--- /dev/null
+# EduMed auto-tagger
+# -*- coding: utf-8 -*-
+import re
+from slughifi import slughifi
+
+
+class Tagger:
+ def __init__(self, state, lines):
+ self.state = state
+ self.lines = lines
+
+ def spawn(self, cls):
+ return cls(self.state, self.lines)
+
+ def line(self, position):
+ return self.lines[position]
+
+ empty_line = re.compile(r"\s+")
+
+ def skip_empty(self, position):
+ while self.line(position) == "" or \
+ self.empty_line.match(self.line(position)):
+ position += 1
+ return position
+
+ def tag(self, position):
+ """
+Return None -- means that we can't tag it in any way
+ """
+ return None
+
+ def wrap(self, tagname, content):
+ return u"<%s>%s</%s>" % (tagname, content, tagname)
+
+
+class Section(Tagger):
+ looks_like = re.compile(r"^[IVX]+[.]\s+(.*)$")
+
+ def tag(self, pos):
+ pos2 = self.skip_empty(pos)
+ pos = pos2
+ m = self.looks_like.match(self.line(pos))
+ if m:
+ self.title = m.groups()[0]
+ return pos + 1
+
+ def __unicode__(self):
+ return self.wrap("naglowek_rozdzial", self.title)
+
+
+class Meta(Tagger):
+ looks_like = re.compile(r"([^:]+): (.*)", re.UNICODE)
+
+ def tag(self, pos):
+ pos = self.skip_empty(pos)
+ m = self.looks_like.match(self.line(pos))
+ if m:
+ k = m.groups()[0]
+ v = m.groups()[1]
+ m = self.state.get('meta', {})
+ m[k] = v
+ self.state['meta'] = m
+ return pos + 1
+
+
+class Informacje(Tagger):
+ def tag(self, pos):
+ self.title = self.spawn(Section)
+ self.meta = []
+ pos = self.title.tag(pos)
+ if pos is None: return
+
+ # collect meta
+ while True:
+ pos = self.skip_empty(pos)
+ meta = self.spawn(Meta)
+ pos2 = meta.tag(pos)
+ if pos2 is None: break
+ self.meta.append(meta)
+ pos = pos2
+
+ return pos
+
+
+class List(Tagger):
+ def tag(self, pos):
+ self.items = []
+ while True:
+ l = self.line(pos)
+ if l and l[0] in ('-', '*'):
+ self.items.append(l[1:].strip())
+ pos += 1
+ else:
+ break
+ if self.items:
+ return pos
+
+ def __unicode__(self):
+ s = "<lista>\n"
+ for i in self.items:
+ s += "<punkt>%s</punkt>\n" % i
+ s += "</lista>\n"
+ return s
+
+
+class Paragraph(Tagger):
+ remove_this = [
+ re.compile(r"[\s]*opis zawartości[\s]*", re.I),
+ re.compile(r"^[\s]*$")
+ ]
+ podrozdzial = [
+ re.compile(r"[\s]*(przebieg zajęć|opcje dodatkowe)[\s]*", re.I),
+ ]
+ def tag(self, pos):
+ self.line = self.lines[pos]
+ self.ignore = False
+ self.is_podrozdzial = False
+
+ for x in self.remove_this:
+ if x.match(self.line):
+ self.ignore = True
+
+ for x in self.podrozdzial:
+ if x.match(self.line):
+ self.is_podrozdzial = True
+
+ return pos + 1
+
+ def __unicode__(self):
+ if not self.ignore:
+ if self.is_podrozdzial:
+ tag = 'naglowek_podrozdzial'
+ else:
+ tag = 'akap'
+ return u"<%s>%s</%s>" % (tag, self.line, tag)
+ else:
+ return u''
+
+
+class Container:
+ def __init__(self, tag_name, *elems):
+ self.tag_name = tag_name
+ self.elems = elems
+
+ def __unicode__(self):
+ s = u"<%s>" % self.tag_name
+ add_nl = False
+ for e in self.elems:
+ if isinstance(e, (str, unicode)):
+ s += unicode(e)
+ else:
+ s += "\n " + unicode(e)
+ add_nl = True
+
+ if add_nl: s += "\n"
+ s += u"</%s>" % self.tag_name
+ return s
+
+
+def eatany(pos, *taggers):
+ try:
+ for t in list(taggers):
+ p = t.tag(pos)
+ if p:
+ return (t, p)
+ except IndexError:
+ pass
+ return (None, pos)
+
+
+def eatseq(pos, *taggers):
+ good = []
+ taggers = list(taggers[:])
+ try:
+ while len(taggers):
+ p = taggers[0].tag(pos)
+ if p is None:
+ return (tuple(good), pos)
+ good.append(taggers.pop(0))
+ # print "%d -> %d" % (pos, p)
+ pos = p
+
+ except IndexError:
+ print "Got index error for pos=%d" % pos
+ return (tuple(good), pos)
+
+
+def tagger(text):
+ """
+tagger(text) function name and signature is a contract.
+returns auto-tagged text
+ """
+ if not isinstance(text, unicode):
+ text = unicode(text.decode('utf-8'))
+ lines = text.split("\n")
+ pos = 0
+ content = []
+ state = {}
+ info = Informacje(state, lines)
+
+ ((info,), pos) = eatseq(pos, info)
+
+ # print "[i] %d. %s" % (pos, lines[pos])
+
+ content.append(info)
+
+ while True:
+ x, pos = eatany(pos, info.spawn(Section),
+ info.spawn(List), info.spawn(Paragraph))
+
+ if x is not None:
+ content.append(x)
+ else:
+ content.append(lines[pos])
+ pos += 1
+ if pos >= len(lines):
+ break
+
+ return toxml(content)
+
+dc_fixed = {
+ 'description': u'Publikacja zrealizowana w ramach projektu Cyfrowa Przyszłość (http://cyfrowaprzyszlosc.pl).',
+ 'relation': u'moduły powiązane linki',
+ 'description.material': u'linki do załączników',
+ 'rights': u'Creative Commons Uznanie autorstwa - Na tych samych warunkach 3.0',
+ }
+
+
+def mark_activities(content):
+ i = 0
+ tl = len(content)
+ is_przebieg = re.compile(r"[\s]*przebieg zaj..[\s]*", re.I)
+ # import pdb; pdb.set_trace()
+ is_next_section = re.compile(r"^[IVX]+[.]? ")
+ is_activity = re.compile(r"^[0-9]+[.]? (.+)")
+
+ is_activity_tools = re.compile(r"^pomoce:[\s]*(.+)")
+ is_activity_work = re.compile(r"^forma pracy:[\s]*(.+)")
+ is_activity_time = re.compile(r"^czas:[\s]*([\d]+).*")
+ activity_props = {
+ 'pomoce': is_activity_tools,
+ 'forma': is_activity_work,
+ 'czas': is_activity_time
+ }
+ activities = []
+
+ in_activities = False
+ ab = -1
+ ae = -1
+ while True:
+ e = content[i]
+ if isinstance(e, Paragraph):
+ if not in_activities and \
+ is_przebieg.match(e.line):
+ in_activities = True
+
+ if in_activities and \
+ is_next_section.match(e.line):
+ in_activities = False
+ if in_activities:
+ m = is_activity.match(e.line)
+ if m:
+ e.line = m.groups()[0]
+ ab = i
+ if is_activity_time.match(e.line):
+ ae = i + 1
+ activities.append((ab, ae))
+ i += 1
+ if i >= tl: break
+
+ activities.reverse()
+ for ab, ae in activities:
+ act_len = ae - ab
+ info_start = ae
+
+ act_els = []
+ act_els.append(Container("opis", content[ab]))
+ for i in range(ab, ae):
+ e = content[i]
+ if isinstance(e, Paragraph):
+ for prop, pattern in activity_props.items():
+ m = pattern.match(e.line)
+ if m:
+ act_els.append(Container(prop, m.groups()[0]))
+ if info_start > i: info_start = i
+ act_els.insert(1, Container('wskazowki',
+ *content[ab + 1:info_start]))
+ content[ab:ae] = [Container('aktywnosc', *act_els)]
+ return content
+
+
+def mark_dictionary(content):
+ db = -1
+ de = -1
+ i = 0
+ is_dictionary = re.compile(r"[\s]*s.owniczek[\s]*", re.I)
+ is_dictentry = re.compile(r"([^-]+) - (.+)")
+ slowniczek = []
+ while i < len(content):
+ e = content[i]
+ if isinstance(e, Section):
+ if is_dictionary.match(e.title):
+ db = i + 1
+ elif db >= 1:
+ de = i
+ content[db:de] = [Container('slowniczek', *slowniczek)]
+ elif db >= 0:
+ if isinstance(e, Paragraph):
+ m = is_dictentry.match(e.line)
+ if m:
+ slowniczek.append(Container('definiendum', m.groups()[0]))
+ slowniczek.append(Container('definiens', m.groups()[1]))
+ else:
+ slowniczek.append(e)
+ i += 1
+
+ return content
+
+
+def toxml(content):
+ content = mark_activities(content)
+ content = mark_dictionary(content)
+ info = content.pop(0)
+
+ state = info.state
+ meta = state['meta']
+ slug = slughifi(meta.get(u'Tytuł modułu', ''))
+ holder = {}
+ holder['xml'] = u""
+
+ def p(t):
+ holder['xml'] += u"%s\n" % t
+
+ def dc(k, v):
+ p(u'<dc:%s xml:lang="pl" xmlns:dc="http://purl.org/dc/elements/1.1/">%s</dc:%s>' % (k, v, k))
+
+ def t(tag, ct):
+ p(u'<%s>%s</%s>' % (tag, ct, tag))
+
+ def a(ct):
+ if ct:
+ t(u'akap', ct)
+
+ p("<utwor>")
+ p(u'<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">')
+ p(u'<rdf:Description rdf:about="http://redakcja.cyfrowaprzyszlosc.pl/documents/">')
+ authors = map(unicode.strip, meta[u'Autorzy'].split(u','))
+ for author in authors:
+ names = author.split(u' ')
+ lastname = names.pop()
+ names.insert(0, lastname + ",")
+ author = u' '.join(names)
+ dc(u'creator', author)
+ dc(u'title', meta.get(u'Tytuł modułu', u''))
+ dc(u'relation.isPartOf', meta.get(u'Dział', u''))
+ dc(u'publisher', u'Fundacja Nowoczesna Polska')
+ dc(u'subject.competence', meta.get(u'Wybrana kompetencja z Katalogu', u''))
+ dc(u'subject.curriculum', meta.get(u'Odniesienie do podstawy programowej', u''))
+ for keyword in meta.get(u'Słowa kluczowe', u'').split(u','):
+ keyword = keyword.strip()
+ dc(u'subject', keyword)
+ dc(u'description', dc_fixed['description'])
+ dc(u'description.material', dc_fixed['description.material'])
+ dc(u'relation', dc_fixed['relation'])
+ dc(u'identifier.url', u'http://cyfrowaprzyszlosc.pl/%s' % slug)
+ dc(u'rights', dc_fixed['rights'])
+ dc(u'rights.license', u'http://creativecommons.org/licenses/by-sa/3.0/')
+ dc(u'format', u'xml')
+ dc(u'type', u'text')
+ dc(u'date', u'2012-11-09') # TODO
+ dc(u'audience', meta.get(u'Poziom edukacyjny', u''))
+ dc(u'language', u'pol')
+ p(u'</rdf:Description>')
+ p(u'</rdf:RDF>')
+
+ p(u'<powiesc>')
+ t(u'nazwa_utworu', meta.get(u'Tytuł modułu', u''))
+ p(u'<nota>')
+ a(u'Numer porządkowy: %s' % meta.get(u'Numer porządkowy', u''))
+ p(u'</nota>')
+
+ p(unicode(info.title))
+ for elm in content:
+ if isinstance(elm, unicode) or isinstance(elm, str):
+ a(elm)
+ continue
+ p(unicode(elm))
+
+ p(u'</powiesc>')
+ p(u'</utwor>')
+
+ return holder['xml']
+
+
+# TODO / TBD
+# ogarnąć podrozdziały
+# Przebieg zajęć
+# opcje dodatkowe
+# usunąć 'opis zawartości'
+# akapit łączony?
fnp / redakcja.git / blobdiff