Indeed country support (#38)

README.md

@@ -23,7 +23,11 @@ import pandas as pd
 jobs: pd.DataFrame = scrape_jobs(
     site_name=["indeed", "linkedin", "zip_recruiter"],
     search_term="software engineer",
-    results_wanted=10
+    location="Dallas, TX",
+    results_wanted=10,
+
+    # country: only needed for indeed
+    country='USA'
 )
 
 if jobs.empty:
@@ -65,8 +69,10 @@ Optional
 ├── is_remote (bool)
 ├── results_wanted (int): number of job results to retrieve for each site specified in 'site_type'
 ├── easy_apply (bool): filters for jobs on LinkedIn that have the 'Easy Apply' option
+├── country (enum): uses the corresponding subdomain on Indeed (e.g. Canada on Indeed is ca.indeed.com
 ```
 
+
 ### JobPost Schema
 ```plaintext
 JobPost
@@ -81,12 +87,91 @@ JobPost
 ├── job_type (enum)
 ├── compensation (object)
 │   ├── interval (CompensationInterval): yearly, monthly, weekly, daily, hourly
-│   ├── min_amount (float)
-│   ├── max_amount (float)
+│   ├── min_amount (int)
+│   ├── max_amount (int)
 │   └── currency (str)
 └── date_posted (datetime)
 ```
 
+## Supported Countries for Job Searching
+
+
+### **LinkedIn**
+
+LinkedIn searches globally. Use the `location` parameter
+
+### **ZipRecruiter**
+
+ZipRecruiter searches for jobs in US/Canada. Use the `location` parameter
+
+
+### **Indeed**
+For Indeed, you `location` along with `country` param
+
+You can specify the following countries when searching on Indeed (use the exact name): 
+
+- Argentina
+- Australia
+- Austria
+- Bahrain
+- Belgium
+- Brazil
+- Canada
+- Chile
+- China
+- Colombia
+- Costa Rica
+- Czech Republic
+- Denmark
+- Ecuador
+- Egypt
+- Finland
+- France
+- Germany
+- Greece
+- Hong Kong
+- Hungary
+- India
+- Indonesia
+- Ireland
+- Israel
+- Italy
+- Japan
+- Kuwait
+- Luxembourg
+- Malaysia
+- Mexico
+- Morocco
+- Netherlands
+- New Zealand
+- Nigeria
+- Norway
+- Oman
+- Pakistan
+- Panama
+- Peru
+- Philippines
+- Poland
+- Portugal
+- Qatar
+- Romania
+- Saudi Arabia
+- Singapore
+- South Africa
+- South Korea
+- Spain
+- Sweden
+- Switzerland
+- Taiwan
+- Thailand
+- Turkey
+- Ukraine
+- United Arab Emirates
+- UK
+- USA
+- Uruguay
+- Venezuela
+- Vietnam
 
 ## Frequently Asked Questions
 

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "python-jobspy"
-version = "1.0.3"
+version = "1.1.0"
 description = "Job scraper for LinkedIn, Indeed & ZipRecruiter"
 authors = ["Zachary Hampton <[email protected]>", "Cullen Watson <[email protected]>"]
 readme = "README.md"

src/jobspy/__init__.py

@@ -1,15 +1,11 @@
 import pandas as pd
 from typing import List, Tuple
 
-from .jobs import JobType
+from .jobs import JobType, Location
 from .scrapers.indeed import IndeedScraper
 from .scrapers.ziprecruiter import ZipRecruiterScraper
 from .scrapers.linkedin import LinkedInScraper
-from .scrapers import (
-    ScraperInput,
-    Site,
-    JobResponse,
-)
+from .scrapers import ScraperInput, Site, JobResponse, Country
 
 
 SCRAPER_MAPPING = {
@@ -32,6 +28,7 @@ def scrape_jobs(
     job_type: JobType = None,
     easy_apply: bool = False,  # linkedin
     results_wanted: int = 15,
+    country: str = "usa",
 ) -> pd.DataFrame:
     """
     Asynchronously scrapes job data from multiple job sites.
@@ -41,9 +38,12 @@ def scrape_jobs(
     if type(site_name) == str:
         site_name = _map_str_to_site(site_name)
 
+    country_enum = Country.from_string(country)
+
     site_type = [site_name] if type(site_name) == Site else site_name
     scraper_input = ScraperInput(
         site_type=site_type,
+        country=country_enum,
         search_term=search_term,
         location=location,
         distance=distance,
@@ -71,22 +71,15 @@ def scrape_site(site: Site) -> Tuple[str, JobResponse]:
         for job in job_response.jobs:
             data = job.dict()
             data["site"] = site
-
-            # Formatting JobType
-            data["job_type"] = data["job_type"].value if data["job_type"] else None
-
-            # Formatting Location
-            location_obj = data.get("location")
-            if location_obj and isinstance(location_obj, dict):
-                data["city"] = location_obj.get("city", "")
-                data["state"] = location_obj.get("state", "")
-                data["country"] = location_obj.get("country", "USA")
+            data["company"] = data["company_name"]
+            if data["job_type"]:
+                # Take the first value from the job type tuple
+                data["job_type"] = data["job_type"].value[0]
             else:
-                data["city"] = None
-                data["state"] = None
-                data["country"] = None
+                data["job_type"] = None
+
+            data["location"] = Location(**data["location"]).display_location()
 
-            # Formatting Compensation
             compensation_obj = data.get("compensation")
             if compensation_obj and isinstance(compensation_obj, dict):
                 data["interval"] = (
@@ -111,13 +104,13 @@ def scrape_site(site: Site) -> Tuple[str, JobResponse]:
         desired_order = [
             "site",
             "title",
-            "company_name",
-            "city",
-            "state",
+            "company",
+            "location",
             "job_type",
             "interval",
             "min_amount",
             "max_amount",
+            "currency",
             "job_url",
             "description",
         ]

src/jobspy/jobs/__init__.py

@@ -6,24 +6,160 @@
 
 
 class JobType(Enum):
-    FULL_TIME = "fulltime"
-    PART_TIME = "parttime"
-    CONTRACT = "contract"
-    TEMPORARY = "temporary"
-    INTERNSHIP = "internship"
-
-    PER_DIEM = "perdiem"
-    NIGHTS = "nights"
-    OTHER = "other"
-    SUMMER = "summer"
-    VOLUNTEER = "volunteer"
+    FULL_TIME = (
+        "fulltime",
+        "períodointegral",
+        "estágio/trainee",
+        "cunormăîntreagă",
+        "tiempocompleto",
+        "vollzeit",
+        "voltijds",
+        "tempointegral",
+        "全职",
+        "plnýúvazek",
+        "fuldtid",
+        "دوامكامل",
+        "kokopäivätyö",
+        "tempsplein",
+        "vollzeit",
+        "πλήρηςαπασχόληση",
+        "teljesmunkaidő",
+        "tempopieno",
+        "tempsplein",
+        "heltid",
+        "jornadacompleta",
+        "pełnyetat",
+        "정규직",
+        "100%",
+        "全職",
+        "งานประจำ",
+        "tamzamanlı",
+        "повназайнятість",
+        "toànthờigian",
+    )
+    PART_TIME = ("parttime", "teilzeit")
+    CONTRACT = ("contract", "contractor")
+    TEMPORARY = ("temporary",)
+    INTERNSHIP = ("internship", "prácticas", "ojt(onthejobtraining)", "praktikum")
+
+    PER_DIEM = ("perdiem",)
+    NIGHTS = ("nights",)
+    OTHER = ("other",)
+    SUMMER = ("summer",)
+    VOLUNTEER = ("volunteer",)
+
+
+class Country(Enum):
+    ARGENTINA = ("argentina", "ar")
+    AUSTRALIA = ("australia", "au")
+    AUSTRIA = ("austria", "at")
+    BAHRAIN = ("bahrain", "bh")
+    BELGIUM = ("belgium", "be")
+    BRAZIL = ("brazil", "br")
+    CANADA = ("canada", "ca")
+    CHILE = ("chile", "cl")
+    CHINA = ("china", "cn")
+    COLOMBIA = ("colombia", "co")
+    COSTARICA = ("costa rica", "cr")
+    CZECHREPUBLIC = ("czech republic", "cz")
+    DENMARK = ("denmark", "dk")
+    ECUADOR = ("ecuador", "ec")
+    EGYPT = ("egypt", "eg")
+    FINLAND = ("finland", "fi")
+    FRANCE = ("france", "fr")
+    GERMANY = ("germany", "de")
+    GREECE = ("greece", "gr")
+    HONGKONG = ("hong kong", "hk")
+    HUNGARY = ("hungary", "hu")
+    INDIA = ("india", "in")
+    INDONESIA = ("indonesia", "id")
+    IRELAND = ("ireland", "ie")
+    ISRAEL = ("israel", "il")
+    ITALY = ("italy", "it")
+    JAPAN = ("japan", "jp")
+    KUWAIT = ("kuwait", "kw")
+    LUXEMBOURG = ("luxembourg", "lu")
+    MALAYSIA = ("malaysia", "malaysia")
+    MEXICO = ("mexico", "mx")
+    MOROCCO = ("morocco", "ma")
+    NETHERLANDS = ("netherlands", "nl")
+    NEWZEALAND = ("new zealand", "nz")
+    NIGERIA = ("nigeria", "ng")
+    NORWAY = ("norway", "no")
+    OMAN = ("oman", "om")
+    PAKISTAN = ("pakistan", "pk")
+    PANAMA = ("panama", "pa")
+    PERU = ("peru", "pe")
+    PHILIPPINES = ("philippines", "ph")
+    POLAND = ("poland", "pl")
+    PORTUGAL = ("portugal", "pt")
+    QATAR = ("qatar", "qa")
+    ROMANIA = ("romania", "ro")
+    SAUDIARABIA = ("saudi arabia", "sa")
+    SINGAPORE = ("singapore", "sg")
+    SOUTHAFRICA = ("south africa", "za")
+    SOUTHKOREA = ("south korea", "kr")
+    SPAIN = ("spain", "es")
+    SWEDEN = ("sweden", "se")
+    SWITZERLAND = ("switzerland", "ch")
+    TAIWAN = ("taiwan", "tw")
+    THAILAND = ("thailand", "th")
+    TURKEY = ("turkey", "tr")
+    UKRAINE = ("ukraine", "ua")
+    UNITEDARABEMIRATES = ("united arab emirates", "ae")
+    UK = ("uk", "uk")
+    USA = ("usa", "www")
+    URUGUAY = ("uruguay", "uy")
+    VENEZUELA = ("venezuela", "ve")
+    VIETNAM = ("vietnam", "vn")
+
+    # internal for ziprecruiter
+    US_CANADA = ("usa/ca", "www")
+
+    # internal for linkeind
+    WORLDWIDE = ("worldwide", "www")
+
+    def __new__(cls, country, domain):
+        obj = object.__new__(cls)
+        obj._value_ = country
+        obj.domain = domain
+        return obj
+
+    @property
+    def domain_value(self):
+        return self.domain
+
+    @classmethod
+    def from_string(cls, country_str: str):
+        """Convert a string to the corresponding Country enum."""
+        country_str = country_str.strip().lower()
+        for country in cls:
+            if country.value == country_str:
+                return country
+        valid_countries = [country.value for country in cls]
+        raise ValueError(
+            f"Invalid country string: '{country_str}'. Valid countries (only include this param for Indeed) are: {', '.join(valid_countries)}"
+        )
 
 
 class Location(BaseModel):
-    country: str = "USA"
-    city: str = None
+    country: Country = None
+    city: Optional[str] = None
     state: Optional[str] = None
 
+    def display_location(self) -> str:
+        location_parts = []
+        if self.city:
+            location_parts.append(self.city)
+        if self.state:
+            location_parts.append(self.state)
+        if self.country and self.country not in (Country.US_CANADA, Country.WORLDWIDE):
+            if self.country.value in ("usa", "uk"):
+                location_parts.append(self.country.value.upper())
+            else:
+                location_parts.append(self.country.value.title())
+        return ", ".join(location_parts)
+
 
 class CompensationInterval(Enum):
     YEARLY = "yearly"
@@ -37,7 +173,7 @@ class Compensation(BaseModel):
     interval: CompensationInterval
     min_amount: int = None
     max_amount: int = None
-    currency: str = "USD"
+    currency: Optional[str] = "USD"
 
 
 class JobPost(BaseModel):