def render_group_webpage(args):
    data = None
    info = {}
    if 'label' in args:
        label = clean_input(args['label'])
        label = label.replace('t', 'T')
        C = base.getDBConnection()
        data = C.transitivegroups.groups.find_one({'label': label})
        if data is None:
            bread = get_bread([("Search error", url_for('.search'))])
            info['err'] = "Group " + label + " was not found in the database."
            info['label'] = label
            return search_input_error(info, bread)
        title = 'Galois Group:' + label
        n = data['n']
        t = data['t']
        data['yesno'] = yesno
        order = data['order']
        data['orderfac'] = latex(ZZ(order).factor())
        orderfac = latex(ZZ(order).factor())
        data['ordermsg'] = "$%s=%s$" % (order, latex(orderfac))
        if ZZ(order) == 1:
            data['ordermsg'] = "$1$"
        if ZZ(order).is_prime():
            data['ordermsg'] = "$%s$ (is prime)" % order
        pgroup = len(ZZ(order).prime_factors()) < 2
        if n == 1:
            G = gap.SmallGroup(n, t)
        else:
            G = gap.TransitiveGroup(n, t)
        if ZZ(order) < ZZ('10000000000'):
            ctable = chartable(n, t)
        else:
            ctable = 'Group too large'
        data['gens'] = generators(n, t)
        if n == 1 and t == 1:
            data['gens'] = 'None needed'
        data['chartable'] = ctable
        data['parity'] = "$%s$" % data['parity']
        data['cclasses'] = conjclasses(G, n)
        data['subinfo'] = subfield_display(C, n, data['subs'])
        data['resolve'] = resolve_display(C, data['resolve'])
#    if len(data['resolve']) == 0: data['resolve'] = 'None'
        data['otherreps'] = otherrep_display(n, t, C, data['repns'])
        prop2 = [
            ('Order:', '\(%s\)' % order),
            ('n:', '\(%s\)' % data['n']),
            ('Cyclic:', yesno(data['cyc'])),
            ('Abelian:', yesno(data['ab'])),
            ('Solvable:', yesno(data['solv'])),
            ('Primitive:', yesno(data['prim'])),
            ('$p$-group:', yesno(pgroup)),
            ('Name:', group_display_short(n, t, C)),
        ]
        info.update(data)

        bread = get_bread([(label, ' ')])
        return render_template("gg-show-group.html", credit=GG_credit, title=title, bread=bread, info=info, properties2=prop2)

def splitint(a,p):
    if a==1:
        return ' '
    j = valuation(a,p)
    if j==0:
        return str(a)
    a = a/p**j
    if a==1:
        return latex(ZZ(p**j).factor())
    return str(a)+r'\cdot'+latex(ZZ(p**j).factor())

 def safe_reduce(f):
     if not m:
         return latex(f)
     try:
         if f in sam.field():
             return latex(redc(f))
         else:
             return latex(redp(f))
     except ZeroDivisionError:
         return '\\textrm{Unable to reduce} \\bmod\\mathfrak{m}'

def make_map_latex(map_str):
    # FIXME: Get rid of nu when map is defined over QQ
    if "nu" not in map_str:
        R0 = QQ
    else:
        R0 = PolynomialRing(QQ,'nu')
    R = PolynomialRing(R0,2,'x,y')
    F = FractionField(R)
    phi = F(map_str)
    num = phi.numerator()
    den = phi.denominator()
    c_num = num.denominator()
    c_den = den.denominator()
    lc = c_den/c_num
    # rescale coeffs to make them integral. then try to factor out gcds
    # numerator
    num_new = c_num*num
    num_cs = num_new.coefficients()
    if R0 == QQ:
        num_cs_ZZ = num_cs
    else:
        num_cs_ZZ = []
        for el in num_cs:
            num_cs_ZZ = num_cs_ZZ + el.coefficients()
    num_gcd = gcd(num_cs_ZZ)
    # denominator
    den_new = c_den*den
    den_cs = den_new.coefficients()
    if R0 == QQ:
        den_cs_ZZ = den_cs
    else:
        den_cs_ZZ = []
        for el in den_cs:
            den_cs_ZZ = den_cs_ZZ + el.coefficients()
    den_gcd = gcd(den_cs_ZZ)
    lc = lc*(num_gcd/den_gcd)
    num_new = num_new/num_gcd
    den_new = den_new/den_gcd
    # make strings for lc, num, and den
    num_str = latex(num_new)
    den_str = latex(den_new)
    if lc==1:
        lc_str=""
    else:
        lc_str = latex(lc)
    if den_new==1:
        if lc ==1:
            phi_str = num_str
        else:
            phi_str = lc_str+"("+num_str+")"
    else:
        phi_str = lc_str+"\\frac{"+num_str+"}"+"{"+den_str+"}"
    return phi_str

def make_curve_latex(crv_str):
    # FIXME: Get rid of nu when map is defined over QQ
    if "nu" not in crv_str:
        R0 = QQ
    else:
        R0 = PolynomialRing(QQ,'nu')
    R = PolynomialRing(R0,2,'x,y')
    F = FractionField(R)
    sides = crv_str.split("=")
    lhs = latex(F(sides[0]))
    rhs = latex(F(sides[1]))
    eqn_str = lhs + '=' + rhs
    return eqn_str

 def _latex_using_dpd_depth1(self, dpd_dct):
     names = [dpd_dct[c] for c in self._consts]
     _gcd = QQ(gcd(self._coeffs))
     coeffs = [c / _gcd for c in self._coeffs]
     coeffs_names = [(c, n) for c, n in zip(coeffs, names) if c != 0]
     tail_terms = ["%s %s %s" % ("+" if c > 0 else "", c, n) for c, n in coeffs_names[1:]]
     c0, n0 = coeffs_names[0]
     head_term = str(c0) + " " + str(n0)
     return r"\frac{{{pol_num}}}{{{pol_dnm}}} \left({terms}\right)".format(
         pol_dnm=latex(_gcd.denominator() * self._scalar_const._polynomial_expr()),
         pol_num=latex(_gcd.numerator()),
         terms=" ".join([head_term] + tail_terms),
     )

 def disc_factored_latex(self):
     D = self.disc()
     s = ''
     if D < 0:
         D = -D
         s = r'-\,'
     return s + latex(D.factor())

    def latex(self, prec=None, name=None, keepzeta=False):
        """
        Change the name of the variable in a polynomial.  If keepzeta, then don't change
        the name of zetaN in the defining polynomial of a cyclotomic field.
        (keepzeta not implemented yet)
        """
        if prec is None:
            qe = self.value()
        else:
            qe = self.value()
            if not qe is None:
                qe = qe.truncate_powerseries(prec)
        wl = web_latex_split_on_re(qe)
        if name is not None and self.value().base_ring().absolute_degree()>1:
            oldname = latex(self.value().base_ring().gen())
            subfrom = oldname.strip()
            subfrom = subfrom.replace("\\","\\\\")  
            subfrom = subfrom.replace("{","\\{")   # because x_{0} means something in a regular expression
            if subfrom[0].isalpha():
                subfrom = "\\b" + subfrom
            subto = name.replace("\\","\\\\") + " "
            if keepzeta and "zeta" in subfrom:
                pass  # keep the variable as-is
            else:
                wl = re.sub(subfrom, subto, wl)
            return wl

        else:
            return wl

def web_latex_split_on_re(x, r = '(q[^+-]*[+-])'):

    def insert_latex(s):
        return s.group(1) + '\) \('

    if isinstance(x, (str, unicode)):
        return x
    else:
        A = "\( %s \)" % latex(x)
        c = re.compile(r)
        A = A.replace('+', '\) \( {}+ ')
        A = A.replace('-', '\) \( {}- ')
#        A = A.replace('\left(','\left( {}\\right.') # parantheses needs to be balanced
#        A = A.replace('\\right)','\left.\\right)')        
        A = A.replace('\left(','\\bigl(')
        A = A.replace('\\right)','\\bigr)')        
        A = c.sub(insert_latex, A)

    # the above will be re-done using a more sophisticated method involving
    # regular expressions.  Below fixes bad spacing when the current approach
    # encounters terms like (-3+x)
    A = A.replace('( {}','(')
    A = A.replace('(\) \(','(')
    A = A.replace('\(+','\(\mathstrut+')
    A = A.replace('\(-','\(\mathstrut-')
    A = A.replace('(  ','(')
    A = A.replace('( ','(')
    A = A.replace('+\) \(O','+O')
    return A

 def eigs_as_seqseq_to_qexp(self, prec_max):
     # Takes a sequence of sequence of integers (or pairs of integers in the hecke_ring_cyclotomic_generator != 0 case) and returns a string for the corresponding q expansion
     # For example, eigs_as_seqseq_to_qexp([[0,0],[1,3]]) returns "\((1+3\beta_{1})q\)\(+O(q^2)\)"
     prec = min(self.qexp_prec, prec_max)
     if prec == 0:
         return 'O(1)'
     eigseq = self.qexp[:prec]
     use_knowl = too_big(eigseq, 10**24)
     s = ''
     for j in range(len(eigseq)):
         term = self._elt(eigseq[j])
         if term != 0:
             latexterm = latex(term)
             if use_knowl:
                 latexterm = make_bigint(latexterm)
             if term.number_of_terms() > 1:
                 latexterm = r"(" +  latexterm + r")"
             if j > 0:
                 if term == 1:
                     latexterm = ''
                 elif term == -1:
                     latexterm = '-'
                 if j == 1:
                     latexterm += ' q'
                 else:
                     latexterm += ' q^{%d}' % j
             if s != '' and latexterm[0] != '-':
                 latexterm = '+' + latexterm
             s += '\(' + latexterm + '\) '
     # Work around bug in Sage's latex
     s = s.replace('betaq', 'beta q')
     return s + '\(+O(q^{%d})\)' % prec

def web_latex_split_on_pm(x):
    on = ['+', '-']
 #   A = "\( %s \)" % latex(x)
    try:
        A = "\(" + x + "\)"  # assume we are given LaTeX to split on
    except:
        A = "\( %s \)" % latex(x)

       # need a more clever split_on_pm that inserts left and right properly
    A = A.replace("\\left","")
    A = A.replace("\\right","")
    for s in on:
  #      A = A.replace(s, '\) ' + s + ' \( ')
   #     A = A.replace(s, '\) ' + ' \( \mathstrut ' + s )
        A = A.replace(s, '\)' + ' \(\mathstrut ' + s + '\mathstrut ')
    # the above will be re-done using a more sophisticated method involving
    # regular expressions.  Below fixes bad spacing when the current approach
    # encounters terms like (-3+x)
    for s in on:
        A = A.replace('(\) \(\mathstrut '+s,'(' + s)
    A = A.replace('( {}','(')
    A = A.replace('(\) \(','(')
    A = A.replace('\(+','\(\mathstrut+')
    A = A.replace('\(-','\(\mathstrut-')
    A = A.replace('(  ','(')
    A = A.replace('( ','(')

    return A

    def eigs_as_seqseq_to_qexp(self, prec_max):
        # Takes a sequence of sequence of integers and returns a string for the corresponding q expansion
        # For example, eigs_as_seqseq_to_qexp([[0,0],[1,3]]) returns "\((1+3\beta_{1})q\)\(+O(q^2)\)"
        prec = min(self.qexp_prec, prec_max)
        if prec == 0:
            return 'O(1)'
        eigseq = self.qexp[:prec]
        d = self.dim
        Rgens = self._get_Rgens()
        s = ''
        for j in range(len(eigseq)):
            term = sum([Rgens[i]*eigseq[j][i] for i in range(d)])
            if term != 0:
                latexterm = latex(term)
                if term.number_of_terms() > 1:
                    latexterm = r"\left(" +  latexterm + r"\right)"

                if j > 0:
                    if term == 1:
                        latexterm = ''
                    elif term == -1:
                        latexterm = '-'
                    if j == 1:
                        latexterm += ' q'
                    else:
                        latexterm += ' q^{%d}' % j
                #print latexterm
                if s != '' and latexterm[0] != '-':
                    latexterm = '+' + latexterm
                s += '\(' + latexterm + '\) '
        # Work around bug in Sage's latex
        s = s.replace('betaq', 'beta q')
        return s + '\(+O(q^{%d})\)' % prec

def nf_knowl_guts(label):
    out = ''
    wnf = WebNumberField(label)
    if wnf.is_null():
        return 'Cannot find global number field %s' % label
    out += "Global number field %s" % label
    out += '<div>'
    out += 'Defining polynomial: '
    out += "\(%s\)" % latex(wnf.poly())
    D = wnf.disc()
    Dfact = wnf.disc_factored_latex()
    if D.abs().is_prime() or D == 1:
        Dfact = "\(%s\)" % str(D)
    else:
        Dfact = '%s = \(%s\)' % (str(D),Dfact)
    out += '<br>Discriminant: '
    out += Dfact
    out += '<br>Signature: '
    out += str(wnf.signature())
    out += '<br>Galois group: '+group_pretty_and_nTj(wnf.degree(),wnf.galois_t(), True)
    out += '<br>Class number: %s ' % str(wnf.class_number_latex())
    if wnf.can_class_number():
        out += wnf.short_grh_string()
    out += '</div>'
    out += '<div align="right">'
    out += '<a href="%s">%s home page</a>' % (str(url_for("number_fields.number_field_render_webpage", natural=label)),label)
    out += '</div>'
    return out

 def get_local_algebra(self, p):
     local_algebra_dict = self._data.get('loc_algebras', None)
     if local_algebra_dict is None:
         return None
     if str(p) in local_algebra_dict:
         R = PolynomialRing(QQ, 'x')
         palg = local_algebra_dict[str(p)]
         palgs = [R(str(s)) for s in palg.split(',')]
         try:
             palgstr = [
                 list2string([int(c) for c in pol.coefficients(sparse=False)])
                 for pol in palgs]
             palgrec = [db.lf_fields.lucky({'p': p, 'coeffs': map(int, c.split(','))}) for c in palgstr]
             return [
                 [
                     LF['label'],
                     latex(f),
                     int(LF['e']),
                     int(LF['f']),
                     int(LF['c']),
                     group_display_knowl(LF['n'], LF['galT']),
                     LF['t'],
                     LF['u'],
                     LF['slopes']
                 ]
                 for LF, f in zip(palgrec, palgs) ]
         except: # we were unable to find the local fields in the database
             return None

def web_latex_split_on(x, on=['+', '-']):
    if isinstance(x, (str, unicode)):
        return x
    else:
        A = "\( %s \)" % latex(x)
        for s in on:
            A = A.replace(s, '\) ' + s + ' \( ')
    return A

def render_hgm_webpage(args):
    data = None
    info = {}
    if 'label' in args:
        label = clean_input(args['label'])
        C = base.getDBConnection()
        data = C.hgm.motives.find_one({'label': label})
        if data is None:
            bread = get_bread([("Search error", url_for('.search'))])
            info['err'] = "Motive " + label + " was not found in the database."
            info['label'] = label
            return search_input_error(info, bread)
        title = 'Hypergeometric Motive:' + label
        A = data['A']
        B = data['B']
        tn,td = data['t']
        t = latex(QQ(str(tn)+'/'+str(td)))
        primes = [2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71]
        locinfo = data['locinfo']
        for j in range(len(locinfo)):
            locinfo[j] = [primes[j]] + locinfo[j]
            locinfo[j][2] = poly_with_factored_coeffs(locinfo[j][2], primes[j])
        hodge = data['hodge']
        prop2 = [
            ('Degree', '\(%s\)' % data['degree']),
            ('Weight',  '\(%s\)' % data['weight']),
            ('Conductor', '\(%s\)' % data['cond']),
        ]
        # Now add factorization of conductor
        Cond = ZZ(data['cond'])
        if not (Cond.abs().is_prime() or Cond == 1):
            data['cond'] = "%s=%s" % (str(Cond), factorint(data['cond']))

        info.update({
                    'A': A,
                    'B': B,
                    't': t,
                    'degree': data['degree'],
                    'weight': data['weight'],
                    'sign': data['sign'],
                    'sig': data['sig'],
                    'hodge': hodge,
                    'cond': data['cond'],
                    'req': data['req'],
                    'locinfo': locinfo
                    })
        AB_data, t_data = data["label"].split("_t")
        #AB_data = data["label"].split("_t")[0]
        friends = [("Motive family "+AB_data.replace("_"," "), url_for(".by_family_label", label = AB_data))]
        friends.append(('L-function', url_for("l_functions.l_function_hgm_page", label=AB_data, t='t'+t_data)))
#        if rffriend != '':
#            friends.append(('Discriminant root field', rffriend))


        bread = get_bread([(label, ' ')])
        return render_template("hgm-show-motive.html", credit=HGM_credit, title=title, bread=bread, info=info, properties2=prop2, friends=friends)

def web_latex(x, enclose=True):
    """
    Convert input to latex string unless it's a string or unicode. The key word
    argument `enclose` indicates whether to surround the string with
    `\(` and `\)` to tag it as an equation in html.

    Note:
    if input is a factored ideal, use web_latex_ideal_fact instead.

    Example:
    >>> x = var('x')
    >>> web_latex(x**23 + 2*x + 1)
    '\\( x^{23} + 2 \\, x + 1 \\)'
    """
    if isinstance(x, (str, unicode)):
        return x
    if enclose == True:
        return "\( %s \)" % latex(x)
    return " %s " % latex(x)

def web_latex_poly(pol, name='x', keepzeta=False):
    """
    Change the name of the variable in a polynomial.  If keepzeta, then don't change
    the name of zetaN in the defining polynomial of a cyclotomic field.
    (keepzeta not implemented yet)
    """
    # the next few lines were adapted from the lines after line 117 of web_newforms.py 
    oldname = latex(pol.parent().gen())
    subfrom = oldname.strip() 
    subfrom = subfrom.replace("\\","\\\\")  
    subfrom = subfrom.replace("{","\\{")   # because x_{0} means somethgn in a regular expression
    if subfrom[0].isalpha():
        subfrom = "\\b" + subfrom
    subto = name.replace("\\","\\\\")  
    subto += " "
#    print "converting from",subfrom,"to", subto, "of", latex(pol)
    newpol = re.sub(subfrom, subto, latex(pol))
#    print "result is",newpol
    return web_latex_split_on_pm(newpol)

 def latex(self, prec=None, name=None):
     if prec is None:
         qe = self.value()
     else:
         qe = self.value().truncate_powerseries(prec)
     wl = web_latex_split_on_re(qe)
     if name is not None and self.value().base_ring().degree()>1:
         return wl.replace(latex(self.value().base_ring().gen()), name)
     else:
         return wl

 def myhelper(self, coefmult):
     coef = string2list(coefmult[0])
     subfield = self.from_coeffs(coef)
     if subfield._data is None:
         deg = len(coef) - 1
         mypol = latex(coeff_to_poly(coef))
         mypol = mypol.replace(' ','').replace('+','%2B').replace('{', '%7B').replace('}','%7d')
         mypol = '<a title = "Field missing" knowl="nf.field.missing" kwargs="poly=%s">Deg %d</a>' % (mypol,deg)
         return [mypol, coefmult[1]]
     return [nf_display_knowl(subfield.get_label(),subfield.field_pretty()), coefmult[1]]