🏴‍☠️ Welcome to scrape_reviews ⚒️

The easiest way to get reviews in your hands.

scrape_reviews is a module for scraping Google Maps reviews that aspires to become a package with a distinct module for each source of reviews (Airbnb, Booking, etc.).

The user can get reviews either by:

• scanning a specific place (e.g. Nauru, Myrina, Aspen, Punta Cana) or
• making a more generic search query (e.g. Myrina Souvlaki or Fiji Bungallows)

Installation

pip install scrape_reviews

Fetching

from scrape_reviews import scraping_gmaps

scan_a_place = scraping_gmaps.fetch_gmaps_reviews(
    place_wanted= "Nauru",
    category_wanted = "hoteles", 
    # Keyword of category In Spanish
    # It would also work with:
    # "hotels", "hotéis", "ξενοδοχεία", "hôtels" 
    #  
    position_of_db_to_append_results = "my_reviews.db" 
    # This will create a database in the current working directory
    )

make_a_search_query = scraping_gmaps.fetch_gmaps_reviews(

    place_wanted = "Myrina Souvlaki", # Acts as a search term,
    
    category_wanted = "doesnt_matter_since_we_want_to_make_a_search_query",
    
    position_of_db_to_append_results = "my_reviews.db" 
    # Append the same database as before
    )

Exploring What's Fetched

A dictionary containing two dictionaries:

scan_a_place.keys()

# Returns:
# dict_keys(['entity_info', 'reviews'])

Zooming into entity_info:

scan_a_place['entity_info']

# A dict containing dictionaries with lists as values:
# {
# {'Name': 
# ['Menen Hotel', 'Ewa Lodge'], 
# 'Number_of_Ratings': ['(217)', '(97)'], 
# 'fetched_at': ['2022-09-09_16.52.54', '2022-09-09_16.52.54'], 
# 'Overall_Rating': ['3,8', '4,3'], 
# 'fetched_looking_at_place': ['Nauru', 'Nauru']
# }

Then, zooming into reviews:

scan_a_place['reviews'].keys()

#dict_keys(
# ['data-review-id', 'date_created', 'reviewer',
#  'regards_entity', 'reviewers_no_of_reviews',
#  'fetched_at', 'text_of_review', 'rating']
# )


# Let's explore the make_a_search_query as well

for reviews_key, reviews_dimension in make_a_search_query['reviews'].items():
    print("First Review's ", reviews_key, " starts with:", reviews_dimension[0][0][:15])
    print("-"*30) 

# Prints:
# 
# First Review's  date_created  starts with: πριν από έναν μ
# ------------------------------
# First Review's  fetched_at  starts with: 2022-09-09_17.3
# ------------------------------
# First Review's  data-review-id  starts with: ChdDSUhNMG9nS0V
# ------------------------------
# First Review's  reviewer  starts with: George Dristass
# ------------------------------
# First Review's  reviewers_no_of_reviews  starts with: 5 κριτικές
# ------------------------------
# First Review's  text_of_review  starts with: Πολύ ποιοτικός
# ------------------------------
# First Review's  regards_entity  starts with: souvlaki.gr
# ------------------------------
# First Review's  rating  starts with: 5/5
# ------------------------------

Simply Storing Results as Pickles

from scrape_reviews.saving import serialize_results

where_are_my_fetched_reviews_stored =  serialize_results(
    reviews_fetched = make_a_search_query, 
    this_time_I_want_as_place = 'Myrina', #will be used in the stored pickle filename
    this_time_I_want_as_category = 'Souvlaki',
    directory_to_save_at = "C:\\Users\\kvoul" # If we don't specify it, it will store it in our current directory
    )

# Prints:
# saved_at: C:\Users\kvoul\Myrina_Souvlaki_saved_at_2022-09-09_17.46.17_reviews.pkl

# And returns the full path
# So that we can assign it to a variable

print(where_are_my_fetched_reviews_stored)

# Prints: 
# C:\Users\kvoul\Myrina_Souvlaki_saved_at_2022-09-09_17.46.17_reviews.pkl

Re-Store From Pickles

from scrape_reviews.saving import deserialize_results

restored_reviews_from_earlier_run = deserialize_results(where_are_my_fetched_reviews_stored)

print(restored_reviews_from_earlier_run.keys())

# Prints:
# dict_keys(['entity_info', 'reviews'])

Or Read our Database

from scrape_reviews.connect_with_sqlite_db import DbConnector

# Initiate a Connector
my_medium = DbConnector(db_name = "my_reviews.db") # We filled with the above lines

# Read a whole table
my_medium.read_a_table(table_wanted = 'entity_info')

# [
# ('Menen Hotel', '3,8', '(217)', 'Nauru', '2022-09-09_16.52.54'),
# ('Ewa Lodge', '4,3', '(97)', 'Nauru', '2022-09-09_16.52.54'),
# ('Od-N Aiwo Hotel', '3,5', '(75)', 'Nauru', '2022-09-09_16.52.54'), 
# ('souvlaki.gr', '4,3', '(186)', 'Myrina Souvlaki', '2022-09-09_17.08.09'), 
# ('ΤΟ ΛΙΜΆΝΙ (ΨΗΤΟΠΩΛΕΙΟ,Γύρος)', '3,9', '(251)', 'Myrina Souvlaki', '2022-09-09_17.31.52'), 
# ('Το Φαγουριό', '4,3', '(213)', 'Myrina Souvlaki', '2022-09-09_17.31.52')
# ]

my_medium.search_reviews(search_keyword = "from Brisbane")

# Returns:
# [
# ('6 κριτικές', 
# 'Alex Lindeman', 
# '4/5', 
# '(Μετάφραση από το Google) Προσβλέπουμε να ακούσουμε τη ζωντανή ψυχαγωγία του Reef Bar. Μια πράξη της κατηγορίας από το Μπρίσμπεϊν εκτελείται πλέον τακτικά.\n\n(Αρχικό κείμενο)\nLooking forward to hearing the Reef Bar s live entertainment.  A class act from Brisbane now performs regularly.', 
# 'Menen Hotel', 
# 'πριν από 8 χρόνια στο\nGoogle', 
# '2022-09-09_16.52.54', 
# 'ChdDSUhNMG9nS0VJQ0FnSUN3bmVDMHRBRRAB')
# ]


my_medium.perform_a_query(
    query = """
    SELECT reviewer, reviewers_no_of_reviews
    FROM reviews
    ORDER BY reviewer
    LIMIT 3
    """)

# Returns:
# [
# ('A k', '8 κριτικές'), 
# ('AF K', '1 κριτική'), 
# ('AGGELIKI ANASTOPOULOU', 
# 'Τοπικός οδηγός · 22 κριτικές')
# ]

And Cleaning

from scrape_reviews import cleaning_gmaps

info_as_df = cleaning_gmaps.from_info_dict_to_info_df(restored_reviews_from_earlier_run)

print(info_as_df.head(1))

#   Overall_Rating           fetched_at fetched_looking_at_place         Name Number_of_Ratings
# 0            4,3  2022-09-09_17.31.52          Myrina Souvlaki  souvlaki.gr             (186)

# [1 rows x 5 columns]

clean_info_df = cleaning_gmaps.clean_info_df(info_as_df)

print(clean_info_df.dtypes)

# Overall_Rating              float64
# fetched_at                   object
# fetched_looking_at_place     object
# Name                         object
# Number_of_Ratings             int32
# dtype: object

And the same for reviews:

from scrape_reviews import cleaning_gmaps

reviews_as_df = cleaning_gmaps.from_reviews_dict_to_reviews_df(restored_reviews_from_earlier_run)

print(reviews_as_df.head(1))
#          date_created           fetched_at  ... regards_entity rating
# 0  πριν από έναν μήνα  2022-09-09_17.31.52  ...    souvlaki.gr    5/5

# [1 rows x 8 columns]


clean_reviews_df = cleaning_gmaps.clean_reviews_df(reviews_as_df)
print(clean_reviews_df.head(1))

#          date_created           fetched_at  ... Created_Before_measure_unit_of_time Review_Listed_at
# 0  πριν από έναν μήνα  2022-09-09_17.31.52  ...                                μήνα

# [1 rows x 15 columns]

print(clean_reviews_df.dtypes)

# Prints:
# date_created                            object
# fetched_at                              object
# data-review-id                          object
# reviewer                                object
# reviewers_no_of_reviews                 object
# text_of_review                          object
# regards_entity                          object
# rating                                  object
# absolute_rating                        float64
# rating_out_of                          float64
# rating_raw                              object
# relative_rating                        float64
# Created_Before_number                   object
# Created_Before_measure_unit_of_time     object
# Review_Listed_at                        object
# dtype: object