def __init__(self, name, root=None):
self.name = name
self.logs = {} #maps log-ids (strings) to logs,
#which now contain lists of cells.
#(for fast access)
if root is None:
self.root = ET.Element(ip.notebook,
version=self.version,
nsmap=nsmap(ip, db))
self.head = ET.SubElement(self.root, ip.head)
self.add_log()
else:
#reconstruct object structure from given xml
self.root = root
version = root.get('version')
if version != self.version:
warnings.warn("Wrong version %s != %s" %
(version, self.version))
self.head = root.find(ip.head)
logelems = root.findall(ip.log)
for logelem in logelems:
logid = logelem.get('id')
self.logs[logid] = Log(self, logid, logelem)
self.tree = ET.ElementTree(self.root)
self.xpath = ET.XPathEvaluator(self.tree,
namespaces=nsmap(ip, db, dc, rdf))
def contiguify(self, log, sheets=None):
"""Renumber cells in a log to make them contiguous.
If sheets are provided, renumber the appropriate elements, too.
"""
numbers = sorted(
int(x)
for x in log.xpath('./ip:cell/@number', namespaces=nsmap(ip)))
old2new = {}
for i, oldnum in enumerate(numbers):
newnum = str(i + 1)
old2new[oldnum] = newnum
for elem in log.xpath('.//[@number="%s"]' % oldnum):
elem.set('number', newnum)
if sheets is not None:
logid = log.get('id')
ns = nsmap(ip)
for sheet in sheets:
cells = sheet.xpath('.//ip:block[@logid="%s"]//' % logid,
namespaces=ns)
figs = sheet.xpath('.//ip:figure[@logid="%s"]' % logid,
namespaces=ns)
elems = cells + figs
for x in elem:
x.set('number', old2new.get(x.get('number')))
def vacuum(self, nb):
"""Remove all unreferenced cells from logs.
"""
sheets = nb.xpath.evaluate('./ip:sheet')
keepers = {}
for sheet in sheets:
# get all unique logids from the relevant elements in the sheet
logids = set(sheet.xpath('.//@logid'))
for logid in logids:
# get the log
log = nb.get_log(logid)
# get all ipython-cell elements from ipython-blocks with this logid
cells = sheet.xpath('.//ip:block[@logid="%s"]/ip:cell' % logid,
namespaces=nsmap(ip))
# get all ipython-figure elements with this logid
figs = sheet.xpath('.//ip:figure[@logid="%s"]' % logid)
logkeepers = keepers.get(logid, set())
logkeepers.update(
(x.get('type'), x.get('number')) for x in cells)
logkeepers.update(('figure', x.get('number')) for x in figs)
keepers[logid] = logkeepers
for logid in logids:
log = nb.get_log(logid)
logkeepers = keepers[logid]
cells = list(log)
for cell in cells:
num = cell.get('number')
for subcell in list(cell):
if (subcell.tag, num) not in logkeepers:
cell.remove(subcell)
if len(cell) == 0:
log.remove(cell)
def prep_latex(cls, tree, name=None):
listings = tree.xpath('//db:programlisting', namespaces=nsmap(db))
for listing in listings:
listing.text = cls.escape_latex(listing.text)
for sub in listing.getiterator():
sub.text = cls.escape_latex(sub.text)
sub.tail = cls.escape_latex(sub.tail)
def equation_images(cls, tree, name):
can_tex = False
try:
from matplotlib import texmanager
except ImportError:
warnings.warn('matplotlib not available')
try:
texmgr = texmanager.TexManager()
texmgr.get_dvipng_version()
can_tex = True
except RuntimeError:
warnings.warn('dvipng not up-to-date')
if can_tex == False:
warnings.warn('Cannot convert equations to images')
return
else:
if name is None:
raise ValueError("must provide a name")
eqdir = name + '_files'
if not os.path.isdir(eqdir):
os.mkdir(eqdir)
equations = tree.xpath('//db:informalequation',
namespaces=nsmap(db))
equations.extend(
tree.xpath('//db:inlineequation', namespaces=nsmap(db)))
for eq in equations:
mo = eq.find(db.mediaobject)
if mo is None:
mo = eq.find(db.inlinemediaobject)
phrase = mo.find("%s/%s" % (db.textobject, phrase))
latex = phrase.text
pngfile = texmgr.make_png(latex, dpi=120)
pngbase = os.path.basename(pngfile)
newpng = os.path.join(eqdir, pngbase)
shutil.copyfile(pngfile, newpng)
io = ET.SubElement(mo, db.imageobject)
data = ET.SubElement(
io,
db.imagedata,
fileref=relpath(newpng,
os.path.split(os.path.abspath(name))[0]),
format='PNG')
def __call__(self, nbxml):
r = nbxml.getroot()
if r.get('version') != '1':
raise ValueError('need notebook with version="1"')
newr = ET.Element(ip.notebook, nsmap=nsmap(ip, db), version='1.1')
for elem in r:
newr.append(elem)
for elem in newr.getiterator():
newtag = self.tagmap.get(elem.tag, elem.tag)
elem.tag = newtag
return ET.ElementTree(newr)
def latex_xsl(self):
sheet = ET.Element(xsl.stylesheet, version="1.0", nsmap=nsmap(xsl))
# configure this later
ET.SubElement(sheet, xsl['import'],
href=os.path.join(XSLDIR, "latex", "docbook.xsl"))
# Add template for <code>
t = ET.SubElement(sheet, xsl.template, match="code")
ET.SubElement(t, xsl['call-template'], name="inline.monoseq")
# turn on fancyvrb
fancyvrb = ET.SubElement(sheet, xsl.param, name="latex.use.fancyvrb")
fancyvrb.text = "1"
fvopt = ET.SubElement(sheet, xsl.template, name="latex.fancyvrb.options")
fvopt.text = r",commandchars=\\\{\}"
# turn off fancyhdr
fancyhdr = ET.SubElement(sheet, xsl.param, name='latex.use.fancyhdr')
fancyhdr.text = '0'
docclass = ET.SubElement(sheet, xsl.param,
name="latex.documentclass.article")
docclass.text = "letterpaper,10pt,twoside"
amble = ET.SubElement(sheet, xsl.param,
name="latex.article.preamble.post")
amble.text = '\n'.join(x.as_latex() for x in self.styles_dict.itervalues())
amble.text = '\n%s\n' % amble.text
# and again for book
amble = ET.SubElement(sheet, xsl.param,
name="latex.book.preamble.post")
amble.text = '\n'.join(x.as_latex() for x in self.styles_dict.itervalues())
amble.text = '\n%s\n' % amble.text
# change to English and UTF-8
lang = ET.SubElement(sheet, xsl.variable, name="l10n.gentext.default.language")
lang.text = "en"
ET.SubElement(sheet, xsl.variable, name="latex.documentclass.common")
ET.SubElement(sheet, xsl.variable, name="latex.babel.language")
encoding = ET.SubElement(sheet, xsl.output, method="text",
encoding="UTF-8", indent="yes")
# add the templates for syntax-highlighting
for textstyle in self.styles_dict.itervalues():
template = ET.SubElement(sheet, xsl.template,
match='phrase[@role="%s"]' % textstyle.name,
mode="latex.verbatim")
template.text = '\\%s{' % textstyle.cmd_name()
value = ET.SubElement(template, xsl["value-of"], select="./text()")
value.tail = '}'
return ET.ElementTree(sheet)
def html_xsl(self):
sheet = ET.Element(xsl.stylesheet, version="1.0", nsmap=nsmap(xsl))
ET.SubElement(sheet, xsl['import'],
href=os.path.join(XSLDIR, "xhtml", "docbook.xsl"))
css = ET.SubElement(sheet, xsl.template, name="user.head.content")
css_style = ET.SubElement(css, "style", type="text/css")
comment = ET.SubElement(css_style, xsl.comment)
comment.text = self.html_css()
ET.SubElement(sheet, xsl.param, name="section.autolabel", select="1")
ET.SubElement(sheet, xsl.param,
name="section.label.includes.component.label", select="1")
return ET.ElementTree(sheet)
def get_title(self):
if self.sheet is not None:
#print "sheet in get_title:", self.sheet, ET.tostring(self.sheet)
try:
title = self.sheet.xpath(
'./db:title',
namespaces=nsmap(db))[0] #single title per file
#print ".. title:", title.text
return title.text
except IndexError:
#print "no title"
return None
else:
#print "no sheet in:", self, ET.tostring(self.root)
return None
def rerun_since(self, log, number):
"""Rerun all inputs >= number and append them to the log.
"""
self.add_hooks()
cells = sorted(
(Cell(x) for x in oldlog.xpath('./ip:cell[@number>=%s]' % number,
namespaces=nsmap(ip))),
key=lambda x: x.number)
curnum = cells[-1].number
for cell in cells:
curnum += 1
newcell = ET.SubElement(log, ip.cell, number=str(curnum))
input = ET.SubElement(newcell, ip.input)
input.text = cell.input
retdict = self.run_one(input.text)
if retdict is not None:
for tag, value in retdict.iteritems():
elem = ET.SubElement(newcell, ip[tag])
elem.text = value
self.rm_hooks()
def book2docbook(bookfilename):
from notabene.docbook import DBFormatter
book = ET.parse(bookfilename)
basedir, basename = os.path.split(bookfilename)
child2parent = dict((c, p) for p in book.getiterator() for c in p)
for link in book.xpath('//*[@xlink:type="simple"]',
namespaces=nsmap(xlink)):
parent = child2parent[link]
path, frag = spliturl(basedir, link.get(xlink.href))
nb = Notebook.from_file(path)
sheet = nb.xpath.evaluate('/notebook/sheet[@id="%s"]' % frag)
if sheet == []:
raise ValueError('could not find <sheet id="%s"> in %s' %
(frag, path))
else:
sheet = sheet[0]
dbf = DBFormatter(nb)
dbxml = dbf.transform_sheet(sheet, nodetype=link.tag)
idx = parent.index(link)
del parent[idx]
parent.insert(idx, dbxml)
return book
def transform_elements(elemtype, func):
elems = sheet2.xpath('.//ip:%s' % elemtype, namespaces=nsmap(ip))
for elem in elems:
parent, idx = parent_and_index(elem)
transformed = func(elem)
parent[idx] = transformed
def fo_xsl(self):
sheet = ET.Element(xsl.stylesheet, version="1.0", nsmap=nsmap(xsl))
ET.SubElement(sheet, xsl['import'],
href=os.path.join(XSLDIR, "fo", "docbook.xsl"))
return ET.ElementTree(sheet)
def get_contents(self):
if self.sheet is not None:
#print "sheet in get_contents:", self.sheet, ET.tostring(self.sheet)
sections = self.sheet.xpath('./db:section', namespaces=nsmap(db))
#print "sections:", sections
return sections