs, dt.microsecond)
if "$regex" in dct:
flags = 0
# PyMongo always adds $options but some other tools may not.
for opt in dct.get("$options", ""):
flags |= _RE_OPT_TABLE.get(opt, 0)
if compile_re:
return re.compile(dct["$regex"], flags)
else:
return Regex(dct["$regex"], flags)
if "$minKey" in dct:
return MinKey()
if "$maxKey" in dct:
return MaxKey()
if "$binary" in dct:
if isinstance(dct["$type"], int):
dct["$type"] = "%d" % dct["$type"]
subtype = int(dct["$type"])
return Binary(base64.b64decode(dct["$binary"].encode()), subtype)
if "$code" in dct:
return Code(dct["$code"], dct.get("$scope"))
if bson.has_uuid() and "$uuid" in dct:
return bson.uuid.UUID(dct["$uuid"])
return dct
def getEndangeredSpecies():
global client, db, continentEntry
continent = endangeredEntry.get()
res = 0
list = []
Species_collection = db.get_collection("endangered_species")
for post in Species_collection.find({"continent": continent}, {
"_id": 0,
"animals_endangered": 1,
"examples": 2
}):
res = post["animals_endangered"]
list = post["examples"]
if continent == "":
map = Code("function () {"
" emit('endangered', this.animals_endangered); "
"}")
reduce = Code("function(key, values) {"
" return Array.sum(values);"
"}")
res = Species_collection.map_reduce(map=map,
reduce=reduce,
out=SON([('inline', 1)]))
res = res["results"][0]["value"]
ans = "Total endangered species: " + str(int(res))
label = Label(window, text=ans)
label.grid(column=3, row=7)
return
examples = ", ".join(list)
examples = "counting: " + examples
ans = "Total endangered species: " + str(int(res))
label = Label(window, text=ans)
label.grid(column=3, row=7)
examplelabel = Label(window, text=examples)
# examplelabel.grid_forget()
examplelabel.grid(column=3, row=8)
def rq1_mapred_handler(event, context):
start_time = timeit.default_timer()
client = MongoClient(MONGO_SERVER, MONGO_PORT)
db = client.stackoverflow
rq1_map = Code("""
function () {
if (this.PostTypeId == 1) {
emit(this.OwnerUserId, 1);
}
}
""")
rq1_reduce = Code('''
function (key, values)
{
return Array.sum(values);
}
''')
db.posts.map_reduce(rq1_map, rq1_reduce, out='rq1')
rq1_map2 = Code("""
function () {
emit(this.Id, 0);
}
""")
rq1_reduce2 = Code("""
function (key, values) {
return Array.sum(values);
}
""")
db.users.map_reduce(rq1_map2, rq1_reduce2, out={'reduce': 'rq1'})
rq1MR = list(db.rq1.find())
elapsed = timeit.default_timer() - start_time
return make_response(rq1MR, elapsed)
def execute(trial=False):
print("starting data retrieval")
startTime = datetime.datetime.now()
client = dml.pymongo.MongoClient()
repo = client.repo
print("repo: ", repo)
repo.authenticate('jzhou94_katerin', 'jzhou94_katerin')
map_function_avg_earnings = Code('''function() {
if (this.postal*1 > 2100 && this.postal*1 < 2300 && this.title == "Police Officer")
emit(this.postal, {tot:this.total_earnings, n: 1, avg: this.total_earnings});
}''')
reduce_function_avg_earnings = Code('''function(k, vs) {
var total = 0;
var counts = 0;
for (var i = 0; i < vs.length; i++)
total += (vs[i].tot*1);
for (var i = 0; i < vs.length; i++)
counts += vs[i].n;
return {tot:total.toFixed(2), n: counts, avg: (total/counts).toFixed(2)};
}''')
repo.dropPermanent('jzhou94_katerin.avg_earnings')
repo.createPermanent('jzhou94_katerin.avg_earnings')
if trial == True:
repo['jzhou94_katerin.avg_earnings'].insert(
repo.jzhou94_katerin.employee_earnings.find().limit(100))
repo.jzhou94_katerin.avg_earnings.map_reduce(
map_function_avg_earnings, reduce_function_avg_earnings,
'jzhou94_katerin.avg_earnings')
else:
repo.jzhou94_katerin.employee_earnings.map_reduce(
map_function_avg_earnings, reduce_function_avg_earnings,
'jzhou94_katerin.avg_earnings')
repo.logout()
endTime = datetime.datetime.now()
return {"start": startTime, "end": endTime}
def get_online_report():
"""获取在线报告"""
ses = get_conn("gps_info")
now = datetime.datetime.now()
year = now.year
month = now.month
begin = get_datetime_from_str("{}-{}-1 0:0:0".format(year, month))
# begin = now - datetime.timedelta(days=30)
query = {"time": {"$gte": begin}}
s = {"time": -1}
out = "online_report_result" # 保存数据的表,每次map_reduce都会提前清空这个表
map_func = Code("""
function(){
emit(this.user_id.$id, 1);
}
""")
reduce_func = Code("""
function(key, values){
return Array.sum(values);
}
""")
result_conn = ses.map_reduce(map=map_func,
reduce=reduce_func,
query=query,
sort=s,
out=out,
full_response=False)
res = result_conn.find(filter=dict())
count_dict = {x['_id']: int(x['value']) for x in res}
ids = list(count_dict.keys())
ses = get_conn("user_info")
f = {"_id": {"$in": ids}}
s = [("last_update", -1)]
users = ses.find(filter=f, sort=s)
res = list()
for user in users:
user_id = user['_id']
temp = to_flat_dict(user)
temp['count'] = count_dict[user_id]
res.append(temp)
res.sort(key=lambda obj: obj['count'], reverse=True)
return res
def get(self):
map = Code("function () {"
" var date = new Date(this.created_t*1000);"
" var parsedDateMonth = date.getMonth();"
" var parsedDateYear = date.getFullYear();"
" var saltLevelsValue = null;"
" var fatLevelsValue = null;"
" var saturatedfatLevelsValue = null;"
" var sugarsLevelsValue = null;"
" if(!parsedDateYear || !parsedDateMonth) return;"
" if(this.hasOwnProperty('nutrient_levels')) {"
" saltLevelsValue = this.nutrient_levels.salt;"
" fatLevelsValue = this.nutrient_levels.fat;"
" saturatedfatLevelsValue = this['nutrient_levels']['saturated-fat'];"
" sugarsLevelsValue = this.nutrient_levels.sugars;"
" } else { saltLevelsValue = null; fatLevelsValue = null; saturatedfatLevelsValue = null, sugarsLevelsValue = null}"
" emit({year : parsedDateYear, month : parsedDateMonth, saltlevels : saltLevelsValue, fatlevels : fatLevelsValue, saturatedfatlevels : saturatedfatLevelsValue, sugarslevels : sugarsLevelsValue}, {count:1});"
"};")
reduce = Code("function (key, values) {"
" var count = 0;"
" var ret = {count : 0};"
" for (index in values) {"
" ret.count += values[index].count;"
" }"
" return ret;"
" };")
listRes = []
result = mongo.db.products.map_reduce(map, reduce, "stats_products")
for doc in result.find():
res = {}
res['dateyear'] = doc['_id']['year']
res['datemonth'] = doc['_id']['month']
res['count'] = doc['value']['count']
res['saltlevels'] = doc['_id']['saltlevels']
res['saturatedfatlevels'] = doc['_id']['saturatedfatlevels']
res['sugarslevels'] = doc['_id']['sugarslevels']
res['fatlevels'] = doc['_id']['fatlevels']
listRes.append(res)
return listRes
def word_count(coll):
result = coll.find({}, {"text": 1}).limit(2)
# pprint.pprint(list(result) )
mapper = Code("""
function () {
txt=this.text ;
wrds = txt.split(" ");
for ( i=0 ; i < wrds.length ; i++) {
if ( wrds[i].length >1 ) {
wrd=wrds[i].toLowerCase().replace('"|,',' ');
wrd=wrd.replace('(\\n)+','');
wrd=wrd.replace('.|;|,)/gi','');
emit(wrd, 1 );
}
}
}
""")
reducer = Code("""
function (key, values) {
result = {
count : 0
};
values.forEach ( function(v) { result.count += 1; } )
return {result};
}
""")
print("\n\n")
pprint.pprint(coll)
result = coll.map_reduce(mapper,
reducer,
"word_cnt",
query={
"sender": "*****@*****.**"
}).find()
for doc in result:
print(doc)
def _map_reduce(self, coll, mapreduce, spec=None):
"""
Perform map/reduce operation over DAS cache using provided
collection, mapreduce name and optional conditions.
"""
self.logger.debug("(%s, %s)" % (mapreduce, spec))
record = find_one(self.mrcol, {'name': mapreduce})
if not record:
raise Exception("Map/reduce function '%s' not found" % mapreduce)
fmap = record['map']
freduce = record['reduce']
if spec:
result = coll.map_reduce(Code(fmap), Code(freduce), query=spec)
else:
result = coll.map_reduce(Code(fmap), Code(freduce))
msg = "found %s records in %s" % (result.count(), result.name)
self.logger.info(msg)
self.logger.debug(fmap)
self.logger.debug(freduce)
return result
def photographer_shots_count():
mapper = Code("""
function() {
emit(this.photographer, 1);
}
""")
reducer = Code("""
function(k, vs) {
return Array.sum(vs);
}
""")
raw = mongo_client.photo.photo.map_reduce(mapper, reducer,
"photographer_phs").find()
res = []
for kv in raw:
res.append({
'name': Photographer.find_by_id(kv['_id'])['name'],
'count': kv['value']
})
return res
def updateDocFreq(patDB, outColName='corpusDict'):
map = Code(open('fresh_docFreqMap.js').read())
reduce = Code(open('fresh_docFreqReduce.js').read())
# finIDF will calculate IDF scores, so I have to pass it
# the number of total docs in a crafty way.
size = patDB.patns.count()
# replace all instances of TOTALDOCS with size in docFreqFinalize.js
finIDF = open('docFreqFinalize.js').read()
finIDF = finIDF.replace('TOTALDOCS', str(size))
# patDB.patns.map_reduce(map, reduce, outColName, finalize=finIDF)
# can either reduce into outColName or replace it. I like replace for now
## I changed 'out' from 'replace' to 'reduce', that should combine the results
## uh-oh, I'm not sure if that works, because of the way docFreqFinalize works
patDB.patns.map_reduce(map,
reduce,
out={'reduce': outColName},
finalize=finIDF)
def setUp(self):
"""Set up class."""
from bson.code import Code
self.encoder = GoodJSONEncoder()
self.expected = {
"code": "console.log('HAAAAAAAAHHHH!!!')",
"scope": {
"data": "test"
}
}
self.data = Code(**self.expected)
def test__group3(self):
reducer=Code("""
function(obj, result) {result.count+=1 }
""")
conditions = {
'foo':{'$in':[self._id1]},
}
self.cmp.compare.group(key=['foo'],
condition=conditions,
initial={"count": 0},
reduce=reducer)
def cutoffSL(doc, cutoff):
m = Code("function(){ emit( this.len , { count : 1 } );}")
r = Code(
"function (key, values) { var count = 0; values.forEach(function (v) {count += v.count;}); return {count: count}; }"
)
if Pippies.count() > 0:
lens = dict([(x['_id'], int(x['value']['count']))
for x in Pippies.map_reduce(
m, r, 'cutoff sparkline', query={
'docs': doc._id
}).find()])
else:
lens = {}
if lens.keys():
return [
str(lens[x]) if x in lens else '0'
for x in xrange(int(max(lens.keys()) + 1))
][4:cutoff]
else:
return []
def sum(self, field):
"""Sum over the values of the specified field.
:param field: the field to sum over; use dot-notation to refer to
embedded document fields
"""
map_func = """
function() {
var path = '{{~%(field)s}}'.split('.'),
field = this;
for (p in path) {
if (typeof field != 'undefined')
field = field[path[p]];
else
break;
}
if (field && field.constructor == Array) {
field.forEach(function(item) {
emit(1, item||0);
});
} else if (typeof field != 'undefined') {
emit(1, field||0);
}
}
""" % dict(field=field)
reduce_func = Code("""
function(key, values) {
var sum = 0;
for (var i in values) {
sum += values[i];
}
return sum;
}
""")
mr_future = self.inline_map_reduce(map_func, reduce_func)
future = get_future(self)
def sum_cb(mr_future):
results = mr_future.result()
for result in results:
r = result.value
break
else:
r = 0
future.set_result(r)
mr_future.add_done_callback(sum_cb)
return future
def top_n_map(n):
return Code('''
function() {
var to_return = Math.min(%d, this.sorted_text.length);
var out_arr = [];
for (var i = 0; i < to_return; i++) {
out_arr[i] = this.sorted_text[i]['tf-idf']
};
emit("tf-idf", {'vals':out_arr})
};''' % n)
return out
def mapFunction(self):
self.mapcode = Code("function() {"
" var key = this.nppes_provider_state;"
" var value = {"
" count: 1,"
" claim: this.average_submitted_chrg_amt,"
" payment: this.average_Medicare_payment_amt"
" };"
" emit(key, value);"
"};")
return self.mapcode
def mapTopDriver(self):
mapper = Code("""
function() {
var key = this.driver.name;
var value = {count : 1};
emit(key, value);
};
""")
reducer = Code("""
function (key, values) {
var count = 0;
for(var i in values){
count += values[i].count;
}
return {count: count};
};
""")
result = self.db.order.map_reduce(mapper, reducer, "result")
res = list(result.find())
print res
def mapTopCompany(self):
mapper = Code("""
function() {
var key = this.Companies.Name;
var value = {count : 1};
emit(key, value);
};
""")
reducer = Code("""
function(key, values) {
var count = 0;
for(var i in values) {
count += values[i].count;
}
return {count: count};
};
""")
result = self.db.Flights.map_reduce(mapper, reducer, 'result')
res = list(result.find())
print res
def __init__(self):
self.mongo = Database()
self.map = Code(
"function() {"
" this.tags.forEach("
" function(tag) {"
" emit(tag, 1);"
" }"
" )"
"}"
)
self.reduce = Code(
"function(key, values) {"
" var total = 0;"
" for (var i=0; i<values.length; i++) {"
" total += values[i];"
" }"
" return total;"
"}"
)
def reduceFunction(self):
self.reducecode = Code("function(keyState, countStVals) {"
" reduceVal = {count: 0, claim: 0, payment: 0};"
" for (var provider = 0; provider < countStVals.length; provider++) {"
" reduceVal.count += countStVals[provider].count;"
" reduceVal.claim += countStVals[provider].claim;"
" reduceVal.payment += countStVals[provider].payment;"
" };"
" return reduceVal;"
"};")
return self.reducecode
def test_qop_ne_6(monty_find, mongo_find):
docs = [
{"a": [{"b": Code("a")}]},
]
spec = {"a.b": {"$ne": "a"}}
monty_c = monty_find(docs, spec)
mongo_c = mongo_find(docs, spec)
assert mongo_c.count() == 1
assert monty_c.count() == mongo_c.count()