From 15efb2d5ed1163fb5f6388646fb167efa7dd1afa Mon Sep 17 00:00:00 2001
From: Nick Sweeting <git@sweeting.me>
Date: Tue, 18 Aug 2020 08:29:05 -0400
Subject: [PATCH] new generic_html parser for extracting hrefs
---
archivebox.egg-info/SOURCES.txt | 1 +
archivebox/index/__init__.py | 4 +-
archivebox/main.py | 6 +--
archivebox/parsers/__init__.py | 81 +++++++++++++++++-------------
archivebox/parsers/generic_html.py | 53 +++++++++++++++++++
5 files changed, 106 insertions(+), 39 deletions(-)
create mode 100644 archivebox/parsers/generic_html.py
diff --git a/archivebox.egg-info/SOURCES.txt b/archivebox.egg-info/SOURCES.txt
index 14e510a1..ee6a2fc5 100644
--- a/archivebox.egg-info/SOURCES.txt
+++ b/archivebox.egg-info/SOURCES.txt
@@ -70,6 +70,7 @@ archivebox/index/json.py
archivebox/index/schema.py
archivebox/index/sql.py
archivebox/parsers/__init__.py
+archivebox/parsers/generic_html.py
archivebox/parsers/generic_json.py
archivebox/parsers/generic_rss.py
archivebox/parsers/generic_txt.py
diff --git a/archivebox/index/__init__.py b/archivebox/index/__init__.py
index b7c8ebff..784c879c 100644
--- a/archivebox/index/__init__.py
+++ b/archivebox/index/__init__.py
@@ -301,14 +301,14 @@ def load_main_index_meta(out_dir: str=OUTPUT_DIR) -> Optional[dict]:
@enforce_types
-def parse_links_from_source(source_path: str) -> Tuple[List[Link], List[Link]]:
+def parse_links_from_source(source_path: str, root_url: Optional[str]=None) -> Tuple[List[Link], List[Link]]:
from ..parsers import parse_links
new_links: List[Link] = []
# parse and validate the import file
- raw_links, parser_name = parse_links(source_path)
+ raw_links, parser_name = parse_links(source_path, root_url=root_url)
new_links = validate_links(raw_links)
if parser_name:
diff --git a/archivebox/main.py b/archivebox/main.py
index 2e17594b..b65c6e64 100644
--- a/archivebox/main.py
+++ b/archivebox/main.py
@@ -548,7 +548,7 @@ def add(urls: Union[str, List[str]],
# save verbatim args to sources
write_ahead_log = save_text_as_source('\n'.join(urls), filename='{ts}-import.txt', out_dir=out_dir)
- new_links += parse_links_from_source(write_ahead_log)
+ new_links += parse_links_from_source(write_ahead_log, root_url=None)
# If we're going one level deeper, download each link and look for more links
new_links_depth = []
@@ -556,9 +556,9 @@ def add(urls: Union[str, List[str]],
log_crawl_started(new_links)
for new_link in new_links:
downloaded_file = save_file_as_source(new_link.url, filename=f'{new_link.timestamp}-crawl-{new_link.domain}.txt', out_dir=out_dir)
- new_links_depth += parse_links_from_source(downloaded_file)
+ new_links_depth += parse_links_from_source(downloaded_file, root_url=new_link.url)
- imported_links = new_links + new_links_depth
+ imported_links = list({link.url: link for link in (new_links + new_links_depth)}.values())
all_links, new_links = dedupe_links(all_links, imported_links)
write_main_index(links=all_links, out_dir=out_dir, finished=not new_links)
diff --git a/archivebox/parsers/__init__.py b/archivebox/parsers/__init__.py
index bde71c27..930e1ade 100644
--- a/archivebox/parsers/__init__.py
+++ b/archivebox/parsers/__init__.py
@@ -11,7 +11,7 @@ import re
import os
from io import StringIO
-from typing import IO, Tuple, List
+from typing import IO, Tuple, List, Optional
from datetime import datetime
from ..system import atomic_write
@@ -38,26 +38,29 @@ from .medium_rss import parse_medium_rss_export
from .netscape_html import parse_netscape_html_export
from .generic_rss import parse_generic_rss_export
from .generic_json import parse_generic_json_export
+from .generic_html import parse_generic_html_export
from .generic_txt import parse_generic_txt_export
PARSERS = (
- # Specialized parsers
- ('Pocket HTML', parse_pocket_html_export),
- ('Pinboard RSS', parse_pinboard_rss_export),
- ('Shaarli RSS', parse_shaarli_rss_export),
- ('Medium RSS', parse_medium_rss_export),
-
- # General parsers
- ('Netscape HTML', parse_netscape_html_export),
- ('Generic RSS', parse_generic_rss_export),
- ('Generic JSON', parse_generic_json_export),
+ # Specialized parsers
+ ('Pocket HTML', parse_pocket_html_export),
+ ('Pinboard RSS', parse_pinboard_rss_export),
+ ('Shaarli RSS', parse_shaarli_rss_export),
+ ('Medium RSS', parse_medium_rss_export),
+
+ # General parsers
+ ('Netscape HTML', parse_netscape_html_export),
+ ('Generic RSS', parse_generic_rss_export),
+ ('Generic JSON', parse_generic_json_export),
+ ('Generic HTML', parse_generic_html_export),
+
+ # Fallback parser
+ ('Plain Text', parse_generic_txt_export),
+)
- # Fallback parser
- ('Plain Text', parse_generic_txt_export),
- )
@enforce_types
-def parse_links_memory(urls: List[str]):
+def parse_links_memory(urls: List[str], root_url: Optional[str]=None):
"""
parse a list of URLS without touching the filesystem
"""
@@ -68,17 +71,16 @@ def parse_links_memory(urls: List[str]):
file = StringIO()
file.writelines(urls)
file.name = "io_string"
- output = _parse(file, timer)
-
- if output is not None:
- return output
-
+ links, parser = run_parser_functions(file, timer, root_url=root_url)
timer.end()
- return [], 'Failed to parse'
+
+ if parser is None:
+ return [], 'Failed to parse'
+ return links, parser
@enforce_types
-def parse_links(source_file: str) -> Tuple[List[Link], str]:
+def parse_links(source_file: str, root_url: Optional[str]=None) -> Tuple[List[Link], str]:
"""parse a list of URLs with their metadata from an
RSS feed, bookmarks export, or text file
"""
@@ -87,28 +89,39 @@ def parse_links(source_file: str) -> Tuple[List[Link], str]:
timer = TimedProgress(TIMEOUT * 4)
with open(source_file, 'r', encoding='utf-8') as file:
- output = _parse(file, timer)
-
- if output is not None:
- return output
+ links, parser = run_parser_functions(file, timer, root_url=root_url)
timer.end()
- return [], 'Failed to parse'
+ if parser is None:
+ return [], 'Failed to parse'
+ return links, parser
+
+
+def run_parser_functions(to_parse: IO[str], timer, root_url: Optional[str]=None) -> Tuple[List[Link], Optional[str]]:
+ most_links: List[Link] = []
+ best_parser_name = None
-def _parse(to_parse: IO[str], timer) -> Tuple[List[Link], str]:
for parser_name, parser_func in PARSERS:
try:
- links = list(parser_func(to_parse))
- if links:
- timer.end()
- return links, parser_name
- except Exception as err: # noqa
- pass
+ parsed_links = list(parser_func(to_parse, root_url=root_url))
+ if not parsed_links:
+ raise Exception('no links found')
+
+ # print(f'[√] Parser {parser_name} succeeded: {len(parsed_links)} links parsed')
+ if len(parsed_links) > len(most_links):
+ most_links = parsed_links
+ best_parser_name = parser_name
+
+ except Exception as err: # noqa
# Parsers are tried one by one down the list, and the first one
# that succeeds is used. To see why a certain parser was not used
# due to error or format incompatibility, uncomment this line:
+
# print('[!] Parser {} failed: {} {}'.format(parser_name, err.__class__.__name__, err))
# raise
+ pass
+ timer.end()
+ return most_links, best_parser_name
@enforce_types
diff --git a/archivebox/parsers/generic_html.py b/archivebox/parsers/generic_html.py
new file mode 100644
index 00000000..4c632f04
--- /dev/null
+++ b/archivebox/parsers/generic_html.py
@@ -0,0 +1,53 @@
+__package__ = 'archivebox.parsers'
+
+
+import re
+
+from typing import IO, Iterable, Optional
+from datetime import datetime
+
+from ..index.schema import Link
+from ..util import (
+ htmldecode,
+ enforce_types,
+ URL_REGEX,
+)
+from html.parser import HTMLParser
+from urllib.parse import urljoin
+
+
+class HrefParser(HTMLParser):
+ def __init__(self):
+ super().__init__()
+ self.urls = []
+
+ def handle_starttag(self, tag, attrs):
+ if tag == "a":
+ for attr, value in attrs:
+ if attr == "href":
+ self.urls.append(value)
+
+
+@enforce_types
+def parse_generic_html_export(html_file: IO[str], root_url: Optional[str]=None, **_kwargs) -> Iterable[Link]:
+ """Parse Pocket-format bookmarks export files (produced by getpocket.com/export/)"""
+
+ html_file.seek(0)
+ for line in html_file:
+ parser = HrefParser()
+ # example line
+ # <li><a href="http://example.com/ time_added="1478739709" tags="tag1,tag2">example title</a></li>
+ parser.feed(line)
+ for url in parser.urls:
+ if root_url:
+ # resolve relative urls /home.html -> https://example.com/home.html
+ url = urljoin(root_url, url)
+
+ for archivable_url in re.findall(URL_REGEX, url):
+ yield Link(
+ url=htmldecode(archivable_url),
+ timestamp=str(datetime.now().timestamp()),
+ title=None,
+ tags=None,
+ sources=[html_file.name],
+ )