webscout 1.2.1__tar.gz → 1.2.3__tar.gz

webscout-1.2.3/DeepWEBS/documents/query_results_extractor.py

@@ -0,0 +1,99 @@
+from bs4 import BeautifulSoup
+from pathlib import Path
+from DeepWEBS.utilsdw.logger import logger
+
+class QueryResultsExtractor:
+    def __init__(self) -> None:
+        self.query_results = []
+        self.related_questions = []
+
+    def load_html(self, html_path):
+        try:
+            with open(html_path, "r", encoding="utf-8") as f:
+                html = f.read()
+            self.soup = BeautifulSoup(html, "html.parser")
+        except FileNotFoundError:
+            logger.error(f"File not found: {html_path}")
+        except Exception as e:
+            logger.error(f"Error loading HTML: {e}")
+
+    def extract_query_results(self):
+        try:
+            self.query = self.soup.find("textarea").text.strip()
+            query_result_elements = self.soup.find_all("div", class_="g")
+            for idx, result in enumerate(query_result_elements):
+                try:
+                    site = result.find("cite").find_previous("span").text.strip()
+                    url = result.find("a")["href"]
+                    title = result.find("h3").text.strip()
+                    abstract_element_conditions = [
+                        {"data-sncf": "1"},
+                        {"class_": "ITZIwc"},
+                    ]
+                    for condition in abstract_element_conditions:
+                        abstract_element = result.find("div", condition)
+                        if abstract_element is not None:
+                            abstract = abstract_element.text.strip()
+                            break
+                    else:
+                        abstract = ""
+                    logger.mesg(
+                        f"{title}\n"
+                        f" - {site}\n"
+                        f" - {url}\n"
+                        f" - {abstract}\n"
+                        f"\n"
+                    )
+                    self.query_results.append(
+                        {
+                            "title": title,
+                            "site": site,
+                            "url": url,
+                            "abstract": abstract,
+                            "index": idx,
+                            "type": "web",
+                        }
+                    )
+                except Exception as e:
+                    logger.error(f"Error extracting query result: {e}")
+            logger.success(f"- {len(query_result_elements)} query results")
+        except Exception as e:
+            logger.error(f"Error extracting query results: {e}")
+
+    def extract_related_questions(self):
+        try:
+            related_question_elements = self.soup.find_all(
+                "div", class_="related-question-pair"
+            )
+            for question_element in related_question_elements:
+                try:
+                    question = question_element.find("span").text.strip()
+                    print(question)
+                    self.related_questions.append(question)
+                except Exception as e:
+                    logger.error(f"Error extracting related question: {e}")
+            logger.success(f"- {len(self.related_questions)} related questions")
+        except Exception as e:
+            logger.error(f"Error extracting related questions: {e}")
+
+    def extract(self, html_path):
+        self.load_html(html_path)
+        self.extract_query_results()
+        self.extract_related_questions()
+        self.search_results = {
+            "query": self.query,
+            "query_results": self.query_results,
+            "related_questions": self.related_questions,
+        }
+        return self.search_results
+
+
+if __name__ == "__main__":
+    html_path_root = Path(__file__).parents[1] / "files"
+    html_filename = "python_tutorials"
+    html_path = html_path_root / f"{html_filename}.html"
+    extractor = QueryResultsExtractor()
+    try:
+        extractor.extract(html_path)
+    except Exception as e:
+        logger.error(f"Error in main function: {e}")

webscout-1.2.3/DeepWEBS/documents/webpage_content_extractor.py

@@ -0,0 +1,145 @@
+import concurrent.futures
+import re
+from pathlib import Path
+from pprint import pprint
+from bs4 import BeautifulSoup
+from tiktoken import get_encoding as tiktoken_get_encoding
+from DeepWEBS.utilsdw.logger import logger
+from markdownify import markdownify
+from DeepWEBS.networks.network_configs import IGNORE_TAGS, IGNORE_CLASSES
+from termcolor import colored
+
+
+class WebpageContentExtractor:
+    def __init__(self):
+        self.tokenizer = tiktoken_get_encoding("cl100k_base")
+
+    def count_tokens(self, text):
+        tokens = self.tokenizer.encode(text)
+        token_count = len(tokens)
+        return token_count
+
+    def html_to_markdown(self, html_str, ignore_links=True):
+        if ignore_links:
+            markdown_str = markdownify(html_str, strip="a")
+        else:
+            markdown_str = markdownify(html_str)
+        markdown_str = re.sub(r"\n{3,}", "\n\n", markdown_str)
+
+        self.markdown_token_count = self.count_tokens(markdown_str)
+        logger.mesg(f'- Tokens: {colored(self.markdown_token_count,"light_green")}')
+
+        self.markdown_str = markdown_str
+
+        return self.markdown_str
+
+    def remove_elements_from_html(self, html_str):
+        soup = BeautifulSoup(html_str, "html.parser")
+        ignore_classes_with_parentheses = [f"({word})" for word in IGNORE_CLASSES]
+        ignore_classes_pattern = f'{"|".join(ignore_classes_with_parentheses)}'
+        removed_element_counts = 0
+        for element in soup.find_all():
+            class_str = ""
+            id_str = ""
+            try:
+                class_attr = element.get("class", [])
+                if class_attr:
+                    class_str = " ".join(list(class_attr))
+                if id_str:
+                    class_str = f"{class_str} {id_str}"
+            except:
+                pass
+
+            try:
+                id_str = element.get("id", "")
+            except:
+                pass
+
+            if (
+                (not element.text.strip())
+                or (element.name in IGNORE_TAGS)
+                or (re.search(ignore_classes_pattern, class_str, flags=re.IGNORECASE))
+                or (re.search(ignore_classes_pattern, id_str, flags=re.IGNORECASE))
+            ):
+                element.decompose()
+                removed_element_counts += 1
+
+        logger.mesg(
+            f"- Elements: "
+            f'{colored(len(soup.find_all()),"light_green")} / {colored(removed_element_counts,"light_red")}'
+        )
+
+        html_str = str(soup)
+        self.html_str = html_str
+
+        return self.html_str
+
+    def extract(self, html_path):
+        logger.note(f"Extracting content from: {html_path}")
+
+        if not Path(html_path).exists():
+            logger.warn(f"File not found: {html_path}")
+            return ""
+
+        encodings = ["utf-8", "latin-1"]
+        for encoding in encodings:
+            try:
+                with open(html_path, "r", encoding=encoding, errors="ignore") as rf:
+                    html_str = rf.read()
+                break
+            except UnicodeDecodeError:
+                pass
+        else:
+            logger.warn(f"No matching encodings: {html_path}")
+            return ""
+
+        html_str = self.remove_elements_from_html(html_str)
+        markdown_str = self.html_to_markdown(html_str)
+        return markdown_str
+
+
+class BatchWebpageContentExtractor:
+    def __init__(self) -> None:
+        self.html_path_and_extracted_content_list = []
+        self.done_count = 0
+
+    def extract_single_html(self, html_path):
+        webpage_content_extractor = WebpageContentExtractor()
+        extracted_content = webpage_content_extractor.extract(html_path)
+        self.html_path_and_extracted_content_list.append(
+            {"html_path": html_path, "extracted_content": extracted_content}
+        )
+        self.done_count += 1
+        logger.success(
+            f"> [{self.done_count}/{self.total_count}] Extracted: {html_path}"
+        )
+
+    def extract(self, html_paths):
+        self.html_path = html_paths
+        self.total_count = len(self.html_path)
+        with concurrent.futures.ThreadPoolExecutor() as executor:
+            futures = [
+                executor.submit(self.extract_single_html, html_path)
+                for html_path in self.html_path
+            ]
+            for idx, future in enumerate(concurrent.futures.as_completed(futures)):
+                result = future.result()
+
+        return self.html_path_and_extracted_content_list
+
+
+if __name__ == "__main__":
+    html_root = Path(__file__).parents[1] / "files" / "urls" / "python tutorials"
+    html_paths = [
+        html_root / html_filename
+        for html_filename in [
+            "docs.python.org_zh-cn_3_tutorial_interpreter.html",
+            "stackoverflow.com_questions_295135_turn-a-string-into-a-valid-filename.html",
+            "www.liaoxuefeng.com_wiki_1016959663602400_1017495723838528.html",
+        ]
+    ]
+    batch_webpage_content_extractor = BatchWebpageContentExtractor()
+    html_path_and_extracted_content_list = batch_webpage_content_extractor.extract(
+        html_paths
+    )
+    # pprint(html_path_and_extracted_content_list)

webscout-1.2.3/DeepWEBS/networks/filepath_converter.py

@@ -0,0 +1,109 @@
+import platform
+import re
+from pathlib import Path
+from urllib.parse import quote, unquote
+
+
+# What characters are forbidden in Windows and Linux directory names?
+#   https://stackoverflow.com/questions/1976007/what-characters-are-forbidden-in-windows-and-linux-directory-names
+
+INVALID_FILE_PATH_CHARS = [
+    "\\",
+    "/",
+    ":",
+    "*",
+    "?",
+    '"',
+    "<",
+    ">",
+    "|",
+    "\n",
+    "\t",
+    "\r",
+    *[chr(i) for i in range(32)],
+]
+
+WINDOWS_INVALID_FILE_PATH_NAMES = [
+    "con",
+    "prn",
+    "aux",
+    "nul",
+    *[f"com{i+1}" for i in range(10)],
+    *[f"lpt{i+1}" for i in range(10)],
+]
+
+
+class FilepathConverter:
+    def __init__(self, parent: str = None):
+        self.output_root = Path(__file__).parents[1] / "files"
+        self.parent = parent
+
+    def preprocess(self, input_string):
+        return input_string
+
+    def validate(self, input_string):
+        if not input_string:
+            return input_string
+        filename = input_string
+        for char in INVALID_FILE_PATH_CHARS:
+            filename = filename.replace(char, "_")
+        if platform.system() == "Windows":
+            filename_base = filename.split(".")[0]
+            if filename_base.lower() in WINDOWS_INVALID_FILE_PATH_NAMES:
+                filename_base = filename_base + "_"
+                filename = ".".join([filename_base, *filename.split(".")[1:]])
+        return filename
+
+    def append_extension(self, filename, accept_exts=[".html", ".htm"], ext=".html"):
+        if ext:
+            filename_ext = "." + filename.split(".")[-1]
+            if filename_ext.lower() not in accept_exts:
+                filename += ext
+        return filename
+
+    def convert(self, input_string, parent=None):
+        filename = self.preprocess(input_string)
+        filename = self.validate(filename)
+        filename = self.append_extension(filename)
+
+        parent = parent or self.parent
+        parent = self.validate(parent)
+        if parent:
+            filepath = self.output_root / parent / filename
+        else:
+            filepath = self.output_root / filename
+
+        self.filename = filename
+        self.filepath = filepath
+
+        return self.filepath
+
+
+class UrlToFilepathConverter(FilepathConverter):
+    def __init__(self, parent: str = None):
+        super().__init__(parent)
+        self.output_root = self.output_root / "urls"
+
+    def preprocess(self, url):
+        filename = unquote(url.split("//")[1])
+        return filename
+
+
+class QueryToFilepathConverter(FilepathConverter):
+    def __init__(self, parent: str = None):
+        super().__init__(parent)
+        self.output_root = self.output_root / "queries"
+
+
+if __name__ == "__main__":
+    query = "python"
+    query_converter = QueryToFilepathConverter()
+    print(query_converter.convert(query))
+
+    # url = "https://trafilatura.readthedocs.io/en/latest/quickstart.html"
+    url = (
+        "https://stackoverflow.com/questions/295135/turn-a-string-into-a-valid-filename"
+    )
+
+    url_converter = UrlToFilepathConverter(parent=query)
+    print(url_converter.convert(url))

webscout-1.2.3/DeepWEBS/networks/google_searcher.py

@@ -0,0 +1,52 @@
+import requests
+from pathlib import Path
+from typing import Optional
+import random
+from DeepWEBS.utilsdw.enver import enver
+from DeepWEBS.utilsdw.logger import logger
+from DeepWEBS.networks.filepath_converter import QueryToFilepathConverter
+from DeepWEBS.networks.network_configs import REQUESTS_HEADERS
+
+class GoogleSearcher:
+    def __init__(self):
+        self.url = "https://www.google.com/search"
+        self.enver = enver
+        self.enver.set_envs(proxies=True)
+        self.filepath_converter = QueryToFilepathConverter()
+
+    def send_request(self, query: str, result_num: int = 10, safe: bool = False) -> requests.Response:
+        params = {
+            "q": query,
+            "num": result_num,
+        }
+        response = requests.get(
+            self.url,
+            headers=REQUESTS_HEADERS,
+            params=params,
+            proxies=self.enver.requests_proxies,
+        )
+        response.raise_for_status()  # Raise an exception for non-2xx status codes
+        return response
+
+    def save_response(self, response: requests.Response, html_path: Path) -> None:
+        html_path.parent.mkdir(parents=True, exist_ok=True)
+        logger.note(f"Saving to: [{html_path}]")
+        with html_path.open("wb") as wf:
+            wf.write(response.content)
+
+    def search(self, query: str, result_num: int = 10, safe: bool = False, overwrite: bool = False) -> Path:
+        html_path = self.filepath_converter.convert(query)
+        logger.note(f"Searching: [{query}]")
+
+        if html_path.exists() and not overwrite:
+            logger.success(f"HTML existed: {html_path}")
+        else:
+            response = self.send_request(query, result_num, safe)
+            self.save_response(response, html_path)
+
+        return html_path
+
+if __name__ == "__main__":
+    searcher = GoogleSearcher()
+    html_path = searcher.search("python tutorials")
+    print(f"HTML file saved at: {html_path}")

webscout-1.2.3/DeepWEBS/networks/network_configs.py

@@ -0,0 +1,30 @@
+IGNORE_TAGS = ["script", "style", "button"]
+IGNORE_CLASSES = [
+    # common
+    "sidebar",
+    "footer",
+    "related",
+    "comment",
+    "topbar",
+    "offcanvas",
+    "navbar",
+    # 163.com
+    "post_(top)|(side)|(recommends)|(crumb)|(statement)|(next)|(jubao)",
+    "ntes\-.*nav",
+    "nav\-bottom",
+    # wikipedia.org
+    "language\-list",
+    "vector\-(header)|(column)|(sticky\-pinned)|(dropdown\-content)",
+    "navbox",
+    "catlinks",
+]
+
+IGNORE_HOSTS = [
+    "weibo.com",
+    "hymson.com",
+    "yahoo.com",
+]
+
+REQUESTS_HEADERS = {
+    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/111.0.0.0 Safari/537.36 Edg/111.0.1661.62",
+}

webscout-1.2.3/DeepWEBS/networks/webpage_fetcher.py

@@ -0,0 +1,97 @@
+import concurrent.futures
+import random
+import requests
+import tldextract
+from pathlib import Path
+from typing import List, Tuple, Dict
+
+from DeepWEBS.utilsdw.enver import enver
+from DeepWEBS.utilsdw.logger import logger
+from DeepWEBS.networks.filepath_converter import UrlToFilepathConverter
+from DeepWEBS.networks.network_configs import IGNORE_HOSTS, REQUESTS_HEADERS
+
+class WebpageFetcher:
+    def __init__(self):
+        self.enver = enver
+        self.enver.set_envs(proxies=True)
+        self.filepath_converter = UrlToFilepathConverter()
+
+    def is_ignored_host(self, url: str) -> bool:
+        host = tldextract.extract(url).registered_domain
+        return host in IGNORE_HOSTS
+
+    def send_request(self, url: str) -> requests.Response:
+        try:
+            user_agent = random.choice(REQUESTS_HEADERS["User-Agent"])
+            response = requests.get(
+                url=url,
+                headers={"User-Agent": user_agent},
+                proxies=self.enver.requests_proxies,
+                timeout=15,
+            )
+            response.raise_for_status()
+            return response
+        except requests.exceptions.RequestException as e:
+            logger.warn(f"Failed to fetch: [{url}] | {e}")
+            return None
+
+    def save_response(self, response: requests.Response, html_path: Path) -> None:
+        if response is None:
+            return
+
+        html_path.parent.mkdir(parents=True, exist_ok=True)
+        logger.success(f"Saving to: [{html_path}]")
+        with html_path.open("wb") as wf:
+            wf.write(response.content)
+
+    def fetch(self, url: str, overwrite: bool = False, output_parent: str = None) -> Path:
+        logger.note(f"Fetching: [{url}]")
+        html_path = self.filepath_converter.convert(url, parent=output_parent)
+
+        if self.is_ignored_host(url):
+            logger.warn(f"Ignored host: [{tldextract.extract(url).registered_domain}]")
+            return html_path
+
+        if html_path.exists() and not overwrite:
+            logger.success(f"HTML existed: [{html_path}]")
+        else:
+            response = self.send_request(url)
+            self.save_response(response, html_path)
+
+        return html_path
+
+class BatchWebpageFetcher:
+    def __init__(self):
+        self.done_count = 0
+        self.total_count = 0
+        self.url_and_html_path_list: List[Dict[str, str]] = []
+
+    def fetch_single_webpage(self, url: str, overwrite: bool = False, output_parent: str = None) -> Tuple[str, Path]:
+        webpage_fetcher = WebpageFetcher()
+        html_path = webpage_fetcher.fetch(url, overwrite, output_parent)
+        self.url_and_html_path_list.append({"url": url, "html_path": str(html_path)})
+        self.done_count += 1
+        logger.success(f"> [{self.done_count}/{self.total_count}] Fetched: {url}")
+        return url, html_path
+
+    def fetch(self, urls: List[str], overwrite: bool = False, output_parent: str = None) -> List[Dict[str, str]]:
+        self.urls = urls
+        self.total_count = len(self.urls)
+
+        with concurrent.futures.ThreadPoolExecutor() as executor:
+            futures = [
+                executor.submit(self.fetch_single_webpage, url, overwrite, output_parent)
+                for url in urls
+            ]
+            concurrent.futures.wait(futures)
+
+        return self.url_and_html_path_list
+
+if __name__ == "__main__":
+    urls = [
+        "https://stackoverflow.com/questions/295135/turn-a-string-into-a-valid-filename",
+        "https://www.liaoxuefeng.com/wiki/1016959663602400/1017495723838528",
+        "https://docs.python.org/zh-cn/3/tutorial/interpreter.html",
+    ]
+    batch_webpage_fetcher = BatchWebpageFetcher()
+    batch_webpage_fetcher.fetch(urls=urls, overwrite=True, output_parent="python tutorials")

webscout-1.2.3/DeepWEBS/utilsdw/enver.py

@@ -0,0 +1,60 @@
+import json
+import os
+
+from pathlib import Path
+from DeepWEBS.utilsdw.logger import logger
+
+
+class OSEnver:
+    def __init__(self):
+        self.envs_stack = []
+        self.envs = os.environ.copy()
+
+    def store_envs(self):
+        self.envs_stack.append(self.envs)
+
+    def restore_envs(self):
+        self.envs = self.envs_stack.pop()
+
+    def set_envs(self, secrets=True, proxies=None, store_envs=True):
+        # caller_info = inspect.stack()[1]
+        # logger.back(f"OS Envs is set by: {caller_info.filename}")
+
+        if store_envs:
+            self.store_envs()
+
+        if secrets:
+            secrets_path = Path(__file__).parents[1] / "secrets.json"
+            if secrets_path.exists():
+                with open(secrets_path, "r") as rf:
+                    secrets = json.load(rf)
+            else:
+                secrets = {}
+
+        if proxies:
+            for proxy_env in ["http_proxy", "https_proxy"]:
+                if isinstance(proxies, str):
+                    self.envs[proxy_env] = proxies
+                elif "http_proxy" in secrets.keys():
+                    self.envs[proxy_env] = secrets["http_proxy"]
+                elif os.getenv("http_proxy"):
+                    self.envs[proxy_env] = os.getenv("http_proxy")
+                else:
+                    continue
+
+        self.proxy = (
+            self.envs.get("all_proxy")
+            or self.envs.get("http_proxy")
+            or self.envs.get("https_proxy")
+            or None
+        )
+        self.requests_proxies = {
+            "http": self.proxy,
+            "https": self.proxy,
+        }
+
+        if self.proxy:
+            logger.note(f"Using proxy: [{self.proxy}]")
+
+
+enver = OSEnver()