More structure and tests around splitting the page into chunks' DOMs.

[paperdoorknob] / paperdoorknob.py

diff --git a/paperdoorknob.py b/paperdoorknob.py
index 3b86cb8da08144c3862ab80b757b0194c086dbad..84ed70228d3d9b4040fc765c3ced2be68d1d58ea 100644 (file)
--- a/paperdoorknob.py
+++ b/paperdoorknob.py
@@ -5,87 +5,34 @@
 # Free Software Foundation, version 3.
 
 
-from argparse import ArgumentParser
-import itertools
-import os.path
-
-from typing import IO, Iterable
-
 from bs4 import BeautifulSoup
-from bs4.element import Tag
-from xdg_base_dirs import xdg_cache_home
-
-from fetch import CachingFetcher, Fetcher
-from texify import PandocTexifier, Texifier
 
-
-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
+from args import spec_from_commandline_args
+from glowfic import chunkDOMs
+from spec import Spec
 
 
 def parse(content: bytes) -> BeautifulSoup:
     return BeautifulSoup(content, '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 process(
-        url: str,
-        fetcher: Fetcher,
-        texifier: Texifier,
-        texout: IO[bytes]) -> None:
-    texout.write(b'\\documentclass{article}\n\\begin{document}\n')
-    html = clean(parse(fetcher.fetch(url)))
-    for r in replies(html):
-        texout.write(texifier.texify(r))
-    texout.write(b'\\end{document}\n')
+def process(spec: Spec) -> None:
+    spec.texout.write(b'\\documentclass{article}\n')
+    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}\n')
+    html = parse(spec.htmlfilter(spec.fetcher.fetch(spec.url)))
+    for r in chunkDOMs(html):
+        spec.domfilter(r)
+        spec.texout.write(spec.texifier.texify(r))
+    spec.texout.write(b'\\end{document}\n')
 
 
 def main() -> None:
-    args = command_line_parser().parse_args()
-    texifier = PandocTexifier(args.pandoc or 'pandoc')
-    with CachingFetcher(args.cache_path, args.timeout) as fetcher:
-        with open(args.out + '.tex', 'wb') as texout:
-            process(args.url, fetcher, texifier, texout)
+    with spec_from_commandline_args() as spec:
+        process(spec)
 
 
 if __name__ == '__main__':