def main():
input_assembled_file = Path(sys.argv[1]).resolve(strict=True)
uuid_to_revised_path = Path(sys.argv[2]).resolve(strict=True)
output_file_path = sys.argv[3]
with open(uuid_to_revised_path, 'r') as f:
uuid_to_revised_map = json.load(f)
json_data = {}
with open(input_assembled_file, "r") as in_file:
binder = reconstitute(in_file)
for doc in flatten_to_documents(binder):
abstract = doc.metadata.get("summary")
# Use the map revised value if available, otherwise expect it from the
# metadata parsed from the assembled XHTML
revised = uuid_to_revised_map.get(doc.id) or doc.metadata["revised"]
json_data[doc.ident_hash] = {
"abstract": abstract,
"revised": utils.ensure_isoformat(revised)
}
with open(output_file_path, "w") as out_file:
json.dump(json_data, out_file)
def parse_collection_binders(input_dir):
"""Create a list of binders from book collections"""
baked_collections = Path(input_dir).glob("*.baked.xhtml")
binders = []
for baked_collection in baked_collections:
with open(baked_collection, "r") as baked_file:
binder = reconstitute(baked_file)
binders.append(binder)
return binders
def main(argv=None):
"""Parse passed in cooked single HTML."""
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument('collated_html',
type=argparse.FileType('r'),
help='Path to the collated html'
' file (use - for stdin)')
parser.add_argument('-d',
'--dump-tree',
action='store_true',
help='Print out parsed model tree.')
parser.add_argument('-o',
'--output',
type=argparse.FileType('w+'),
help='Write out epub of parsed tree.')
parser.add_argument('-i',
'--input',
type=argparse.FileType('r'),
help='Read and copy resources/ for output epub.')
args = parser.parse_args(argv)
if args.input and args.output == sys.stdout:
raise ValueError('Cannot output to stdout if reading resources')
from cnxepub.collation import reconstitute
binder = reconstitute(args.collated_html)
if args.dump_tree:
print(pformat(cnxepub.model_to_tree(binder)), file=sys.stdout)
if args.output:
cnxepub.adapters.make_epub(binder, args.output)
if args.input:
args.output.seek(0)
zout = ZipFile(args.output, 'a', ZIP_DEFLATED)
zin = ZipFile(args.input, 'r')
for res in zin.namelist():
if res.startswith('resources'):
zres = zin.open(res)
zi = zin.getinfo(res)
zout.writestr(zi, zres.read(), ZIP_DEFLATED)
zout.close()
# TODO Check for documents that have no identifier.
# These should likely be composite-documents
# or the the metadata got wiped out.
# docs = [x for x in cnxepub.flatten_to(binder, only_documents_filter)
# if x.ident_hash is None]
return 0
def main():
"""Main function"""
in_dir = Path(sys.argv[1]).resolve(strict=True)
out_dir = (in_dir / "disassembled").resolve(strict=True)
baked_file = (in_dir / "collection.baked.xhtml").resolve(strict=True)
baked_metdata_file = (in_dir / "collection.baked-metadata.json").resolve(
strict=True)
with open(baked_file, "rb") as file:
html_root = etree.parse(file)
binder = reconstitute(file)
# It's important that we generate slug metadata in parallel with disassemble
# so that where ident_hash values are based upon potentially randomly
# generated UUIDs we can still use them as unique keys in JSON outputs
# without diverging
slugs = extract_slugs_from_binder(binder)
nav = html_root.xpath("//xhtml:nav",
namespaces=HTML_DOCUMENT_NAMESPACES)[0]
toc_maker = ElementMaker(namespace=None,
nsmap={None: "http://www.w3.org/1999/xhtml"})
toc = toc_maker.html(E.head(E.title("Table of Contents")), E.body(nav))
with open(f"{out_dir}/collection.toc.xhtml", "wb") as out:
out.write(etree.tostring(toc, encoding="utf8", pretty_print=True))
with open(baked_metdata_file, "r") as baked_json:
baked_metadata = json.load(baked_json)
for doc in flatten_to(binder, lambda d: isinstance(d, Document)):
with open(f"{out_dir / doc.ident_hash}.xhtml", "wb") as out:
out.write(bytes(DocumentContentFormatter(doc)))
with open(f"{out_dir / doc.ident_hash}-metadata.json",
"w") as json_out:
# Incorporate metadata from disassemble step while setting defaults
# for cases like composite pages which may not have metadata from
# previous stages
json_metadata = {
"slug": slugs.get(doc.ident_hash),
"title": doc.metadata.get("title"),
"abstract": None
}
# Add / override metadata from baking if available
json_metadata.update(baked_metadata.get(doc.ident_hash, {}))
json.dump(json_metadata, json_out)
def main(argv=None):
"""Parse passed in cooked single HTML."""
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument('collated_html', type=argparse.FileType('r'),
help='Path to the collated html'
' file (use - for stdin)')
parser.add_argument('-d', '--dump-tree', action='store_true',
help='Print out parsed model tree.')
parser.add_argument('-o', '--output', type=argparse.FileType('w+'),
help='Write out epub of parsed tree.')
parser.add_argument('-i', '--input', type=argparse.FileType('r'),
help='Read and copy resources/ for output epub.')
args = parser.parse_args(argv)
if args.input and args.output == sys.stdout:
raise ValueError('Cannot output to stdout if reading resources')
from cnxepub.collation import reconstitute
binder = reconstitute(args.collated_html)
if args.dump_tree:
print(pformat(cnxepub.model_to_tree(binder)),
file=sys.stdout)
if args.output:
cnxepub.adapters.make_epub(binder, args.output)
if args.input:
args.output.seek(0)
zout = ZipFile(args.output, 'a', ZIP_DEFLATED)
zin = ZipFile(args.input, 'r')
for res in zin.namelist():
if res.startswith('resources'):
zres = zin.open(res)
zi = zin.getinfo(res)
zout.writestr(zi, zres.read(), ZIP_DEFLATED)
zout.close()
# TODO Check for documents that have no identifier.
# These should likely be composite-documents
# or the the metadata got wiped out.
# docs = [x for x in cnxepub.flatten_to(binder, only_documents_filter)
# if x.ident_hash is None]
return 0
def main():
"""Main function"""
xhtml_file = Path(sys.argv[1]).resolve(strict=True)
metadata_file = Path(sys.argv[2]).resolve(strict=True)
book_slug = sys.argv[3]
out_dir = Path(sys.argv[4])
with open(xhtml_file, "rb") as file:
html_root = etree.parse(file)
binder = reconstitute(file)
slugs = extract_slugs_from_binder(binder)
with open(metadata_file, "r") as baked_json:
baked_metadata = json.load(baked_json)
book_toc_metadata = baked_metadata.get(binder.ident_hash)
nav = html_root.xpath("//xhtml:nav",
namespaces=HTML_DOCUMENT_NAMESPACES)[0]
toc_maker = ElementMaker(namespace=None,
nsmap={None: "http://www.w3.org/1999/xhtml"})
toc = toc_maker.html(E.head(E.title("Table of Contents")), E.body(nav))
nav_links = toc.xpath("//xhtml:a", namespaces=HTML_DOCUMENT_NAMESPACES)
for doc in flatten_to_documents(binder):
id_with_context = f'{binder.ident_hash}:{doc.id}'
module_etree = content_to_etree(doc.content)
for link in nav_links:
link_href = link.attrib['href']
if not link_href.startswith('#'):
continue
if module_etree.xpath(
f"/xhtml:body/xhtml:div[@id='{link_href[1:]}']",
namespaces=HTML_DOCUMENT_NAMESPACES):
link.attrib['href'] = f'./{id_with_context}.xhtml'
# Add metadata to same-book-different-module links.
# The module in which same-book link targets reside is only fully known
# at time of disassembly. Different pipelines can make use of this
# metadata in different ways
for node in module_etree.xpath(
'//xhtml:a[@href and starts-with(@href, "/contents/")]',
namespaces=HTML_DOCUMENT_NAMESPACES):
print('BEFORE:')
print(node.attrib)
page_link = node.attrib["href"].split("/")[-1]
# Link may have fragment
if "#" in page_link:
page_uuid, page_fragment = page_link.split("#")
else:
page_uuid = page_link
page_fragment = ''
# This is either an intra-book link or inter-book link. We can
# differentiate the latter by data-book-uuid attrib).
if not node.attrib.get("data-book-uuid"):
node.attrib["data-page-slug"] = slugs.get(page_uuid)
node.attrib["data-page-uuid"] = page_uuid
node.attrib["data-page-fragment"] = page_fragment
print('AFTER:')
print(node.attrib)
doc.content = etree_to_content(module_etree)
# Inject some styling and JS for QA
xml_parser = etree.XMLParser(ns_clean=True)
root = etree.XML(bytes(DocumentContentFormatter(doc)), xml_parser)
head = root.xpath("//xhtml:head", namespaces=HTML_DOCUMENT_NAMESPACES)
if not head:
head = etree.Element("head")
root.insert(0, head)
style = etree.Element("style")
script = etree.Element("script")
style.text = u'''
/* STYLING_FOR_DEVS */
/* Linking to a specific element should highlight the element */
:target {
background-color: #ffffcc;
border: 1px dotted #000000;
animation-name: cssAnimation;
animation-duration: 10s;
animation-timing-function: ease-out;
animation-delay: 0s;
animation-fill-mode: forwards;
}
@keyframes cssAnimation {
to {
background-color: initial;
border: initial;
}
}
/* Style footnotes so that they stand out */
[role="doc-footnote"] {
background-color: #ffcccc;
border: 1px dashed #ff0000;
}
[role="doc-footnote"]:before { content: "FOOTNOTE " ; }
/* Show a permalink when hovering over a heading or paragraph */
*:not(:hover) > a.-dev-permalinker { display: none; }
* > a.-dev-permalinker {
margin-left: .1rem;
text-decoration: none;
}
'''
script.text = u'''//<![CDATA[
// SCRIPTS_FOR_DEVS
window.addEventListener('load', () => {
const pilcrow = '¶'
function addPermalink(parent, id) {
const link = window.document.createElement('a')
link.classList.add('-dev-permalinker')
link.setAttribute('href', '#' + id)
link.textContent = pilcrow
parent.appendChild(link)
}
const paragraphs = Array.from(
document.querySelectorAll('p[id]')
)
paragraphs.forEach(p => addPermalink(p, p.getAttribute('id')) )
const headings = Array.from(
document.querySelectorAll(
'*[id] > h1, *[id] > h2, *[id] > h3, ' +
'*[id] > h4, *[id] > h5, *[id] > h6'
)
)
headings.forEach(h => addPermalink(
h, h.parentElement.getAttribute('id'))
)
})
// ]]>'''
head.append(style)
head.append(script)
with open(f"{out_dir / id_with_context}.xhtml", "wb") as out:
out.write(etree.tostring(root))
with open(f"{out_dir / id_with_context}-metadata.json",
"w") as json_out:
# Incorporate metadata from disassemble step while setting defaults
# for cases like composite pages which may not have metadata from
# previous stages
json_metadata = {
"slug": slugs.get(doc.id),
"title": doc.metadata.get("title"),
"abstract": None,
"id": doc.id,
"revised": datetime.now(timezone.utc).isoformat()
}
# Add / override metadata from baking if available
json_metadata.update(baked_metadata.get(doc.ident_hash, {}))
json.dump(json_metadata, json_out)
with open(f"{out_dir}/{book_slug}.toc.xhtml", "wb") as out:
out.write(etree.tostring(toc, encoding="utf8", pretty_print=True))
with open(f"{out_dir}/{book_slug}.toc-metadata.json", "w") as toc_json:
json.dump(book_toc_metadata, toc_json)
import sys
import json
from cnxepub.collation import reconstitute
from cnxepub.models import flatten_to_documents
in_path, out_path = sys.argv[1:3]
json_data = {}
with open(in_path, "r") as in_file:
binder = reconstitute(in_file)
for doc in flatten_to_documents(binder):
abstract = doc.metadata.get("summary")
json_data[doc.ident_hash] = {"abstract": abstract}
with open(out_path, "w") as out_file:
json.dump(json_data, out_file)
def test_html(self):
page_path = os.path.join(TEST_DATA_DIR, 'desserts-single-page.html')
with open(page_path) as html:
from cnxepub.collation import reconstitute
desserts = reconstitute(html)
self.check_desserts(desserts)
def test_html(self):
page_path = os.path.join(TEST_DATA_DIR, 'desserts-single-page.html')
with open(page_path) as html:
from cnxepub.collation import reconstitute
desserts = reconstitute(html)
self.check_desserts(desserts)
