Version 0.7.0. Extended tweet model

HISTORY.rst

@@ -110,3 +110,18 @@ Version 0.6.2:
 No changes, just for trying new hooks.
 Version 0.6.3 & 0.6.4:
 Updated tweetmanager dependency.
+
+0.7.0 (2019-10-14)
+------------------
+
+Main differences with last major release:
+
+ * Accept embedded tweets in "quoted_status" and "retweeted_status".
+ * Accept some other new fields found in some tweets (although they don't seem
+   to appear in the Twitter documentation) such as "full_text" or
+   "display_text_range"
+ * Accept any other argument without throwing an error, but don't do shit with
+   them. Thus, if unexpected and undocumented fields arrive, they won't be used
+   to instantiate the Tweet object, but it won't crash either.
+ * Some reformatting
+ * 

requirements.txt

 alabaster==0.7.12
 argh==0.26.2
 arrow==0.13.1
+atomicwrites==1.3.0
+attrs==19.2.0
 Babel==2.6.0
 binaryornot==0.4.4
 bleach==3.1.0
@@ -18,14 +20,16 @@ flake8==3.5.0
 future==0.17.1
 idna==2.8
 imagesize==1.1.0
+importlib-metadata==0.23
 Jinja2==2.10
 jinja2-time==0.2.0
 MarkupSafe==1.1.1
 mccabe==0.6.1
+more-itertools==7.2.0
 packaging==19.0
 pathtools==0.1.2
 pkginfo==1.5.0.1
-pluggy==0.9.0
+pluggy==0.13.0
 poyo==0.4.2
 py==1.8.0
 pycodestyle==2.3.1
@@ -34,12 +38,13 @@ PyGithub==1.43.5
 Pygments==2.3.1
 PyJWT==1.7.1
 pyparsing==2.3.1
-pysnooper==0.0.38
+PySnooper==0.0.38
+pytest==5.2.1
 python-dateutil==2.8.0
 python-gitlab==1.8.0
 pytz==2018.9
 pyupio==1.0.2
-PyYAML==5.1
+PyYAML==5.1.2
 readme-renderer==24.0
 requests==2.21.0
 requests-toolbelt==0.9.1
@@ -56,6 +61,8 @@ twine==1.12.1
 urllib3==1.24.1
 virtualenv==16.4.3
 watchdog==0.9.0
+wcwidth==0.1.7
 webencodings==0.5.1
 whichcraft==0.5.2
-wrapt==1.11.1
+wrapt==1.11.2
+zipp==0.6.0

setup.py

@@ -51,6 +51,6 @@ setup(
     test_suite='tests',
     tests_require=test_requirements,
     url='https://github.com/Serbaf/tweet_model',
-    version='0.6.4',
+    version='0.7.0',
     zip_safe=False,
 )

tweet_model/tweet.py

@@ -25,7 +25,8 @@ class Tweet():
                  filter_level=None, lang=None, matching_rules=None,
                  current_user_retweet=None, scopes=None,
                  withheld_copyright=None, withheld_in_countries=None,
-                 withheld_scope=None, geo=None,
+                 withheld_scope=None, geo=None, full_text=None,
+                 display_text_range=None,
 
 
                  # User object
@@ -54,6 +55,9 @@ class Tweet():
                  user__withheld_in_countries=None, user__withheld_scope=None,
                  user__is_translator=None, user__following=None,
                  user__notifications=None,
+                 user__is_translation_enabled=None,
+                 user__has_extended_profile=None,
+                 user__translator_type=None,
 
 
                  # Coordinates object
@@ -134,22 +138,42 @@ class Tweet():
                  extended_entities__media__source_status_id_str=None,
                  extended_entities__media__type=None,
                  extended_entities__media__url=None,
+                 extended_entities__media__sizes__thumb__h=None,
+                 extended_entities__media__sizes__thumb__resize=None,
+                 extended_entities__media__sizes__thumb__w=None,
+                 extended_entities__media__sizes__large__h=None,
+                 extended_entities__media__sizes__large__resize=None,
+                 extended_entities__media__sizes__large__w=None,
+                 extended_entities__media__sizes__medium__h=None,
+                 extended_entities__media__sizes__medium__resize=None,
+                 extended_entities__media__sizes__medium__w=None,
+                 extended_entities__media__sizes__small__h=None,
+                 extended_entities__media__sizes__small__resize=None,
+                 extended_entities__media__sizes__small__w=None,
+
+
+                 # Metadata object
+                 metadata__result_type=None,
+                 metadata__iso_language_code=None,
 
                  # Additional fields (not from the Tweeter model)
                  polarity=None,
-                 trtext=None
+                 trtext=None,
+
+                 # Ignore possible additional arguments
+                 **kwargs
                  ):
 
         # Basic attributes
         try:
             self.created_at =\
                 datetime.datetime.strptime(created_at, "%Y-%m-%d %H:%M:%S")
-        except Exception as exc:
+        except Exception:
             self.created_at = created_at
 
         try:
             self.id = int(id)
-        except Exception as exc:
+        except Exception:
             self.id = id
 
         if type(truncated) is str:
@@ -162,17 +186,17 @@ class Tweet():
 
         try:
             self.in_reply_to_status_id = int(in_reply_to_status_id)
-        except Exception as exc:
+        except Exception:
             self.in_reply_to_status_id = in_reply_to_status_id
 
         try:
             self.in_reply_to_user_id = int(in_reply_to_user_id)
-        except Exception as exc:
+        except Exception:
             self.in_reply_to_user_id = in_reply_to_user_id
 
         try:
             self.quoted_status_id = int(quoted_status_id)
-        except Exception as exc:
+        except Exception:
             self.quoted_status_id = quoted_status_id
 
         if type(is_quote_status) is str:
@@ -184,15 +208,17 @@ class Tweet():
             self.is_quote_status = is_quote_status
         try:
             self.retweet_count = int(retweet_count)
-        except Exception as exc:
+        except Exception:
             self.retweet_count = retweet_count
 
         try:
             self.favorite_count = int(favorite_count)
-        except Exception as exc:
+        except Exception:
             self.favorite_count = favorite_count
 
         self.text = text
+        if full_text is not None:
+            self.text = full_text
         self.source = source
         self.in_reply_to_screen_name = in_reply_to_screen_name
 
@@ -215,19 +241,20 @@ class Tweet():
         self.withheld_in_countries = withheld_in_countries
         self.withheld_scope = withheld_scope
         self.geo = geo
+        self.display_text_range = display_text_range
 
         # User object
         self.user = {}
 
         try:
             self.user["id"] = int(user__id)
-        except Exception as exc:
+        except Exception:
             self.user["id"] = user__id
 
         try:
             self.user["created_at"] = datetime.datetime.strptime(
                 user__created_at, "%Y-%m-%d %H:%M:%S")
-        except Exception as exc:
+        except Exception:
             self.user["created_at"] = user__created_at
 
         if type(user__verified) is str:
@@ -240,27 +267,27 @@ class Tweet():
 
         try:
             self.user["followers_count"] = int(user__followers_count)
-        except Exception as exc:
+        except Exception:
             self.user["followers_count"] = user__followers_count
 
         try:
             self.user["friends_count"] = int(user__friends_count)
-        except Exception as exc:
+        except Exception:
             self.user["friends_count"] = user__friends_count
 
         try:
             self.user["listed_count"] = int(user__listed_count)
-        except Exception as exc:
+        except Exception:
             self.user["listed_count"] = user__listed_count
 
         try:
             self.user["favourites_count"] = int(user__favourites_count)
-        except Exception as exc:
+        except Exception:
             self.user["favourites_count"] = user__favourites_count
 
         try:
             self.user["statuses_count"] = int(user__statuses_count)
-        except Exception as exc:
+        except Exception:
             self.user["statuses_count"] = user__statuses_count
 
         if type(user__geo_enabled) is str:
@@ -309,6 +336,26 @@ class Tweet():
         self.user["following"] = user__following
         self.user["notifications"] = user__notifications
 
+        if type(user__is_translation_enabled) is str:
+            if user__is_translation_enabled == "True":
+                self.user["is_translation_enabled"] = True
+            elif user__is_translation_enabled == "False":
+                self.user["is_translation_enabled"] = False
+        else:
+            self.user["is_translation_enabled"] =\
+                user__is_translation_enabled
+
+        if type(user__has_extended_profile) is str:
+            if user__has_extended_profile == "True":
+                self.user["has_extended_profile"] = True
+            elif user__has_extended_profile == "False":
+                self.user["has_extended_profile"] = False
+        else:
+            self.user["has_extended_profile"] =\
+                user__has_extended_profile
+
+        self.user["translator_type"] = user__translator_type
+
         # Coordinates object
         self.coordinates = {}
         self.coordinates["type"] = coordinates__type
@@ -316,7 +363,7 @@ class Tweet():
             self.coordinates["coordinates"] =\
                 [float(coords)
                     for coords in json.loads(coordinates__coordinates)]
-        except Exception as exc:
+        except Exception:
             self.coordinates["coordinates"] = coordinates__coordinates
 
         # Place object
@@ -339,7 +386,7 @@ class Tweet():
                     for y in range(len(coords[x]))]
                     for x in range(len(coords))]
 
-        except Exception as exc:
+        except Exception:
             self.place["bounding_box"]["coordinates"] =\
                 place__bounding_box__coordinates
         self.place["bounding_box"]["type"] = place__bounding_box__type
@@ -351,7 +398,7 @@ class Tweet():
         try:
             self.entities["hashtags"]["text"] =\
                 json.loads(entities__hashtags__text)
-        except Exception as exc:
+        except Exception:
             self.entities["hashtags"]["text"] = entities__hashtags__text
 
         self.entities["hashtags"]["indices"] = entities__hashtags__indices
@@ -360,7 +407,7 @@ class Tweet():
         try:
             self.entities["media"]["media_url"] =\
                 json.loads(entities__media__media_url)
-        except Exception as exc:
+        except Exception:
             self.entities["media"]["media_url"] = entities__media__media_url
         self.entities["media"]["display_url"] = entities__media__display_url
         self.entities["media"]["expanded_url"] = entities__media__expanded_url
@@ -410,7 +457,7 @@ class Tweet():
         try:
             self.entities["urls"]["expanded_url"] =\
                 json.loads(entities__urls__expanded_url)
-        except Exception as exc:
+        except Exception:
             self.entities["urls"]["expanded_url"] =\
                 entities__urls__expanded_url
         self.entities["urls"]["display_url"] = entities__urls__display_url
@@ -430,7 +477,7 @@ class Tweet():
         try:
             self.entities["user_mentions"]["screen_name"] =\
                 json.loads(entities__user_mentions__screen_name)
-        except Exception as exc:
+        except Exception:
             self.entities["user_mentions"]["screen_name"] =\
                 entities__user_mentions__screen_name
         self.entities["user_mentions"]["id"] = entities__user_mentions__id
@@ -479,6 +526,40 @@ class Tweet():
         self.extended_entities["media"]["type"] =\
             extended_entities__media__type
         self.extended_entities["media"]["url"] = extended_entities__media__url
+        self.extended_entities["media"]["sizes"] = {}
+        self.extended_entities["media"]["sizes"]["thumb"] = {}
+        self.extended_entities["media"]["sizes"]["large"] = {}
+        self.extended_entities["media"]["sizes"]["medium"] = {}
+        self.extended_entities["media"]["sizes"]["small"] = {}
+        self.extended_entities["media"]["sizes"]["thumb"]["h"] =\
+            extended_entities__media__sizes__thumb__h
+        self.extended_entities["media"]["sizes"]["thumb"]["resize"] =\
+            extended_entities__media__sizes__thumb__resize
+        self.extended_entities["media"]["sizes"]["thumb"]["w"] =\
+            extended_entities__media__sizes__thumb__w
+        self.extended_entities["media"]["sizes"]["large"]["h"] =\
+            extended_entities__media__sizes__large__h
+        self.extended_entities["media"]["sizes"]["large"]["resize"] =\
+            extended_entities__media__sizes__large__resize
+        self.extended_entities["media"]["sizes"]["large"]["w"] =\
+            extended_entities__media__sizes__large__w
+        self.extended_entities["media"]["sizes"]["medium"]["h"] =\
+            extended_entities__media__sizes__medium__h
+        self.extended_entities["media"]["sizes"]["medium"]["resize"] =\
+            extended_entities__media__sizes__medium__resize
+        self.extended_entities["media"]["sizes"]["medium"]["w"] =\
+            extended_entities__media__sizes__medium__w
+        self.extended_entities["media"]["sizes"]["small"]["h"] =\
+            extended_entities__media__sizes__small__h
+        self.extended_entities["media"]["sizes"]["small"]["resize"] =\
+            extended_entities__media__sizes__small__resize
+        self.extended_entities["media"]["sizes"]["small"]["w"] =\
+            extended_entities__media__sizes__small__w
+
+        # Metadata object
+        self.metadata = {}
+        self.metadata["result_type"] = metadata__result_type
+        self.metadata["iso_language_code"] = metadata__iso_language_code
 
         # Additional fields
         self.polarity = polarity
@@ -506,7 +587,7 @@ class Tweet():
                 pass
         return tweet_subset
 
-    def as_short_json(self, dictionary: Dict=None) -> Dict:
+    def as_short_json(self, dictionary: Dict = None) -> Dict:
         """
         Return the Tweet object in a short JSON-like representation
         but without all the null key-value pairs

tweet_model/utils.py

 import logging
+import re
 from typing import Union, Dict, List, Generator
 
 from tweet_model.tweet import Tweet
@@ -24,7 +25,7 @@ def get_tweet_from_csv_raw_line(header, line):
     return get_tweet_from_csv_line(header_fields, line_fields)
 
 
-def get_tweet_from_csv_line(header_fields, line_fields):
+def get_tweet_from_csv_line_OLD(header_fields, line_fields):
     """
     Given the fields of a CSV line and header, the function instances a Tweet
     object with all the non-empty attributes initialized to the values
@@ -40,6 +41,45 @@ def get_tweet_from_csv_line(header_fields, line_fields):
     return Tweet(**tweet_contents)
 
 
+def get_tweet_from_csv_line(header_fields, line_fields):
+    """
+    Given the fields of a CSV line and header, the function instances a Tweet
+    object with all the non-empty attributes initialized to the values
+    indicated in the CSV entry.
+    Accepts embedded tweets in "quoted_status" and "retweeted_statusW
+    Returns a Tweet object
+    """
+
+    tweet_contents = {}
+    quoted_contents = {}
+    retweeted_contents = {}
+    quoted_pattern = re.compile(r"^(quoted_status\.)(.*)$")
+    retweeted_pattern = re.compile(r"^(retweeted_status\.)(.*)$")
+
+    for i in range(len(line_fields)):
+        if line_fields[i] != '':
+            quoted_match = quoted_pattern.match(header_fields[i])
+            retweeted_match = retweeted_pattern.match(header_fields[i])
+            if quoted_match is not None:
+                quoted_contents[
+                    quoted_match.group(2).replace(".", "__")] =\
+                    line_fields[i]
+            elif retweeted_match is not None:
+                retweeted_contents[
+                    retweeted_match.group(2).replace(".", "__")] =\
+                    line_fields[i]
+            else:
+                tweet_contents[header_fields[i].replace(".", "__")] =\
+                    line_fields[i]
+
+    if bool(quoted_contents):    # Check non empty
+        tweet_contents["quoted_status"] = Tweet(**quoted_contents)
+    if bool(retweeted_contents):    # Check non empty
+        tweet_contents["retweeted_status"] = Tweet(**retweeted_contents)
+
+    return Tweet(**tweet_contents)
+
+
 def get_tweets_from_csv(csv_file):
     """
     Take one argument: a path pointing to a valid CSV file.
@@ -67,8 +107,8 @@ def get_tweets_from_csv(csv_file):
         for component in field_components:
             error_string += component
             if (checking_dict is None) or (component not in checking_dict):
-                logger.error('The field in the header ' + error_string +
-                             'is not a valid element of a Tweet')
+                logger.error('The field in the header ' + error_string
+                             + 'is not a valid element of a Tweet')
                 raise NotValidTweetError("Header contains field which doesn't"
                                          + " belong to tweet specification: "
                                          + error_string)
@@ -86,8 +126,8 @@ def get_tweets_from_csv(csv_file):
 
 def get_tweet_collection_fields_subset(
         tweet_collection: Union[List[Tweet], Generator[Tweet, None, None]],
-        fields: List[str]
-        ) -> Generator[Dict, None, None]:
+        fields: List[str])\
+        -> Generator[Dict, None, None]:
     """
     Given a list of Tweet objects, keep just the specified fields and
     return a generator of dicts with just the information specified