wxpath 0.1.1__py3-none-any.whl → 0.3.0__py3-none-any.whl

wxpath/__init__.py

@@ -0,0 +1,9 @@
+from .core.runtime.engine import wxpath_async, wxpath_async_blocking, wxpath_async_blocking_iter
+from .util.logging import configure_logging
+
+__all__ = [
+    'wxpath_async', 
+    'wxpath_async_blocking',
+    'wxpath_async_blocking_iter',
+    'configure_logging',
+]

wxpath/cli.py

@@ -0,0 +1,92 @@
+import argparse
+import json
+import sys
+
+from wxpath.core import parser as wxpath_parser
+from wxpath.core.runtime.engine import WXPathEngine, wxpath_async_blocking_iter
+from wxpath.hooks import builtin, registry
+from wxpath.http.client.crawler import Crawler
+from wxpath.util.serialize import simplify
+
+
+def main():
+    registry.register(builtin.SerializeXPathMapAndNodeHook)
+    arg_parser = argparse.ArgumentParser(description="Run wxpath expression.")
+    arg_parser.add_argument("expression", help="The wxpath expression")
+    arg_parser.add_argument("--depth", type=int, default=1, help="Recursion depth")
+    # debug
+    arg_parser.add_argument("--debug", action="store_true", help="Debug mode")
+    # verbose
+    arg_parser.add_argument("--verbose", action="store_true", help="Verbose mode")
+    
+    arg_parser.add_argument(
+        "--concurrency", 
+        type=int, 
+        default=16, 
+        help="Number of concurrent fetches"
+    )
+    arg_parser.add_argument(
+        "--concurrency-per-host", 
+        type=int,
+        default=8,
+        help="Number of concurrent fetches per host"
+    )
+    arg_parser.add_argument(
+        "--header",
+        action="append",
+        dest="header_list",
+        default=[],
+        help="Add a custom header (e.g., 'Key:Value'). Can be used multiple times.",
+    )
+    arg_parser.add_argument(
+        "--respect-robots",
+        action="store_true",
+        help="Respect robots.txt",
+        default=True
+    )
+
+    args = arg_parser.parse_args()
+
+    if args.verbose:
+        segments = wxpath_parser.parse(args.expression)
+        print("parsed expression:\n\nSegments([")
+        for s in segments:
+            print(f"\t{s},")
+        print("])")
+        print()
+
+    if args.debug:
+        from wxpath import configure_logging
+        configure_logging('DEBUG')
+
+    custom_headers = {}
+    if args.header_list:
+        for header_item in args.header_list:
+            try:
+                key, value = header_item.split(':', 1)
+                custom_headers[key.strip()] = value.strip()
+            except ValueError:
+                print(f"Warning: Invalid header format '{header_item}'. Use 'Key:Value'.")
+
+    if custom_headers and args.verbose:
+        print(f"Using custom headers: {custom_headers}")
+        print()
+
+    crawler = Crawler(
+        concurrency=args.concurrency,
+        per_host=args.concurrency_per_host,
+        respect_robots=args.respect_robots,
+        headers=custom_headers
+    )
+    engine = WXPathEngine(crawler=crawler)
+
+    try:
+        for r in wxpath_async_blocking_iter(args.expression, args.depth, engine):
+            clean = simplify(r)
+            print(json.dumps(clean, ensure_ascii=False), flush=True)
+    except BrokenPipeError:
+        sys.exit(0)
+
+
+if __name__ == "__main__":
+    main()

wxpath/core/__init__.py

@@ -0,0 +1,13 @@
+from wxpath.core.runtime.engine import (
+    WXPathEngine,
+    wxpath_async,
+    wxpath_async_blocking,
+    wxpath_async_blocking_iter,
+)
+
+__all__ = [
+    'wxpath_async', 
+    'wxpath_async_blocking',
+    'wxpath_async_blocking_iter',
+    'WXPathEngine',
+]

wxpath/core/dom.py

@@ -0,0 +1,22 @@
+from urllib.parse import urljoin
+
+
+def _make_links_absolute(links: list[str], base_url: str) -> list[str]:
+    """
+    Convert relative links to absolute links based on the base URL.
+
+    Args:
+        links (list): List of link strings.
+        base_url (str): The base URL to resolve relative links against.
+
+    Returns:
+        List of absolute URLs.
+    """
+    if base_url is None:
+        raise ValueError("base_url must not be None when making links absolute.")
+    return [urljoin(base_url, link) for link in links if link]
+
+
+def get_absolute_links_from_elem_and_xpath(elem, xpath):
+    base_url = getattr(elem, 'base_url', None)
+    return _make_links_absolute(elem.xpath3(xpath), base_url)

wxpath/core/models.py

@@ -0,0 +1,74 @@
+from dataclasses import dataclass, field
+from typing import Any, List, Optional, Tuple
+
+
+@dataclass(slots=True)
+class CrawlTask:
+    """A unit of work for the crawler."""
+    elem: Any
+    url: str
+    segments: List[Tuple[str, str]]
+    depth: int
+    backlink: Optional[str] = None
+    base_url: Optional[str] = None
+    
+    # Priority for the queue (lower number = higher priority)
+    # Useful if you want Depth-First behavior in a shared queue
+    priority: int = field(default=0)
+
+    def __post_init__(self):
+        # Automatically sync priority with depth for BFS behavior
+        self.priority = self.depth
+
+    def __lt__(self, other):
+        return self.priority < other.priority
+
+    def __iter__(self):
+        return iter((self.elem, self.segments, self.depth, self.backlink))
+
+
+@dataclass(slots=True)
+class Intent:
+    pass
+
+
+@dataclass(slots=True)
+class Result(Intent):
+    """A container for an extracted item or error."""
+    value: Any
+    url: str
+    depth: int
+    error: Optional[Exception] = None
+    backlink: Optional[str] = None
+
+
+@dataclass(slots=True)
+class CrawlIntent(Intent):
+    url: str             # "I found this link"
+    next_segments: list  # "Here is what to do next if you go there"
+
+
+@dataclass(slots=True)
+class ProcessIntent(Intent):
+    elem: Any
+    next_segments: list
+
+
+@dataclass(slots=True)
+class InfiniteCrawlIntent(ProcessIntent):
+    pass
+
+
+@dataclass(slots=True)
+class ExtractIntent(ProcessIntent):
+    pass
+
+
+@dataclass(slots=True)
+class CrawlFromAttributeIntent(ProcessIntent):
+    pass
+
+
+@dataclass(slots=True)
+class DataIntent(Intent):
+    value: Any

wxpath/core/ops.py

@@ -0,0 +1,278 @@
+from typing import Callable, Iterable
+from urllib.parse import urljoin
+
+import elementpath
+from elementpath.datatypes import AnyAtomicType
+from elementpath.xpath3 import XPath3Parser
+from lxml import html
+
+from wxpath.core.dom import get_absolute_links_from_elem_and_xpath
+from wxpath.core.models import (
+    CrawlIntent,
+    DataIntent,
+    ExtractIntent,
+    InfiniteCrawlIntent,
+    Intent,
+    ProcessIntent,
+)
+from wxpath.core.parser import (
+    Binary,
+    Call,
+    ContextItem,
+    Segment,
+    Segments,
+    String,
+    Url,
+    UrlCrawl,
+    Xpath,
+)
+from wxpath.util.logging import get_logger
+
+log = get_logger(__name__)
+
+
+class WxStr(str):
+    """A string with associated base_url and depth metadata for debugging."""
+    def __new__(cls, value, base_url=None, depth=-1):
+        obj = super().__new__(cls, value)
+        obj.base_url = base_url
+        obj.depth = depth
+        return obj
+
+    def __repr__(self):
+        return f"WxStr({super().__repr__()}, base_url={self.base_url!r}, depth={self.depth})"
+
+
+class RuntimeSetupError(Exception):
+    pass
+
+
+OPS_REGISTER: dict[str, Callable] = {}
+
+def register(func_name_or_type: str | type, args_types: tuple[type, ...] | None = None):
+    def _register(func: Callable) -> Callable:
+        global OPS_REGISTER
+        _key = (func_name_or_type, args_types) if args_types else func_name_or_type
+        if _key in OPS_REGISTER:
+            raise RuntimeSetupError(f"The operation handler for \"{_key}\" already registered")
+        OPS_REGISTER[_key] = func
+        return func
+    return _register
+
+
+def get_operator(
+        binary_or_segment: Binary | Segment
+    ) -> Callable[[html.HtmlElement, list[Url | Xpath], int], Iterable[Intent]]:
+    func_name_or_type = getattr(binary_or_segment, 'func', None) or binary_or_segment.__class__
+
+    args_types = None
+    if isinstance(binary_or_segment, Binary):
+        args_types = (binary_or_segment.left.__class__, binary_or_segment.right.__class__)
+    elif isinstance(binary_or_segment, Call):
+        args_types = tuple(arg.__class__ for arg in binary_or_segment.args)
+
+    _key = (func_name_or_type, args_types) if args_types else func_name_or_type
+    if _key not in OPS_REGISTER:
+        raise ValueError(f"Unknown operation: {_key}")
+    return OPS_REGISTER[_key]
+
+
+@register('url', (String,))
+@register('url', (String, Xpath))
+def _handle_url_str_lit(curr_elem: html.HtmlElement, 
+                        curr_segments: list[Url | Xpath], 
+                        curr_depth: int, **kwargs) -> Iterable[Intent]:
+    """Handle `url('<literal>')` segments and optional follow xpath."""
+    url_call = curr_segments[0] # type: Url
+
+    next_segments = curr_segments[1:]
+
+    if len(url_call.args) == 2:
+        _segments = [
+            UrlCrawl('///url', [url_call.args[1], url_call.args[0].value])
+        ] + next_segments
+        
+        yield CrawlIntent(url=url_call.args[0].value, next_segments=_segments)
+    else:
+        yield CrawlIntent(url=url_call.args[0].value, next_segments=next_segments)
+
+
+# @register2('url', (Xpath,))
+@register(Xpath)
+def _handle_xpath(curr_elem: html.HtmlElement,
+                  curr_segments: Segments,
+                  curr_depth: int,
+                  **kwargs) -> Iterable[Intent]:
+    """Execute an xpath step and yield data or chained processing intents."""
+    xpath_node = curr_segments[0] # type: Xpath
+
+    expr = xpath_node.value
+
+    if curr_elem is None:
+        raise ValueError("Element must be provided when path_expr does not start with 'url()'.")
+    base_url = getattr(curr_elem, 'base_url', None)
+    log.debug("base url", extra={"depth": curr_depth, "op": 'xpath', "base_url": base_url})
+
+    _backlink_str = f"string('{curr_elem.get('backlink')}')"
+    # We use the root tree's depth and not curr_depth because curr_depth accounts for a +1 
+    # increment after each url*() hop
+    _depth_str = f"number({curr_elem.getroottree().getroot().get('depth')})"
+    expr = expr.replace('wx:backlink()', _backlink_str)
+    expr = expr.replace('wx:backlink(.)', _backlink_str)
+    expr = expr.replace('wx:depth()', _depth_str)
+    expr = expr.replace('wx:depth(.)', _depth_str)
+
+    elems = curr_elem.xpath3(expr)
+    
+    next_segments = curr_segments[1:]
+    for elem in elems:
+        value_or_elem = WxStr(
+            elem, base_url=base_url, 
+            depth=curr_depth
+        ) if isinstance(elem, str) else elem
+        if len(curr_segments) == 1:
+            yield DataIntent(value=value_or_elem)
+        else:
+            yield ProcessIntent(elem=value_or_elem, next_segments=next_segments)
+
+
+@register('//url', (ContextItem,))
+@register('//url', (Xpath,))
+@register('/url', (ContextItem,))
+@register('/url', (Xpath,))
+@register('url', (ContextItem,))
+@register('url', (Xpath,))
+def _handle_url_eval(curr_elem: html.HtmlElement | str, 
+                     curr_segments: list[Url | Xpath], 
+                     curr_depth: int, 
+                     **kwargs) -> Iterable[Intent]:
+    """Resolve dynamic url() arguments and enqueue crawl intents.
+    
+    Yields:
+        CrawlIntent
+    """
+    url_call = curr_segments[0] # type: Url
+
+    if isinstance(url_call.args[0], ContextItem):
+        urls = [urljoin(getattr(curr_elem, 'base_url', None) or '', curr_elem)]
+    else:
+        _path_exp = url_call.args[0].value
+        # TODO: If prior xpath operation is XPATH_FN_MAP_FRAG, then this will likely fail.
+        # It should be handled in the parser.
+        urls = get_absolute_links_from_elem_and_xpath(curr_elem, _path_exp)
+        urls = dict.fromkeys(urls)
+
+    next_segments = curr_segments[1:]
+    for url in urls:
+        # log.debug("queueing", extra={"depth": curr_depth, "op": op, "url": url})
+        yield CrawlIntent(url=url, next_segments=next_segments)
+
+
+@register('///url', (Xpath,))
+def _handle_url_inf(curr_elem: html.HtmlElement, 
+                    curr_segments: list[Url | Xpath], 
+                    curr_depth: int, 
+                    **kwargs) -> Iterable[CrawlIntent]:
+    """Handle the ``///url()`` segment of a wxpath expression.
+
+    This operation is also generated internally by the parser when a
+    ``///<xpath>/[/]url()`` segment is encountered.
+
+    Instead of fetching URLs directly, this operator XPaths the current
+    element for URLs and queues them for further processing via
+    ``_handle_url_inf_and_xpath``.
+    """
+    url_call = curr_segments[0] # type: Url
+
+    _path_exp = url_call.args[0].value
+
+    urls = get_absolute_links_from_elem_and_xpath(curr_elem, _path_exp)
+
+    tail_segments = curr_segments[1:]
+    for url in dict.fromkeys(urls):
+        _segments = [
+            UrlCrawl('///url', [url_call.args[0], url])
+        ] + tail_segments
+        
+        yield CrawlIntent(url=url, next_segments=_segments)
+
+
+@register('///url', (Xpath, str))
+def _handle_url_inf_and_xpath(curr_elem: html.HtmlElement, 
+                              curr_segments: list[Url | Xpath], 
+                              curr_depth: int, **kwargs) \
+                                -> Iterable[DataIntent | ProcessIntent | InfiniteCrawlIntent]:
+    """Handle infinite-crawl with an xpath extraction step.
+
+    This operation is generated internally by the parser; there is no explicit
+    wxpath expression that produces it directly.
+
+    Yields:
+        DataIntent: If the current element is not None and no next segments are provided.
+        ExtractIntent: If the current element is not None and next segments are provided.
+        InfiniteCrawlIntent: If the current element is not None and next segments are provided.
+
+    Raises:
+        ValueError: If the current element is None.
+    """
+    url_call = curr_segments[0]
+
+    try:
+        if curr_elem is None:
+            raise ValueError("Missing element when op is 'url_inf_and_xpath'.")
+
+        next_segments = curr_segments[1:]
+
+        if not next_segments:
+            yield DataIntent(value=curr_elem)
+        else:
+            yield ExtractIntent(elem=curr_elem, next_segments=next_segments)
+
+        # For url_inf, also re-enqueue for further infinite expansion
+        _segments = [UrlCrawl('///url', url_call.args[:-1])] + next_segments
+        crawl_intent = InfiniteCrawlIntent(elem=curr_elem, next_segments=_segments)
+
+        yield crawl_intent
+
+    except Exception:
+        log.exception("error fetching url inf and xpath", 
+                      extra={"depth": curr_depth, "url": url_call.args[1]})
+
+@register(Binary, (Xpath, Segments))
+def _handle_binary(curr_elem: html.HtmlElement | str, 
+                              curr_segments: list[Url | Xpath] | Binary, 
+                              curr_depth: int, 
+                              **kwargs) -> Iterable[DataIntent | ProcessIntent]:
+    """Execute XPath expressions suffixed with the ``!`` (map) operator.
+
+    Yields:
+        ProcessIntent: Contrains either a WxStr or lxml or elementpath element.
+    """
+    left = curr_segments.left
+    _ = curr_segments.op
+    right = curr_segments.right
+
+    if len(right) == 0:
+        # Binary operation on segments expects non-empty segments
+        raise ValueError("Binary operation on segments expects non-empty segments")
+
+    base_url = getattr(curr_elem, 'base_url', None)
+    next_segments = right
+
+    results = elementpath.select(
+        curr_elem,
+        left.value,
+        parser=XPath3Parser,
+        item='' if curr_elem is None else None
+    )
+
+    if isinstance(results, AnyAtomicType):
+        results = [results]
+
+    for result in results:
+        if isinstance(result, str):
+            value_or_elem = WxStr(result, base_url=base_url, depth=curr_depth) 
+        else:
+            value_or_elem = result
+
+        yield ProcessIntent(elem=value_or_elem, next_segments=next_segments)