pdf2epub/scripts/pdf2html.py

#!/usr/bin/env python3
"""PDF (calibre-generated) -> semantic XHTML, preserving italic/bold/headings."""
import html
import re
import sys
from pathlib import Path

import pymupdf

if sys.argv[1] == "--fonts":  # разведка: какие шрифты/кегли в PDF
    import collections

    c = collections.Counter()
    d = pymupdf.open(sys.argv[2])
    for p in d:
        for b in p.get_text("dict")["blocks"]:
            for l in b.get("lines", []):
                for s in l["spans"]:
                    c[(s["font"], round(s["size"], 1))] += len(s["text"])
    for (f, sz), n in c.most_common(25):
        print("%-32s %5.1f  %8d симв." % (f, sz, n))
    sys.exit()

SRC = Path(sys.argv[1])
OUT = Path(sys.argv[2])
IMGDIR = OUT.parent / "images"
IMGDIR.mkdir(parents=True, exist_ok=True)

# Пороги подобраны под вёрстку 15pt/letter. Для другой книги сначала
# посмотреть реальные шрифты и кегли: pdf2html.py --fonts file.pdf
INDENT_X = 88  # x0 первой строки: больше — абзац с красной строки


def style(font):
    f = font.lower()
    return ("bold" in f or "semibold" in f, "-it" in f or "italic" in f)


def span_html(s):
    t = html.escape(s["text"])
    if not t:
        return ""
    bold, ital = style(s["font"])
    if ital:
        t = "<i>%s</i>" % t
    if bold:
        t = "<b>%s</b>" % t
    return t


def block_kind(b):
    """(tag, css class) from dominant span font/size."""
    spans = [s for l in b["lines"] for s in l["spans"] if s["text"].strip()]
    if not spans:
        return None
    top = max(spans, key=lambda s: len(s["text"]))
    f, sz = top["font"], top["size"]
    if sz >= 28 or f.startswith("Arvo") or f.startswith("CloisterBlack"):
        return ("h1", "title")
    if sz >= 24:
        return ("h1", None)
    if sz >= 17:
        return ("h2", None)
    if f.startswith("MyriadPro"):
        return ("p", "note")  # Maxine's journal / chat / slides
    return ("p", None)


def block_text(b):
    out = []
    for i, l in enumerate(b["lines"]):
        line = "".join(span_html(s) for s in l["spans"])
        if not line:
            continue
        if out:
            prev = out[-1]
            if prev.endswith("-") and not prev.endswith("--"):
                out[-1] = prev[:-1]  # de-hyphenate
            else:
                out.append(" ")
        out.append(line)
    txt = "".join(out)
    txt = re.sub(r"</i>(\s*)<i>", r"\1", txt)
    txt = re.sub(r"</b>(\s*)<b>", r"\1", txt)
    return txt.strip()


doc = pymupdf.open(SRC)
parts = []
open_para = None  # (tag, cls, text) still collecting

# cover
pix = doc[0].get_pixmap(dpi=150)
pix.save(IMGDIR / "cover.jpg")

img_n = 0
for pno, page in enumerate(doc):
    if pno == 0:
        continue
    blocks = sorted(page.get_text("dict")["blocks"], key=lambda b: b["bbox"][1])
    for bi, b in enumerate(blocks):
        if b["type"] == 1:
            img_n += 1
            name = "img%02d.png" % img_n
            (IMGDIR / name).write_bytes(b["image"])
            if open_para:
                parts.append(open_para)
                open_para = None
            parts.append(("figure", None, '<img src="images/%s"/>' % name))
            continue
        kind = block_kind(b)
        if not kind:
            continue
        tag, cls = kind
        txt = block_text(b)
        if not txt:
            continue
        indented = b["lines"][0]["bbox"][0] >= INDENT_X
        cont = (
            open_para
            and tag == "p"
            and open_para[0] == "p"
            and open_para[1] == cls
            and bi == 0
            and not indented
        )
        if cont:
            join = "" if open_para[2].endswith("-") else " "
            open_para = (tag, cls, open_para[2] + join + txt)
            continue
        if open_para:
            parts.append(open_para)
        open_para = (tag, cls, txt)
if open_para:
    parts.append(open_para)

CSS = """
body { margin: 0 1em; }
h1 { text-align: center; margin: 2em 0 0.2em; page-break-before: always; }
h1.title { page-break-before: avoid; }
h2 { text-align: center; font-style: italic; font-weight: normal;
     font-size: 1.1em; margin: 0.2em 0 1.5em; }
p { text-indent: 1.2em; margin: 0; text-align: justify; }
p.note { text-indent: 0; margin: 1em 2em; font-size: 0.9em;
         font-family: sans-serif; }
p.figure { text-indent: 0; text-align: center; margin: 1em 0; }
img { max-width: 100%; }
"""

buf = ['<?xml version="1.0" encoding="utf-8"?>',
       '<html xmlns="http://www.w3.org/1999/xhtml"><head>',
       "<title>%s</title>" % html.escape(doc.metadata.get("title") or SRC.stem),
       "<style>%s</style></head><body>" % CSS]
for tag, cls, txt in parts:
    if tag == "figure":
        buf.append('<p class="figure">%s</p>' % txt)
    else:
        c = ' class="%s"' % cls if cls else ""
        buf.append("<%s%s>%s</%s>" % (tag, c, txt, tag))
buf.append("</body></html>")
doc_html = "\n".join(buf)
# merge tag runs split at page/line boundaries, collapse doubled spaces
doc_html = re.sub(r"</(i|b)>(\s*)<\1>", r"\2", doc_html)
doc_html = re.sub(r"(?<=\S) {2,}(?=\S)", " ", doc_html)
OUT.write_text(doc_html, encoding="utf-8")

print("blocks:", len(parts), "images:", img_n)
print("h1:", sum(1 for p in parts if p[0] == "h1"),
      "p:", sum(1 for p in parts if p[0] == "p"))