]> git.scottworley.com Git - paperdoorknob/blobdiff - paperdoorknob.py
Support _ in URLs
[paperdoorknob] / paperdoorknob.py
index 32daf4f4d9372b07ee1e5eb801037ea2efa129ad..72ec1a5669f48595fba7b640aff98de2da1b6069 100644 (file)
 # under the terms of the GNU General Public License as published by the
 # Free Software Foundation, version 3.
 
-
-from argparse import ArgumentParser
-import itertools
-import os.path
-import subprocess
-
-from typing import IO, Iterable
+from typing import Any, Iterable
 
 from bs4 import BeautifulSoup
-from bs4.element import Tag
-import requests
-import requests_cache
-from xdg_base_dirs import xdg_cache_home
-
-
-def command_line_parser() -> ArgumentParser:
-    parser = ArgumentParser(prog='paperdoorknob', description='Print glowfic')
-    parser.add_argument(
-        '--cache_path',
-        metavar='PATH',
-        help='Where to keep the http cache (instead of %(default)s)',
-        default=os.path.join(xdg_cache_home(), "paperdoorknob"))
-    parser.add_argument(
-        '--out',
-        help='The filename stem at which to write output ' +
-             '(eg: "%(default)s" produces %(default)s.text, %(default)s.pdf, etc.)',
-        default='book')
-    parser.add_argument('--pandoc', help='Location of the pandoc executable')
-    parser.add_argument(
-        '--timeout',
-        help='How long to wait for HTTP requests, in seconds',
-        default=30)
-    parser.add_argument('url', help='URL to retrieve')
-    return parser
-
-
-def fetch(url: str, session: requests.Session, timeout: int) -> BeautifulSoup:
-    with session.get(url, timeout=timeout) as r:
-        r.raise_for_status()
-        return BeautifulSoup(r.text, 'html.parser')
-
-
-def clean(html: BeautifulSoup) -> BeautifulSoup:
-    for eb in html.find_all("div", class_="post-edit-box"):
-        eb.decompose()
-    for footer in html.find_all("div", class_="post-footer"):
-        footer.decompose()
-    return html
-
-
-def replies(html: BeautifulSoup) -> Iterable[Tag]:
-    def text() -> Tag:
-        body = html.body
-        assert body
-        text = body.find_next("div", class_="post-post")
-        assert isinstance(text, Tag)
-        return text
-
-    def the_replies() -> Iterable[Tag]:
-        rs = html.find_all("div", class_="post-reply")
-        assert all(isinstance(r, Tag) for r in rs)
-        return rs
-
-    return itertools.chain([text()], the_replies())
-
-
-def html_to_tex(pandoc: str, tag: Tag) -> bytes:
-    return subprocess.run([pandoc, '--from=html', '--to=latex'],
-                          input=tag.encode(),
-                          stdout=subprocess.PIPE,
-                          check=True).stdout
-
 
-def process(
-        url: str,
-        session: requests.Session,
-        timeout: int,
-        texout: IO[bytes],
-        pandoc: str,
-) -> None:
-    html = clean(fetch(url, session, timeout))
-    for r in replies(html):
-        texout.write(html_to_tex(pandoc, r))
+from args import spec_from_commandline_args
+from glowfic import chunkDOMs, flatURL, makeChunk
+from spec import Spec
+
+
+def parse(content: bytes) -> BeautifulSoup:
+    return BeautifulSoup(content, 'html.parser')
+
+
+def ilen(it: Iterable[Any]) -> int:
+    return sum(1 for _ in it)
+
+
+def process(spec: Spec) -> None:
+    spec.texout.write(b'''\\documentclass{article}
+\\usepackage{booktabs}
+\\usepackage{graphicx}
+\\usepackage{longtable}
+\\usepackage{soul}
+\\usepackage{varwidth}
+\\usepackage{wrapstuff}
+''')
+    if spec.geometry is not None:
+        spec.texout.write(b'\\usepackage[' +
+                          spec.geometry.encode('UTF-8') +
+                          b']{geometry}\n')
+    spec.texout.write(b'''\\begin{document}
+\\newcommand{\\href}[2]{#2\\footnote{\\detokenize{#1}}}
+''')
+    url = flatURL(spec.url)
+    spec.log('Fetching HTML...\r')
+    html = spec.fetcher.fetch(url)
+    spec.log('Parsing HTML...\r')
+    dom = parse(spec.htmlfilter(html))
+    spec.log('Counting chunks...\r')
+    num_chunks = ilen(chunkDOMs(dom))
+    for i, r in enumerate(chunkDOMs(dom)):
+        percent = 100.0 * i / num_chunks
+        spec.log(f'Processing chunk {i} of {num_chunks} ({percent:.1f}%)\r')
+        spec.domfilter(r)
+        chunk = makeChunk(r, spec.images)
+        spec.texout.write(spec.texfilter(spec.layout.renderChunk(chunk)))
+    spec.log('')
+    spec.texout.write(b'\\end{document}\n')
 
 
 def main() -> None:
-    args = command_line_parser().parse_args()
-    with requests_cache.CachedSession(args.cache_path, cache_control=True) as session:
-        with open(args.out + '.tex', 'wb') as texout:
-            process(
-                args.url,
-                session,
-                args.timeout,
-                texout,
-                args.pandoc or 'pandoc')
+    with spec_from_commandline_args() as spec:
+        process(spec)
 
 
 if __name__ == '__main__':