diora-web/e2e/fixtures/make_test_pdf.py

"""A small PDF with the cases the PDF reader has to get right (e2e and by hand).

  python e2e/fixtures/make_test_pdf.py e2e/fixtures/test.pdf

12 pages: A4 portrait text pages, one A5 page and one A4 landscape page (mixed
sizes), page 7 nearly empty (only a page number: the smart crop must not zoom
it to infinity), an outline (table of contents) with three entries, internal
links from page 1 to pages 5 and 9 and one external link. Plain PDF 1.4,
Helvetica, no dependencies.
"""
import sys

A4 = (595, 842)
A5 = (420, 595)
A4L = (842, 595)

WORDS = ('Die Seite trägt Text, damit Auswahl, Suche und Vorlesen etwas finden. '
         'Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor. ').split()


def esc(t):
    return t.replace('\\', '\\\\').replace('(', '\\(').replace(')', '\\)')


def text_lines(n, words_per_line=11):
    out, i = [], 0
    for _ in range(n):
        out.append(' '.join(WORDS[(i + k) % len(WORDS)] for k in range(words_per_line)))
        i += words_per_line
    return out


def page_content(no, size):
    w, h = size
    ops = ['BT', '/F1 22 Tf', f'50 {h - 70} Td', f'(Kapitel {no} - Seite {no}) Tj', 'ET']
    if no == 7:
        # Nearly empty: just the number at the bottom.
        return '\n'.join(['BT', '/F1 10 Tf', f'{w / 2 - 10:.0f} 30 Td', f'({no}) Tj', 'ET'])
    y = h - 110
    lines = text_lines(28 if h > 700 else 18, 11 if w > 500 else 8)
    ops += ['BT', '/F1 11 Tf', f'50 {y} Td', '14 TL']
    for ln in lines:
        ops.append(f'({esc(ln)}) Tj T*')
    ops.append('ET')
    if no == 1:
        ops += ['BT', '/F1 12 Tf', f'50 {y - 30 - 14 * len(lines)} Td',
                '(Link zu Seite 5) Tj', '0 -20 Td', '(Link zu Seite 9) Tj', '0 -20 Td', '(Externer Link: example.org) Tj', 'ET']
    ops += ['BT', '/F1 10 Tf', f'{w / 2 - 10:.0f} 30 Td', f'({no}) Tj', 'ET']
    return '\n'.join(ops)


def build(path):
    sizes = [A4] * 12
    sizes[3] = A5          # page 4
    sizes[8] = A4L         # page 9
    objs = []              # index = object number - 1

    def add(body):
        objs.append(body)
        return len(objs)

    catalog = add(None)
    pages_obj = add(None)
    font = add('<< /Type /Font /Subtype /Type1 /BaseFont /Helvetica /Encoding /WinAnsiEncoding >>')
    page_ids = []
    for i, size in enumerate(sizes):
        content = page_content(i + 1, size).encode('cp1252')
        cid = add(b'<< /Length %d >>\nstream\n' % len(content) + content + b'\nendstream')
        pid = add(None)
        page_ids.append((pid, cid, size))
    # links on page 1 (placed where the text is: bottom part of the text block)
    p1_h = sizes[0][1]
    base_y = p1_h - 110 - 30 - 14 * 28

    def link_to(page_no, y):
        return add(f'<< /Type /Annot /Subtype /Link /Rect [48 {y - 4} 220 {y + 14}] /Border [0 0 0] '
                   f'/Dest [{page_ids[page_no - 1][0]} 0 R /XYZ 0 {sizes[page_no - 1][1]} null] >>')
    annots = [link_to(5, base_y), link_to(9, base_y - 20),
              add(f'<< /Type /Annot /Subtype /Link /Rect [48 {base_y - 44} 300 {base_y - 26}] /Border [0 0 0] '
                  f'/A << /S /URI /URI (https://example.org/) >> >>')]
    for n, (pid, cid, (w, h)) in enumerate(page_ids):
        extra = f' /Annots [{" ".join(f"{a} 0 R" for a in annots)}]' if n == 0 else ''
        objs[pid - 1] = (f'<< /Type /Page /Parent {pages_obj} 0 R /MediaBox [0 0 {w} {h}] '
                         f'/Resources << /Font << /F1 {font} 0 R >> >> /Contents {cid} 0 R{extra} >>')
    objs[pages_obj - 1] = f'<< /Type /Pages /Kids [{" ".join(f"{p} 0 R" for p, _, _ in page_ids)}] /Count {len(page_ids)} >>'
    # outline: three entries
    outline = add(None)
    entries = [('Anfang', 1), ('Querformat', 9), ('Fast leer', 7)]
    ids = [add(None) for _ in entries]
    for k, ((title, page), oid) in enumerate(zip(entries, ids)):
        nav = ''
        if k > 0:
            nav += f' /Prev {ids[k - 1]} 0 R'
        if k < len(ids) - 1:
            nav += f' /Next {ids[k + 1]} 0 R'
        objs[oid - 1] = (f'<< /Title ({esc(title)}) /Parent {outline} 0 R{nav} '
                         f'/Dest [{page_ids[page - 1][0]} 0 R /XYZ 0 {sizes[page - 1][1]} null] >>')
    objs[outline - 1] = f'<< /Type /Outlines /First {ids[0]} 0 R /Last {ids[-1]} 0 R /Count {len(ids)} >>'
    objs[catalog - 1] = f'<< /Type /Catalog /Pages {pages_obj} 0 R /Outlines {outline} 0 R /PageMode /UseOutlines >>'
    info = add('<< /Title (Testdokument PDF) /Author (diora Test) >>')

    out = bytearray(b'%PDF-1.4\n%\xe2\xe3\xcf\xd3\n')
    offsets = []
    for i, body in enumerate(objs):
        offsets.append(len(out))
        data = body if isinstance(body, bytes) else body.encode('cp1252')
        out += b'%d 0 obj\n' % (i + 1) + data + b'\nendobj\n'
    xref = len(out)
    out += b'xref\n0 %d\n0000000000 65535 f \n' % (len(objs) + 1)
    for off in offsets:
        out += b'%010d 00000 n \n' % off
    out += (b'trailer\n<< /Size %d /Root %d 0 R /Info %d 0 R >>\nstartxref\n%d\n%%%%EOF\n'
            % (len(objs) + 1, catalog, info, xref))
    open(path, 'wb').write(out)


if __name__ == '__main__':
    build(sys.argv[1] if len(sys.argv) > 1 else 'test.pdf')