Initial release v1.0.0

.gitignore

@@ -127,3 +127,5 @@ dmypy.json
 
 # Pyre type checker
 .pyre/
+
+.csv

README.md

@@ -2,19 +2,41 @@
 
 [![Python 3.8](https://img.shields.io/badge/python-3.8-blue.svg)](https://www.python.org/downloads/release/python-380/)
 
-requirements:
-* requests
-
-To run script first you need to set the API_KEY environment variable, like this:
-
+# Installation
+## Using Pip
+```bash
+  $ pip install geocli
+```
+## Manual
+```bash
+  $ git clone https://github.com/expandaventures/geocoding-cli
+  $ cd geocoding-cli
+  $ python setup.py install
+```
+# Usage
+```bash
+$ geocli
 ```
-export API_KEY='YOUR_API_KEY'
+## Geocode a single address
+`Usage: geocli geocode [OPTIONS] ADDRESS [STATE] [CITY]`
+```bash
+$ geocli geocode --dry_run 'Av. Reforma 222' cdmx mexico
 ```
+## Batch: geocode a addresses in a file
 
-and next use
+`Usage: geocli batch [OPTIONS] INPUT_FILE OUTPUT_PATH`
+```bash
+$ geocli batch --dry_run sample.csv out.csv
+```
+
+The expected schema in the file is:
 
 ```
-python3 geocoding_imss.py file.csv output_name
+id, state, unused, city, ... , address
+1, CDMX,,Mexico,,Av. Reforma 222
+
 ```
 
-where ```file.csv``` is the file that you want to read and parse address to coordinates and ```output_name``` is the file name that will saved after the script runs
+The **ONLY** columns used are 2nd, 4th and last. 
+(e.g. row[1], row[3], row[-1])
+

geocli/__main__.py

@@ -0,0 +1,45 @@
+import sys
+import click
+from geocli_driver import process_file, get_location_by_address
+
+
+@click.group()
+@click.version_option("1.0.0")
+def main():
+    """A Geocoding CLI"""
+    pass
+
+
+@main.command()
+@click.argument('input_file', required=True, type=click.Path(exists=True))
+@click.argument('output_path', required=True)
+@click.option('--dry_run', is_flag=True)
+def batch(**kwargs):
+    """Process a file and output the results to a file"""
+    results = process_file(kwargs.get("input_file"),
+                           kwargs.get("output_path"),
+                           kwargs.get("dry_run", False),
+                           )
+    click.echo(f'Processed {results} results')
+
+
+@main.command()
+@click.argument('address', required=True)
+@click.argument('state', required=False, default='')
+@click.argument('city', required=False, default='')
+@click.option('--dry_run', is_flag=True)
+def geocode(**kwargs):
+    """process a single address"""
+    lat, lon = get_location_by_address(kwargs.get("state", ""),
+                                       kwargs.get("city", ""),
+                                       address_string=kwargs.get("address"),
+                                       dry_run=kwargs.get("dry_run", False),
+                                       )
+    click.echo(f'Lat, Lon: {lat}, {lon}')
+
+
+if __name__ == '__main__':
+    args = sys.argv
+    if "--help" in args or len(args) == 1:
+        print("GEOCLI")
+    main()

geocli/main.py → geocli/geocli_driver.py

@@ -1,14 +1,12 @@
+import csv
 import os
 import re
-import sys
-import csv
-import time
-
 import requests
+import time
 
-start = time.time()
-
-API_KEY = os.getenv("API_KEY", "")
+API_KEY = os.getenv("GOOGLE_API_KEY", False)
+LIMIT_LON = tuple(map(float, os.getenv("LIMIT_LON", "-117.12776,14.5388286402").split(',')))
+LIMIT_LAT = tuple(map(float, os.getenv("LIMIT_LAT", "-86.811982388,32.72083").split(',')))
 
 
 def clean_and_format(string):
@@ -24,11 +22,7 @@ def extract_cp(string):
     return cp_str
 
 
-LIMIT_LON = -117.12776, 14.5388286402
-LIMIT_LAT = -86.811982388, 32.72083
-
-
-def get_location_by_address(state, city, address_string=None):
+def get_location_by_address(state, city, address_string=None, dry_run=False):
     if address_string:
         address_string = re.sub(r'ENTRE\s(.*)\sY\s(.*)\sCOLONIA:', 'COLONIA:', address_string)
         address_string = address_string.replace("C.P.", "postal_code=")
@@ -40,6 +34,9 @@ def get_location_by_address(state, city, address_string=None):
 
     url = "https://maps.googleapis.com/maps/api/geocode/json?address={}&region=mx&key={}".format(address_string,
                                                                                                  API_KEY)
+    if dry_run:
+        print(url)
+        return None, None
     response = requests.get(url)
     try:
         if response.status_code == 200:
@@ -55,33 +52,32 @@ def get_location_by_address(state, city, address_string=None):
     return None, None
 
 
-args = sys.argv
-if len(args) != 3:
-    raise Exception("Scripts takes exactly two argument ({} given) [file_path] and [output_name]".format(len(args) - 1))
-file_path = args[1]
-output_name = args[2]
-
-with open(file_path) as f:
-    csv_data = csv.reader(f, delimiter=',', quotechar='"')
-    rows = list(csv_data)
-    rows[0].append("Latitud")
-    rows[0].append("Longitud")
-    for row in rows[1:]:
-        address = row[-1]
-        cp = extract_cp(address)
-        lat, lng = get_location_by_address(row[1], row[3], address_string=address)
-        if (not lat and not lng) or (
-                (LIMIT_LAT[0] > lat > LIMIT_LAT[1]) and (LIMIT_LON[0] > lng > LIMIT_LON[1])):
-            lat, lng = get_location_by_address(row[1], row[3], address_string=cp)
-        row.append(lat)
-        row.append(lng)
-        print(row[0], lat, lng)
-
-
-with open(output_name, 'w', newline='') as csvfile:
-    csv_file = csv.writer(csvfile, delimiter=',',
-                          quotechar='"', quoting=csv.QUOTE_MINIMAL)
-    csv_file.writerows(rows)
-
-end = time.time()
-print(end - start)
+def process_file(file_path, output_name, dry_run=False):
+    if not API_KEY and not dry_run:
+        raise ValueError("Environment Variable GOOGLE_API_KEY must be set")
+    start = time.time()
+    with open(file_path) as f:
+        csv_data = csv.reader(f, delimiter=',', quotechar='"')
+        rows = list(csv_data)
+        rows[0].append("Latitud")
+        rows[0].append("Longitud")
+        for row in rows[1:]:
+            address = row[-1]
+            cp = extract_cp(address)
+            lat, lng = get_location_by_address(row[1], row[3], address_string=address, dry_run=dry_run)
+            if (not lat and not lng) or (
+                    (LIMIT_LAT[0] > lat > LIMIT_LAT[1]) and (LIMIT_LON[0] > lng > LIMIT_LON[1])):
+                lat, lng = get_location_by_address(row[1], row[3], address_string=cp, dry_run=dry_run)
+            row.append(lat)
+            row.append(lng)
+            print(row[0], lat, lng)
+            if dry_run:
+                return 0
+
+    with open(output_name, 'w', newline='') as csvfile:
+        csv_file = csv.writer(csvfile, delimiter=',',
+                              quotechar='"', quoting=csv.QUOTE_MINIMAL)
+        csv_file.writerows(rows)
+    end = time.time()
+    print(end - start)
+    return len(rows)

requirements.txt

@@ -1 +1,6 @@
+certifi==2020.12.5
+chardet==4.0.0
+click==7.1.2
+idna==2.10
 requests==2.25.1
+urllib3==1.26.3

sample.csv

@@ -0,0 +1,2 @@
+id, state, unused, city, ... , address
+1, CDMX,,Mexico,,Reforma 222

setup.py

@@ -0,0 +1,47 @@
+from setuptools import setup, find_packages
+from io import open
+from os import path
+
+import pathlib
+
+# The directory containing this file
+HERE = pathlib.Path(__file__).parent
+
+# The text of the README file
+README = (HERE / "README.md").read_text()
+
+# automatically captured required modules for install_requires in requirements.txt
+with open(path.join(HERE, 'requirements.txt'), encoding='utf-8') as f:
+    all_reqs = f.read().split('\n')
+
+install_requires = [x.strip() for x in all_reqs if ('git+' not in x) and (
+    not x.startswith('#')) and (not x.startswith('-'))]
+dependency_links = [x.strip().replace('git+', '') for x in all_reqs \
+                    if 'git+' not in x]
+setup(
+    name='geocli',
+    description='A simple commandline app for geocoding',
+    version='1.0.0',
+    packages=find_packages(),  # list of all packages
+    install_requires=install_requires,
+    python_requires='>=2.7',  # any python greater than 2.7
+    entry_points='''
+        [console_scripts]
+        geocli=geocli.__main__:main
+    ''',
+    author="Paco Valdez",
+    keyword="google maps, georeference, address, geocoding",
+    long_description=README,
+    long_description_content_type="text/markdown",
+    license='GPLv3',
+    url='https://github.com/expandaventures/geocoding-cli',
+    download_url='https://github.com/expandaventures/geocoding-cli/archive/1.0.0.tar.gz',
+    dependency_links=dependency_links,
+    author_email='[email protected]',
+    classifiers=[
+        "License :: OSI Approved :: GNU General Public License v3 or later (GPLv3+)",
+        "Programming Language :: Python :: 2.7",
+        "Programming Language :: Python :: 3",
+        "Programming Language :: Python :: 3.7",
+    ]
+)