Adding Geo Location and improve internet responses

docs/devel/askai-questions.txt

@@ -1,6 +1,14 @@
 Questions:
 
+en_US
+
 1. Summarize my markdown files at my HomeSetup docs folder.
 2. What are the current weather conditions in San Francisco, U.S today?
 3. When is the upcoming Los Angeles Lakers match?
-4. Who currently holds the office of President of the United States?
+4. Who currently holds the office of President of the United States?
+
+pt_BR
+
+1. Qual a previsao do tempo hoje para Belo Horizonte?
+2. Quem e o atual presidente do Brasil?
+3. Quem e Hugo Saporetti junior?

src/demo/components/internet-demo.py

@@ -12,25 +12,33 @@
 
    Copyright·(c)·2024,·HSPyLib
 """
-from askai.core.component.cache_service import cache
-from askai.core.component.internet_service import internet
-from askai.core.support.shared_instances import shared
+import logging as log
+import re
+
 from clitt.core.tui.line_input.line_input import line_input
 from hspylib.core.tools.commons import log_init, sysout
 
-import logging as log
+from askai.core.component.cache_service import cache
+from askai.core.component.geo_location import geo_location
+from askai.core.component.internet_service import internet
+from askai.core.model.search_result import SearchResult
+from askai.core.support.shared_instances import shared
 
 if __name__ == "__main__":
+
     log_init("internet-demo.log", level=log.INFO)
     cache.read_query_history()
     sysout("-=" * 40)
     sysout("AskAI Internet Demo")
     sysout("-=" * 40)
     shared.create_engine(engine_name="openai", model_name="gpt-3.5-turbo")
-    urls = ["https://www.accuweather.com"]
     sysout(f"READY to search")
     sysout("--" * 40)
+
     while (query := line_input("You: ")) not in ["exit", "q", "quit"]:
-        answer = internet.search_google(query, *urls)
+        kw: list[str] = re.split('[ ,;]', query)
+        sites: list[str] = ['accuweather.com', 'weather.com']
+        q = SearchResult(query, geo_location.now, kw, sites)
+        answer = internet.search_google(q)
         sysout(f"%GREEN%AI: {answer}")
     cache.save_query_history()

src/main/askai/core/component/geo_location.py

@@ -0,0 +1,97 @@
+import json
+import logging as log
+from datetime import datetime
+from json import JSONDecodeError
+from textwrap import dedent
+
+import pytz
+from hspylib.core.metaclass.singleton import Singleton
+from hspylib.core.namespace import Namespace
+from hspylib.modules.fetch import fetch
+from requests.exceptions import ConnectionError
+
+
+class GeoLocation(metaclass=Singleton):
+    """TODO"""
+
+    INSTANCE: 'GeoLocation' = None
+
+    GEO_LOC_URL: str = "http://ip-api.com/json"
+
+    EMPTY_JSON_RESP: str = dedent('''
+    {
+        "status": "failure", "country": "", "countryCode": "", "region": "", "regionName": "",
+        "city": "", "zip": "", "lat": 0.0, "lon": 0.0, "timezone": "",
+        "isp": "", "org": "", "as": "", "query": ""
+    }
+    ''')
+
+    # Date format used in prompts, e.g: Fri 22 Mar 19:47 2024.
+    DATE_FMT: str = "%a %d %b %-H:%M %Y"
+
+    @classmethod
+    def get_location(cls, ip: str = None) -> Namespace:
+        """TODO"""
+        try:
+            url = f"{cls.GEO_LOC_URL}{'/' + ip if ip else ''}"
+            log.debug('Fetching the Geo Position from: %s', url)
+            geo_req = fetch.get(url)
+        except (JSONDecodeError, ConnectionError) as err:
+            log.error('Failed to retrieve geo location => %s', str(err))
+            geo_req = Namespace(body=cls.EMPTY_JSON_RESP)
+        geo_json = json.loads(geo_req.body)
+        geo_location: Namespace = Namespace(**geo_json)
+        return geo_location
+
+    def __init__(self, ip: str = None):
+        self._geo_location = self.get_location(ip)
+
+    def __str__(self):
+        geo_loc = self._geo_location
+        geo_loc.setattr('zoned_datetime', self.now)
+        return str(self._geo_location)
+
+    @property
+    def latitude(self) -> float:
+        return self._geo_location.lat
+
+    @property
+    def longitude(self) -> float:
+        return self._geo_location.lon
+
+    @property
+    def country(self) -> str:
+        return self._geo_location.country
+
+    @property
+    def country_code(self) -> str:
+        return self._geo_location.country_code
+
+    @property
+    def region(self) -> str:
+        return self._geo_location.region
+
+    @property
+    def region_name(self) -> str:
+        return self._geo_location.region_name
+
+    @property
+    def city(self) -> float:
+        return self._geo_location.city
+
+    @property
+    def zip(self) -> str:
+        return self._geo_location.zip
+
+    @property
+    def timezone(self) -> str:
+        return self._geo_location.timezone
+
+    @property
+    def now(self) -> str:
+        utc_datetime = datetime.utcnow().replace(tzinfo=pytz.utc)
+        zoned_datetime = utc_datetime.astimezone(pytz.timezone(self.timezone))
+        return zoned_datetime.strftime(GeoLocation.DATE_FMT)
+
+
+assert (geo_location := GeoLocation().INSTANCE) is not None

src/main/askai/core/component/internet_service.py

@@ -70,21 +70,35 @@ def __init__(self):
 
     def search_google(self, search: SearchResult) -> Optional[str]:
         """Search the web using google search API.
+        Google search operators: https://ahrefs.com/blog/google-advanced-search-operators/
         :param search: The AI search parameters.
         """
-        query = '+'.join(search.keywords)
         if len(search.sites) > 0:
             search_results: List[Document] = []
-            websites: str = ' OR '.join(['site: ' + url for url in search.sites])
+            query = self._build_query(search.keywords, search.filters, search.sites)
             AskAiEvents.ASKAI_BUS.events.reply.emit(message=msg.searching())
             log.info("Searching GOOGLE for '%s'  url: '%s'", query, str(', '.join(search.sites)))
-            content = str(self._tool.run(f"{query} {websites}"))
+            content = str(self._tool.run(query))
             search_results.append(Document(content))
             prompt = ChatPromptTemplate.from_messages([("system", "{query}\n\n{context}")])
             chain = create_stuff_documents_chain(lc_llm.create_chat_model(), prompt)
             return chain.invoke({"query": query, "context": search_results})
 
         return None
 
+    def _build_query(self, keywords: List[str], filters: List[str], sites: List[str]) -> str:
+        """TODO"""
+        query = ''
+        # Weather is a filter that does not require any other search parameter.
+        if filters and any(f.find("weather:") >= 0 for f in filters):
+            return ' AND '.join(filters)
+        if sites:
+            query += ' OR '.join(['site:' + url for url in sites])
+        if filters and any(f.find("people:") >= 0 for f in filters):
+            query += f" intext:\"{' + '.join([f.split(':')[1] for f in filters])}\" "
+        if keywords:
+            query += ' + '.join(keywords)
+        return query
+
 
 assert (internet := InternetService().INSTANCE) is not None

src/main/askai/core/model/search_result.py

@@ -25,6 +25,7 @@ class SearchResult:
     datetime: str = None
     keywords: List[str] = None
     sites: str | List[str] = None
+    filters: List[str] = None
     response: str = None
 
     def __str__(self):

src/main/askai/core/processor/instances/generic_processor.py

@@ -21,6 +21,7 @@
 from askai.core.askai_messages import msg
 from askai.core.askai_prompt import prompt
 from askai.core.component.cache_service import cache
+from askai.core.component.geo_location import geo_location
 from askai.core.engine.openai.temperatures import Temperatures
 from askai.core.model.chat_context import ContextRaw
 from askai.core.model.processor_response import ProcessorResponse
@@ -69,10 +70,10 @@ def template(self) -> str:
     def process(self, query_response: ProcessorResponse) -> Tuple[bool, Optional[str]]:
         status = False
         template = PromptTemplate(input_variables=['user', 'datetime', 'idiom'], template=self.template())
-        final_prompt: str = template.format(user=prompt.user, datetime=shared.now, idiom=shared.idiom)
+        final_prompt: str = template.format(user=prompt.user, datetime=geo_location.now, idiom=shared.idiom)
         shared.context.set("SETUP", final_prompt, "system")
         shared.context.set("QUESTION", f"\n\nQuestion: {query_response.question}\n\nHelpful Answer:")
-        context: ContextRaw = shared.context.join("GENERAL", "SETUP", "QUESTION")
+        context: ContextRaw = shared.context.join("SETUP", "GENERAL", "QUESTION")
         log.info("Setup::[GENERIC] '%s'  context=%s", query_response.question, context)
 
         if (response := shared.engine.ask(context, *Temperatures.CREATIVE_WRITING.value)) and response.is_success:

src/main/askai/core/processor/instances/internet_processor.py

@@ -21,6 +21,7 @@
 from askai.core.askai_messages import msg
 from askai.core.askai_prompt import prompt
 from askai.core.component.cache_service import cache
+from askai.core.component.geo_location import geo_location
 from askai.core.component.internet_service import internet
 from askai.core.engine.openai.temperatures import Temperatures
 from askai.core.model.chat_context import ContextRaw
@@ -60,7 +61,7 @@ def template(self) -> str:
     def process(self, query_response: ProcessorResponse) -> Tuple[bool, Optional[str]]:
         status = False
         template = PromptTemplate(input_variables=['idiom', 'datetime'], template=self.template())
-        final_prompt: str = template.format(idiom=shared.idiom, datetime=shared.now)
+        final_prompt: str = template.format(idiom=shared.idiom, datetime=geo_location.now)
         shared.context.set("SETUP", final_prompt, "system")
         shared.context.set("QUESTION", f"\n\nQuestion: {query_response.question}\n\nHelpful Answer:")
         context: ContextRaw = shared.context.join("SETUP", "QUESTION")
@@ -74,6 +75,7 @@ def process(self, query_response: ProcessorResponse) -> Tuple[bool, Optional[str
                     output = response.message.strip()
                 else:
                     if output := internet.search_google(search):
+                        output = msg.translate(output)
                         shared.context.push("GENERAL", query_response.question)
                         shared.context.push("GENERAL", output, "assistant")
                         cache.save_reply(query_response.question, output)

src/main/askai/core/support/shared_instances.py

@@ -4,7 +4,6 @@
 from clitt.core.tui.line_input.line_input import line_input
 from hspylib.core.metaclass.singleton import Singleton
 from hspylib.core.preconditions import check_state
-from hspylib.core.zoned_datetime import now
 from hspylib.modules.cli.keyboard import Keyboard
 
 from askai.core.askai_configs import configs
@@ -26,9 +25,6 @@ class SharedInstances(metaclass=Singleton):
     # This is the uuid used in prompts that require internet.
     INTERNET_ID: str = 'e35057db-f690-4299-ad4d-147d6124184c'
 
-    # Date format used in prompts, e.g: Fri 22 Mar 19:47 2024.
-    DATE_FMT: str = "%a %d %b %-H:%M %Y"
-
     def __init__(self) -> None:
         self._engine: Optional[AIEngine] = None
         self._context: Optional[ChatContext] = None
@@ -64,10 +60,6 @@ def nickname(self) -> str:
     def username(self) -> str:
         return f"%WHITE%  {prompt.user.title()}%NC%"
 
-    @property
-    def now(self) -> str:
-        return now(self.DATE_FMT)
-
     def create_engine(self, engine_name: str, model_name: str) -> AIEngine:
         """TODO"""
         if self._engine is None:

src/main/askai/resources/assets/prompts/generic-prompt.txt

@@ -13,15 +13,11 @@ Before responding to the user, it is imperative that you follow the step-by-step
 
 1. Review the entire conversation history thoroughly, but refrain from generating a response at this time.
 
-2. Determine if you can craft your response using solely the information available in the conversation history.
+2. Determine if you can craft your response using the conversation history or using the information available in your database. Remember that your training data includes information up to January 2022, and we are currently in: '{datetime}'.
 
-3. Determine if you can craft your response from your database. Remember that your training data includes information up to January 2022, and we are currently in: '{datetime}'.
+3. If you possess pertinent information rather than responses like 'I don't have details...' or 'I lack information about...', please proceed with completing all tasks delineated in 'WORKFLOW 1'.
 
-4. If you're certain of the accuracy of your response, proceed with all tasks outlined in "WORKFLOW 1".
-
-5. If you're unsure or lack an answer, please refrain from guessing. Instead, proceed with all tasks outlined in "WORKFLOW 2".
-
-6. Ensure completion of all tasks assigned to you in the previous steps, following the designated workflow.
+4. If you don't know the answer, or if you don't have a useful answer, please refrain from guessing. Instead, proceed with all tasks outlined in "WORKFLOW 2".
 
 
 "WORKFLOW 1":

src/main/askai/resources/assets/prompts/internet-prompt.txt

@@ -1,6 +1,6 @@
-Act as an internet facilitator by performing tasks such as web crawling, page scraping, and conducting google searches. Respond consistently in the following idiom: '{idiom}'.
+Act as an internet facilitator by performing tasks such as web crawling, page scraping, and conducting Google searches. Respond consistently in the following idiom: '{idiom}'.
 
-Before responding to the user, it is imperative that you follow the step-by-step instructions provided below in sequential order:
+Before responding to the user, you must follow the step-by-step instructions provided below in sequential order:
 
 1. Interpret the question thoroughly to discern the user's underlying intent.
 
@@ -10,16 +10,16 @@ Before responding to the user, it is imperative that you follow the step-by-step
 
 4. Locate credible sources relevant to '{idiom}' to collect essential information for creating a thoroughly researched response. Offer a curated list comprising a minimum of three and a maximum of five website URLs tailored to the user's locale. Please include only the base URL of each website without specifying any particular paths.
 
-5. If the inquiry pertains to programming languages, operating systems, or other IT topics, append 'stackoverflow.com' and 'github.com' to your list.
+5. If the inquiry pertains to programming languages, operating systems, or other IT topics, append 'stackoverflow.com' and 'github.com' to your sites list.
 
-6. If the inquiry pertains to whether forecast or related topics, append 'accuweather.com' to your list.
+6. If the inquiry pertains to general information about individuals who are not notable, extract the name of the person mentioned in the prompt and append this filter: 'people:name' to your list.  Where name represents the extracted person's name. In addition to that, append 'linkedin.com', 'facebook.com', and 'instagram.com' to your sites list.
 
-7. If the inquiry pertains to general information about individuals who are not notable, append 'linkedin.com', 'facebook.com', and 'instagram.com' to your list.
+7. If the inquiry pertains to weather forecasts or related topics, extract the city location mentioned in the prompt and append this filter: 'weather:city' to your list.  Where city represents the extracted city name.
 
 8. The final response is a formatted JSON with no additional description or context.
 
 9. Do not use markdown to format the response message. Use plain JSON.
 
 10. The final response 'JSON' must contain the string fields: 'question', 'datetime'.
 
-11. The final response 'JSON' must contain the array fields, 'keywords' and 'sites'.
+11. The final response 'JSON' must contain the array fields, 'keywords', 'filters', and 'sites'.