Rename html → dom

[paperdoorknob] / paperdoorknob.py

diff --git a/paperdoorknob.py b/paperdoorknob.py
index 31ce02c494ea29aadc8794d75f11ee8397729a2c..ade0b885d42addb973f415599798bd2e153ac961 100644 (file)
--- a/paperdoorknob.py
+++ b/paperdoorknob.py
@@ -4,94 +4,76 @@
 # under the terms of the GNU General Public License as published by the
 # Free Software Foundation, version 3.
 
 # 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
 
 from bs4 import BeautifulSoup
 from bs4.element import Tag

-from xdg_base_dirs import xdg_cache_home
-
-from fetch import CachingFetcher, Fetcher

 
 

-
-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 flatURL, makeChunk, renderChunk, Thread
+from spec import Spec

 
 
 def parse(content: bytes) -> BeautifulSoup:
     return BeautifulSoup(content, 'html.parser')
 
 
 
 
 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 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,
-        fetcher: Fetcher,
-        texout: IO[bytes],
-        pandoc: str) -> None:
-    texout.write(b'\\documentclass{article}\n\\begin{document}\n')
-    html = clean(parse(fetcher.fetch(url)))
-    for r in replies(html):
-        texout.write(html_to_tex(pandoc, r))
-    texout.write(b'\\end{document}\n')
+def ilen(it: Iterable[Any]) -> int:
+    return sum(1 for _ in it)
+
+
+def get_title(dom: BeautifulSoup) -> str | None:
+    span = dom.findChild("span", id="post-title")
+    if not isinstance(span, Tag):
+        return None
+    return span.text
+
+
+def process(spec: Spec) -> None:
+    spec.texout.write(br'''\documentclass{article}
+\usepackage{booktabs}
+\usepackage{graphicx}
+\usepackage{longtable}
+\usepackage{soul}
+\usepackage{varwidth}
+\usepackage{wrapstuff}
+''')
+    if spec.geometry is not None:
+        spec.texout.write(br'\usepackage[' +
+                          spec.geometry.encode('UTF-8') +
+                          b']{geometry}\n')
+    spec.texout.write(br'''\begin{document}
+\newcommand{\href}[2]{#2\footnote{\detokenize{#1}}}
+\def\glowiconsize{%fmm}
+\newcommand{\glowicon}[1]{\includegraphics[
+  width=\glowiconsize,height=\glowiconsize,keepaspectratio
+]{#1}}
+\newcommand{\ifnotempty}[2]{\expandafter\ifx\expandafter\relax
+  \detokenize{#1}\relax\else #2\fi}
+%s
+''' % (spec.icon_size, spec.layout))
+    url = flatURL(spec.url)
+    spec.log('Fetching HTML...\r')
+    html = spec.fetcher.fetch(url)
+    spec.log('Parsing HTML...\r')
+    dom = parse(spec.htmlfilter(html))
+    thread = Thread(dom)
+    spec.log('Counting chunks...\r')
+    num_chunks = ilen(thread.chunkDOMs())
+    title = get_title(dom) or "chunk"
+    for i, r in enumerate(thread.chunkDOMs()):
+        percent = 100.0 * i / num_chunks
+        spec.log(f'Processing {title} {i} of {num_chunks} ({percent:.1f}%)\r')
+        spec.domfilter(r)
+        chunk = makeChunk(r, spec.images)
+        spec.texout.write(spec.texfilter(renderChunk(spec.texifier, chunk)))
+    spec.log('')
+    spec.texout.write(b'\\end{document}\n')

 
 
 def main() -> None:
 
 
 def main() -> None:

-    args = command_line_parser().parse_args()
-    with CachingFetcher(args.cache_path, args.timeout) as fetcher:
-        with open(args.out + '.tex', 'wb') as texout:
-            process(args.url, fetcher, texout, args.pandoc or 'pandoc')
+    with spec_from_commandline_args() as spec:
+        process(spec)

 
 
 if __name__ == '__main__':
 
 
 if __name__ == '__main__':