diff --git a/archivebox/config/__init__.py b/archivebox/config/__init__.py index 6627c736..82d401cd 100644 --- a/archivebox/config/__init__.py +++ b/archivebox/config/__init__.py @@ -81,6 +81,7 @@ CONFIG_DEFAULTS: Dict[str, ConfigDefaultDict] = { 'SAVE_WGET_REQUISITES': {'type': bool, 'default': True, 'aliases': ('FETCH_WGET_REQUISITES',)}, 'SAVE_SINGLEFILE': {'type': bool, 'default': True, 'aliases': ('FETCH_SINGLEFILE',)}, 'SAVE_READABILITY': {'type': bool, 'default': True, 'aliases': ('FETCH_READABILITY',)}, + 'SAVE_MERCURY': {'type': bool, 'default': True, 'aliases': ('FETCH_MERCURY',)}, 'SAVE_PDF': {'type': bool, 'default': True, 'aliases': ('FETCH_PDF',)}, 'SAVE_SCREENSHOT': {'type': bool, 'default': True, 'aliases': ('FETCH_SCREENSHOT',)}, 'SAVE_DOM': {'type': bool, 'default': True, 'aliases': ('FETCH_DOM',)}, @@ -112,6 +113,7 @@ CONFIG_DEFAULTS: Dict[str, ConfigDefaultDict] = { 'USE_WGET': {'type': bool, 'default': True}, 'USE_SINGLEFILE': {'type': bool, 'default': True}, 'USE_READABILITY': {'type': bool, 'default': True}, + 'USE_MERCURY': {'type': bool, 'default': True}, 'USE_GIT': {'type': bool, 'default': True}, 'USE_CHROME': {'type': bool, 'default': True}, 'USE_NODE': {'type': bool, 'default': True}, @@ -122,6 +124,7 @@ CONFIG_DEFAULTS: Dict[str, ConfigDefaultDict] = { 'WGET_BINARY': {'type': str, 'default': 'wget'}, 'SINGLEFILE_BINARY': {'type': str, 'default': 'single-file'}, 'READABILITY_BINARY': {'type': str, 'default': 'readability-extractor'}, + 'MERCURY_BINARY': {'type': str, 'default': 'mercury-parser'}, 'YOUTUBEDL_BINARY': {'type': str, 'default': 'youtube-dl'}, 'CHROME_BINARY': {'type': str, 'default': None}, }, @@ -265,6 +268,9 @@ DERIVED_CONFIG_DEFAULTS: ConfigDefaultDict = { 'USE_READABILITY': {'default': lambda c: c['USE_READABILITY'] and c['SAVE_READABILITY']}, 'READABILITY_VERSION': {'default': lambda c: bin_version(c['READABILITY_BINARY']) if c['USE_READABILITY'] else None}, + 'USE_MERCURY': {'default': lambda c: c['USE_MERCURY'] and c['SAVE_MERCURY']}, + 'MERCURY_VERSION': {'default': lambda c: bin_version(c['MERCURY_BINARY']) if c['USE_MERCURY'] else None}, + 'USE_GIT': {'default': lambda c: c['USE_GIT'] and c['SAVE_GIT']}, 'GIT_VERSION': {'default': lambda c: bin_version(c['GIT_BINARY']) if c['USE_GIT'] else None}, 'SAVE_GIT': {'default': lambda c: c['USE_GIT'] and c['SAVE_GIT']}, @@ -283,6 +289,7 @@ DERIVED_CONFIG_DEFAULTS: ConfigDefaultDict = { 'SAVE_DOM': {'default': lambda c: c['USE_CHROME'] and c['SAVE_DOM']}, 'SAVE_SINGLEFILE': {'default': lambda c: c['USE_CHROME'] and c['USE_SINGLEFILE'] and c['USE_NODE']}, 'SAVE_READABILITY': {'default': lambda c: c['USE_READABILITY'] and c['USE_NODE']}, + 'SAVE_MERCURY': {'default': lambda c: c['USE_MERCURY'] and c['USE_NODE']}, 'DEPENDENCIES': {'default': lambda c: get_dependency_info(c)}, 'CODE_LOCATIONS': {'default': lambda c: get_code_locations(c)}, @@ -728,6 +735,13 @@ def get_dependency_info(config: ConfigDict) -> ConfigValue: 'enabled': config['USE_READABILITY'], 'is_valid': bool(config['READABILITY_VERSION']), }, + 'MERCURY_BINARY': { + 'path': bin_path(config['MERCURY_BINARY']), + 'version': config['MERCURY_VERSION'], + 'hash': bin_hash(config['MERCURY_BINARY']), + 'enabled': config['USE_MERCURY'], + 'is_valid': bool(config['MERCURY_VERSION']), + }, 'GIT_BINARY': { 'path': bin_path(config['GIT_BINARY']), 'version': config['GIT_VERSION'], diff --git a/archivebox/config/stubs.py b/archivebox/config/stubs.py index 4a58110d..4f09c1e0 100644 --- a/archivebox/config/stubs.py +++ b/archivebox/config/stubs.py @@ -57,6 +57,7 @@ class ConfigDict(BaseConfig, total=False): SAVE_WGET_REQUISITES: bool SAVE_SINGLEFILE: bool SAVE_READABILITY: bool + SAVE_MERCURY: bool SAVE_PDF: bool SAVE_SCREENSHOT: bool SAVE_DOM: bool @@ -81,6 +82,7 @@ class ConfigDict(BaseConfig, total=False): USE_WGET: bool USE_SINGLEFILE: bool USE_READABILITY: bool + USE_MERCURY: bool USE_GIT: bool USE_CHROME: bool USE_YOUTUBEDL: bool @@ -89,6 +91,7 @@ class ConfigDict(BaseConfig, total=False): WGET_BINARY: str SINGLEFILE_BINARY: str READABILITY_BINARY: str + MERCURY_BINARY: str YOUTUBEDL_BINARY: str CHROME_BINARY: Optional[str] diff --git a/archivebox/extractors/mercury.py b/archivebox/extractors/mercury.py new file mode 100644 index 00000000..437bd955 --- /dev/null +++ b/archivebox/extractors/mercury.py @@ -0,0 +1,95 @@ +__package__ = 'archivebox.extractors' + +from pathlib import Path +from tempfile import NamedTemporaryFile + +from typing import Optional +import json + +from ..index.schema import Link, ArchiveResult, ArchiveError +from ..system import run, atomic_write +from ..util import ( + enforce_types, + download_url, + is_static_file, + +) +from ..config import ( + TIMEOUT, + CURL_BINARY, + SAVE_MERCURY, + DEPENDENCIES, + MERCURY_VERSION, +) +from ..logging_util import TimedProgress + +@enforce_types +def should_save_mercury(link: Link, out_dir: Optional[str]=None) -> bool: + out_dir = out_dir or link.link_dir + if is_static_file(link.url): + return False + + output = Path(out_dir or link.link_dir) / 'mercury' + return SAVE_MERCURY and MERCURY_VERSION and (not output.exists()) + + +@enforce_types +def save_mercury(link: Link, out_dir: Optional[str]=None, timeout: int=TIMEOUT) -> ArchiveResult: + """download reader friendly version using @postlight/mercury-parser""" + + out_dir = Path(out_dir or link.link_dir) + output_folder = out_dir.absolute() / "mercury" + output = str(output_folder) + + status = 'succeeded' + timer = TimedProgress(timeout, prefix=' ') + try: + cmd = [ + DEPENDENCIES['MERCURY_BINARY']['path'], + link.url, + "--format=text" + ] + result = run(cmd, cwd=out_dir, timeout=timeout) + txtresult_json = json.loads(result.stdout) + + cmd = [ + DEPENDENCIES['MERCURY_BINARY']['path'], + link.url + ] + result = run(cmd, cwd=out_dir, timeout=timeout) + result_json = json.loads(result.stdout) + + output_folder.mkdir(exist_ok=True) + atomic_write(str(output_folder / "content.html"), result_json.pop("content")) + atomic_write(str(output_folder / "content.txt"), txtresult_json["content"]) + atomic_write(str(output_folder / "article.json"), result_json) + + # parse out number of files downloaded from last line of stderr: + # "Downloaded: 76 files, 4.0M in 1.6s (2.52 MB/s)" + output_tail = [ + line.strip() + for line in (result.stdout + result.stderr).decode().rsplit('\n', 20)[-20:] + if line.strip() + ] + hints = ( + 'Got mercury response code: {}.'.format(result.returncode), + *output_tail, + ) + + # Check for common failure cases + if (result.returncode > 0): + raise ArchiveError('Mercury parser was not able to archive the page', hints) + except (Exception, OSError) as err: + status = 'failed' + output = err + finally: + timer.end() + + return ArchiveResult( + cmd=cmd, + pwd=str(out_dir), + cmd_version=MERCURY_VERSION, + output=output, + status=status, + **timer.stats, + )