def uploadDocument(uv_task): task_tag = "DOCUMENTCLOUD UPLOAD" print "\n\n************** %s [START] ******************\n" % task_tag print "uploading doc %s to DocumentCloud" % uv_task.doc_id uv_task.setStatus(302) from lib.Worker.Models.cp_documentcloud_client import CompassDocumentCloudClient from lib.Worker.Models.uv_document import UnveillanceDocument from conf import DEBUG document = UnveillanceDocument(_id=uv_task.doc_id) if document is None: print "\n\n************** %s [ERROR] ******************\n" % task_tag print "Document is None" uv_task.fail() return if not hasattr(uv_task, "auth_string"): print "\n\n************** %s [ERROR] ******************\n" % task_tag print "DocumentCloud upload needs an auth string" uv_task.fail() return dc_client = CompassDocumentCloudClient(auth_string=uv_task.auth_string) upload = dc_client.upload(document) if DEBUG: print upload if upload is None: print "\n\n************** %s [ERROR] ******************\n" % task_tag print "DocumentCloud upload needs an auth string" uv_task.fail() return document.dc_id = upload['id'] document.save() document.addCompletedTask(uv_task.task_path) uv_task.finish() print "\n\n************** %s [END] ******************\n" % task_tag print "Uploaded document %s" % document._id
def evaluateText(task):
task_tag = "TEXT EVALUATION"
print "\n\n************** %s [START] ******************\n" % task_tag
print "evaluating text at %s" % task.doc_id
task.setStatus(302)
from lib.Worker.Models.uv_document import UnveillanceDocument
from conf import DEBUG
from vars import MIME_TYPE_TASKS
document = UnveillanceDocument(_id=task.doc_id)
"""
limited choices: json, pgp, or txt
"""
if hasattr(task, "text_file"):
content = document.loadAsset(task.text_file)
else:
content = document.loadFile(document.file_name)
if content is None:
print "no text to evaluate :("
print "\n\n************** %s [ERROR] ******************\n" % task_tag
task.fail()
return
new_mime_type = None
import json
try:
json_txt = json.loads(content)
new_mime_type = "application/json"
print "THIS IS JSON"
except Exception as e:
print "NOT JSON: %s" % e
task_path = None
if new_mime_type is not None:
document.mime_type = new_mime_type
document.save()
if document.mime_type in MIME_TYPE_TASKS.keys():
task_path = MIME_TYPE_TASKS[document.mime_type][0]
else:
try:
from lib.Core.Utils.funcs import cleanLine
from vars import ASSET_TAGS
txt_json = []
txt_pages = []
line_count = 0
# this is arbitrary
MAX_LINES_PER_PAGE = 80
for line in content.splitlines():
txt_pages.append(cleanLine(line))
line_count += 1
if line_count == MAX_LINES_PER_PAGE:
txt_json.append(" ".join(txt_pages))
txt_pages = []
line_count = 0
txt_json.append(" ".join(txt_pages))
document.total_pages = len(txt_json)
document.save()
asset_path = document.addAsset(txt_json, "doc_texts.json", as_literal=False,
description="jsonified text of original document, segment by segment",
tags=[ASSET_TAGS['TXT_JSON']])
from lib.Worker.Models.uv_text import UnveillanceText
uv_text = UnveillanceText(inflate={
'media_id' : document._id,
'searchable_text' : txt_json,
'file_name' : asset_path
})
document.text_id = uv_text._id
document.save()
except Exception as e:
if DEBUG:
print "ERROR HERE GENERATING DOC TEXTS:"
print e
document.addCompletedTask(task.task_path)
task.finish()
task.routeNext()
print "\n\n************** %s [END] ******************\n" % task_tag
def massageJ3M(task):
task_tag = "MASSAGING J3M"
print "\n\n************** %s [START] ******************\n" % task_tag
print "massaging j3m at %s" % task.doc_id
task.setStatus(302)
from lib.Worker.Models.uv_document import UnveillanceDocument
from conf import DEBUG
from vars import ASSET_TAGS
media = UnveillanceDocument(_id=task.doc_id)
if media is None:
print "DOC IS NONE"
print "\n\n************** %s [ERROR] ******************\n" % task_tag
task.fail()
return
if hasattr(task, "j3m_name"):
j3m_name = task.j3m_name
else:
j3m_name = "j3m.json"
j3m = media.loadAsset(j3m_name)
if j3m is None:
print "J3M IS NONE"
print "\n\n************** %s [ERROR] ******************\n" % task_tag
task.fail()
return
from json import loads
try:
j3m = loads(j3m)
except Exception as e:
print "J3M IS INVALID"
print "\n\n************** %s [ERROR] ******************\n" % task_tag
task.fail(status=412)
return
try:
media.date_created = j3m['genealogy']['dateCreated']
media.saveFields("date_created")
except KeyError as e:
print "J3M HAS NO DATE CREATED: %s" % e
print "\n\n************** %s [WARN] ******************\n" % task_tag
from hashlib import sha1
try:
j3m['public_hash'] = sha1("".join(
[j3m['genealogy']['createdOnDevice'],
"".join(j3m['genealogy']['hashes'])])).hexdigest()
except KeyError as e:
if DEBUG:
print "no key %s" % e
pass
if 'data' in j3m.keys():
try:
location = j3m['data']['exif']['location']
j3m['data']['exif'].update({
'location' : [location[1], location[0]]
})
except KeyError as e:
if DEBUG:
print "no key %s" % e
pass
try:
if type(j3m['data']['sensorCapture']) is list: pass
except KeyError as e:
if DEBUG: print "no key %s" % e
pass
if 'sensorCapture' in j3m['data'].keys():
for playback in j3m['data']['sensorCapture']:
if 'gps_coords' in playback['sensorPlayback'].keys():
try:
gps = str(playback['sensorPlayback']['gps_coords'])[1:-1].split(",")
if DEBUG:
print "REPLACING %s as geopoint" % gps
print type(gps)
playback['sensorPlayback'].update({
'gps_coords' : [float(gps[1]), float(gps[0])]
})
except Exception as e:
if DEBUG: print e
pass
if 'regionLocationData' in playback['sensorPlayback'].keys():
try:
gps = str(playback['sensorPlayback']['regionLocationData']['gps_coords'])
gps = gps[1:-1].split(",")
if DEBUG:
print "REPLACING %s as geopoint" % gps
playback['sensorPlayback']['regionLocationData'].update({
'gps_coords' : [float(gps[1]), float(gps[0])]
})
except Exception as e:
if DEBUG: print e
pass
if 'visibleWifiNetworks' in playback['sensorPlayback'].keys():
try:
for i,b in enumerate(playback['sensorPlayback']['visibleWifiNetworks']):
playback['sensorPlayback']['visibleWifiNetworks'][i].update({
'bt_hash' : sha1(b['bssid']).hexdigest()
})
except Exception as e:
if DEBUG: print e
pass
import os, json
from conf import getConfig
from lib.Core.Utils.funcs import b64decode
from lib.Worker.Utils.funcs import getFileType, unGzipBinary
searchable_text = []
if 'userAppendedData' in j3m['data'].keys():
try:
with open(os.path.join(getConfig('informacam.forms_root'), "forms.json"), 'rb') as F:
form_data = json.loads(F.read())['forms']
for udata in j3m['data']['userAppendedData']:
for aForms in udata['associatedForms']:
st_keys = aForms['answerData'].keys()
for f in form_data:
if f['namespace'] == aForms['namespace']:
try:
for mapping in f['mapping']:
try:
group = mapping.keys()[0]
key = aForms['answerData'][group].split(" ")
for m in mapping[group]:
if m.keys()[0] in key:
key[key.index(m.keys()[0])] = m[m.keys()[0]]
aForms['answerData'][group] = " ".join(key)
except KeyError as e:
if DEBUG: print "no key %s" % e
pass
except KeyError as e:
if DEBUG: print "no key %s" % e
pass
try:
idx = 0
for audio in f['audio_form_data']:
try:
while audio in st_keys: st_keys.remove(audio)
except Exception as e: pass
try:
audio_data = b64decode(
aForms['answerData'][audio])
if audio_data is None:
if DEBUG: print "could not unb64 audio"
continue
if getFileType(audio_data, as_buffer=True) != MIME_TYPES['gzip']:
if DEBUG: print "audio is not gzipped"
continue
audio_f = "audio_%d.3gp" % idx
idx += 1
media.addAsset(unGzipBinary(audio_data), audio_f,
tags=[ASSET_TAGS['A_3GP']],
description="3gp audio file from form")
'''
new_task=UnveillanceTask(inflate={
'task_path' : "Media.convert.audioConvert",
'doc_id' : media._id,
'formats' : ["3gp", "wav"],
'src_file' : "audio_%d.3gp" % idx,
'queue' : task.queue
})
new_task.run()
'''
aForms['answerData'][audio] = "audio_%d.wav"
except KeyError as e:
if DEBUG: print "no key %s" % e
pass
except KeyError as e:
if DEBUG: print "no key %s" % e
pass
if len(st_keys) > 0:
for key in st_keys:
searchable_text.append(aForms['answerData'][key])
except KeyError as e:
if DEBUG: print "no key %s" % e
pass
except IOError as e:
print "\n\n************** %s [WARN] ******************\n" % task_tag
if DEBUG: print "no forms to go over: %s" % e
except ValueError as e:
print "\n\n************** %s [WARN] ******************\n" % task_tag
if DEBUG: print "for some reason, forms.json is not legible?\n%s" % e
if media.addAsset(j3m, "j3m.json", as_literal=False) is False:
print "J3M COULD NOT BE ADDED"
print "\n\n************** %s [ERROR] ******************\n" % task_tag
task.fail()
return
from lib.Worker.Models.ic_j3m import InformaCamJ3M
j3m['media_id'] = media._id
if len(searchable_text) > 0:
j3m['searchable_text'] = searchable_text
j3m = InformaCamJ3M(inflate=j3m)
print "\n\n***NEW J3M CREATED***\n\n"
j3m.save()
media.j3m_id = j3m._id
media.save()
media.addCompletedTask(task.task_path)
task.routeNext()
task.finish()
print "\n\n************** %s [END] ******************\n" % task_tag
def verifySignature(task):
task_tag = "VERIFYING SIGNATURE"
print "\n\n************** %s [START] ******************\n" % task_tag
print "image preprocessing at %s" % task.doc_id
task.setStatus(302)
from lib.Worker.Models.uv_document import UnveillanceDocument
from conf import DEBUG
from vars import ASSET_TAGS
media = UnveillanceDocument(_id=task.doc_id)
if media is None:
print "DOC IS NONE"
print "\n\n************** %s [ERROR] ******************\n" % task_tag
task.fail()
return
sig = media.getAsset("j3m.sig", return_only="path")
j3m = media.getAsset("j3m.json", return_only="path")
if DEBUG:
print "j3m path: %s, sig path: %s" % (j3m, sig)
if sig is None or j3m is None:
err_msg = "NO SIGNATURE or J3M"
print err_msg
print "\n\n************** %s [ERROR] ******************\n" % task_tag
task.fail(message=err_msg)
return
import gnupg
from conf import getConfig
try:
gpg = gnupg.GPG(homedir=getConfig('gpg_homedir'))
except Exception as e:
print "ERROR INITING GPG"
print "\n\n************** %s [ERROR] ******************\n" % task_tag
task.fail()
return
media.j3m_verified = False
verified = gpg.verify_file(j3m, sig_file=sig)
if DEBUG:
print "verified fingerprint: %s" % verified.fingerprint
if verified.fingerprint is not None:
from json import loads
supplied_fingerprint = str(loads(
media.loadAsset("j3m.json"))['genealogy']['createdOnDevice'])
if verified.fingerprint.upper() == supplied_fingerprint.upper():
if DEBUG:
print "SIGNATURE VALID for %s" % verified.fingerprint.upper()
media.j3m_verified = True
media.save()
media.addCompletedTask(task.task_path)
task.routeNext()
task.finish()
print "\n\n************** %s [END] ******************\n" % task_tag
def getAssets(uv_task):
task_tag = "FETCHING DOCUMENTCLOUD ASSETS"
print "\n\n************** %s [START] ******************\n" % task_tag
print "getting DocumentCloud assets for %s" % uv_task.doc_id
uv_task.setStatus(412)
from lib.Worker.Models.cp_documentcloud_client import CompassDocumentCloudClient
from lib.Worker.Models.uv_document import UnveillanceDocument
from conf import DEBUG
document = UnveillanceDocument(_id=uv_task.doc_id)
if document is None:
print "\n\n************** %s [ERROR] ******************\n" % task_tag
print "Document is None"
return
if not hasattr(document, "dc_id"):
print "\n\n************** %s [ERROR] ******************\n" % task_tag
print "Document has not document cloud id!"
return
if not hasattr(uv_task, "auth_string"):
print "\n\n************** %s [ERROR] ******************\n" % task_tag
print "DocumentCloud upload needs an auth string"
return
dc_client = CompassDocumentCloudClient(auth_string=uv_task.auth_string)
dc_manifest = dc_client.download("documents/%s.json" % document.dc_id)
if dc_manifest is None:
print "\n\n************** %s [ERROR] ******************\n" % task_tag
print "No DocumentCloud manifest yet for %s." % document._id
return
document.addAsset(dc_manifest, "document_cloud_manifest.json", as_literal=False,
description="description of document on DocumentCloud",
tags=[ASSET_TAGS['DOC_CLOUD_MANIFEST'], ASSET_TAGS['DOC_CLOUD_DOC']])
dc_entities = dc_client.download("documents/%s/entities.json" % document.dc_id)
if dc_entities is None:
print "\n\n************** %s [WARN] ******************\n" % task_tag
print "No DocumentCloud entiteis yet for %s." % document._id
else:
entity_asset = document.addAsset(dc_entities, "document_cloud_entities.json",
as_literal=False, description="entites pulled from DocumentCloud",
tags=[ASSET_TAGS['DOC_CLOUD_ENTITIES'], ASSET_TAGS['DOC_CLOUD_DOC']])
from lib.Worker.Models.uv_text import UnveillanceText
if not hasattr(document, "text_id"):
text = UnveillanceText(inflate={
'file_name' : entity_asset,
'entities' : dc_entities['entities'],
'media_id' : document._id
})
document.text_id = text._id
document.save()
else:
text = UnveillanceText(_id=document.text_id)
text.entities = dc_entities['entities']
text.save()
document.addCompletedTask(uv_task.task_path)
uv_task.finish()
def verifyVisualContent(task):
task_tag = "VERIFYING VISUAL CONTENT"
print "\n\n************** %s [START] ******************\n" % task_tag
print "image preprocessing at %s" % task.doc_id
task.setStatus(302)
from lib.Worker.Models.uv_document import UnveillanceDocument
from conf import DEBUG
from vars import ASSET_TAGS
media = UnveillanceDocument(_id=task.doc_id)
if media is None:
print "DOC IS NONE"
print "\n\n************** %s [ERROR] ******************\n" % task_tag
task.fail()
return
j3m = media.loadAsset("j3m.json")
if j3m is None:
print "NO J3M AT ALL"
print "\n\n************** %s [ERROR] ******************\n" % task_tag
task.fail()
return
import os
from json import loads
from subprocess import Popen, PIPE
from conf import ANNEX_DIR, getConfig
from vars import MIME_TYPES
try:
supplied_hashes = loads(j3m)['genealogy']['hashes']
except KeyError as e:
print "NO HASHES"
print "\n\n************** %s [WARNING] ******************\n" % task_tag
task.finish()
return
media.media_verified = False
if media.mime_type == MIME_TYPES['image']:
cmd = ["java", "-jar",
os.path.join(getConfig('jpeg_tools_dir'), "JavaMediaHasher.jar"),
os.path.join(ANNEX_DIR, media.file_name)]
elif media.mime_type == MIME_TYPES['video']:
cmd = ["ffmpeg", "-y", "-i",
os.path.join(ANNEX_DIR, media.file_name), "-vcodec", "copy",
"-an", "-f", "md5", "-"]
p = Popen(cmd, stdout=PIPE, close_fds=True)
verified_hash = p.stdout.readline().strip().replace("MD5=", "")
p.stdout.close()
if type(supplied_hashes) is list:
for hash in supplied_hashes:
if type(hash) is unicode:
hash = str(hash)
if hash == verified_hash:
media.media_verified = True
media.save()
media.addCompletedTask(task.task_path)
task.routeNext()
task.finish()
print "\n\n************** %s [END] ******************\n" % task_tag
