1
0
forked from ScoDoc/ScoDoc
ScoDoc/app/scodoc/sco_poursuite_dut.py

225 lines
8.4 KiB
Python
Raw Normal View History

2020-09-26 16:19:37 +02:00
# -*- mode: python -*-
# -*- coding: utf-8 -*-
##############################################################################
#
# Gestion scolarite IUT
#
2021-01-01 17:51:08 +01:00
# Copyright (c) 1999 - 2021 Emmanuel Viennet. All rights reserved.
2020-09-26 16:19:37 +02:00
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
#
# Emmanuel Viennet emmanuel.viennet@gmail.com
#
##############################################################################
"""Extraction de données pour poursuites d'études
Recapitule tous les semestres validés dans une feuille excel.
"""
2021-02-03 22:00:41 +01:00
import collections
2020-09-26 16:19:37 +02:00
import app.scodoc.sco_utils as scu
from app.scodoc import sco_abs
from app.scodoc import sco_core
from app.scodoc import sco_formsemestre
from app.scodoc import sco_groups
from app.scodoc import sco_preferences
from app.scodoc import sco_etud
from app.scodoc import VERSION
from app.scodoc.gen_tables import GenTable
from app.scodoc.sco_codes_parcours import code_semestre_validant, code_semestre_attente
2020-09-26 16:19:37 +02:00
def etud_get_poursuite_info(context, sem, etud):
2021-01-01 18:40:47 +01:00
"""{ 'nom' : ..., 'semlist' : [ { 'semestre_id': , 'moy' : ... }, {}, ...] }"""
2020-09-26 16:19:37 +02:00
I = {}
I.update(etud) # copie nom, prenom, civilite, ...
2020-09-26 16:19:37 +02:00
# Now add each semester, starting from the first one
semlist = []
current_id = sem["semestre_id"]
for sem_id in range(1, current_id + 1):
sem_descr = None
for s in etud["sems"]:
if s["semestre_id"] == sem_id:
etudid = etud["etudid"]
2021-06-13 23:37:14 +02:00
nt = sco_core.get_notes_cache(
context,
).get_NotesTable(context, s["formsemestre_id"])
2020-09-26 16:19:37 +02:00
dec = nt.get_etud_decision_sem(etudid)
# Moyennes et rangs des UE
ues = nt.get_ues(filter_sport=True)
moy_ues = [
(
ue["acronyme"],
2021-02-04 20:02:44 +01:00
scu.fmt_note(nt.get_etud_ue_status(etudid, ue["ue_id"])["moy"]),
2020-09-26 16:19:37 +02:00
)
for ue in ues
]
rg_ues = [
("rang_" + ue["acronyme"], nt.ue_rangs[ue["ue_id"]][0][etudid])
for ue in ues
]
# Moyennes et rang des modules
modimpls = nt.get_modimpls() # recupération des modules
modules = []
rangs = []
for ue in ues: # on parcourt chaque UE
for modimpl in modimpls: # dans chaque UE les modules
if modimpl["module"]["ue_id"] == ue["ue_id"]:
codeModule = modimpl["module"]["code"]
2021-02-04 20:02:44 +01:00
noteModule = scu.fmt_note(
2020-09-26 16:19:37 +02:00
nt.get_etud_mod_moy(modimpl["moduleimpl_id"], etudid)
)
if noteModule != "NI": # si étudiant inscrit au module
rangModule = nt.mod_rangs[modimpl["moduleimpl_id"]][0][
etudid
]
modules.append([codeModule, noteModule])
rangs.append(["rang_" + codeModule, rangModule])
# Absences
2021-01-10 18:54:39 +01:00
AbsSemEtud = sco_abs.getAbsSemEtud(context, nt.sem, etudid)
2020-09-26 16:19:37 +02:00
NbAbs = AbsSemEtud.CountAbs()
NbAbsJust = AbsSemEtud.CountAbsJust()
if (
dec
and not sem_descr # not sem_descr pour ne prendre que le semestre validé le plus récent
and (
code_semestre_validant(dec["code"])
or code_semestre_attente(dec["code"])
)
and nt.get_etud_etat(etudid) == "I"
):
d = [
2021-02-04 20:02:44 +01:00
("moy", scu.fmt_note(nt.get_etud_moy_gen(etudid))),
("moy_promo", scu.fmt_note(nt.moy_moy)),
2020-09-26 16:19:37 +02:00
("rang", nt.get_etud_rang(etudid)),
("effectif", len(nt.T)),
("date_debut", s["date_debut"]),
("date_fin", s["date_fin"]),
("periode", "%s - %s" % (s["mois_debut"], s["mois_fin"])),
("AbsNonJust", NbAbs - NbAbsJust),
("AbsJust", NbAbsJust),
]
d += (
moy_ues + rg_ues + modules + rangs
) # ajout des 2 champs notes des modules et classement dans chaque module
sem_descr = collections.OrderedDict(d)
if not sem_descr:
sem_descr = collections.OrderedDict(
[
("moy", ""),
("moy_promo", ""),
("rang", ""),
("effectif", ""),
("date_debut", ""),
("date_fin", ""),
("periode", ""),
]
)
sem_descr["semestre_id"] = sem_id
semlist.append(sem_descr)
I["semlist"] = semlist
return I
def _flatten_info(info):
# met la liste des infos semestres "a plat"
# S1_moy, S1_rang, ..., S2_moy, ...
ids = []
for s in info["semlist"]:
for k, v in s.items():
if k != "semestre_id":
label = "S%s_%s" % (s["semestre_id"], k)
info[label] = v
ids.append(label)
return ids
def _getEtudInfoGroupes(context, group_ids, etat=None):
"""liste triée d'infos (dict) sur les etudiants du groupe indiqué.
Attention: lent, car plusieurs requetes SQL par etudiant !
"""
etuds = []
for group_id in group_ids:
members = sco_groups.get_group_members(context, group_id, etat=etat)
for m in members:
etud = sco_etud.get_etud_info(etudid=m["etudid"], filled=True)[0]
etuds.append(etud)
return etuds
2020-09-26 16:19:37 +02:00
def formsemestre_poursuite_report(
context, formsemestre_id, format="html", REQUEST=None
):
2021-01-01 18:40:47 +01:00
"""Table avec informations "poursuite" """
2020-09-26 16:19:37 +02:00
sem = sco_formsemestre.get_formsemestre(context, formsemestre_id)
etuds = _getEtudInfoGroupes(
context, [sco_groups.get_default_group(context, formsemestre_id)]
2020-09-26 16:19:37 +02:00
)
infos = []
ids = []
for etud in etuds:
etud["_nom_target"] = "ficheEtud?etudid=" + etud["etudid"]
etud["_prenom_target"] = "ficheEtud?etudid=" + etud["etudid"]
etud["_nom_td_attrs"] = 'id="%s" class="etudinfo"' % (etud["etudid"])
info = etud_get_poursuite_info(context, sem, etud)
idd = _flatten_info(info)
# On recupere la totalite des UEs dans ids
for id in idd:
if id not in ids:
ids += [id]
infos.append(info)
#
column_ids = (
("civilite_str", "nom", "prenom", "annee", "date_naissance")
2020-09-26 16:19:37 +02:00
+ tuple(ids)
+ ("debouche",)
)
titles = {}
for c in column_ids:
titles[c] = c
tab = GenTable(
titles=titles,
columns_ids=column_ids,
rows=infos,
# html_col_width='4em',
html_sortable=True,
html_class="table_leftalign table_listegroupe",
pdf_link=False, # pas d'export pdf
2021-06-13 19:12:20 +02:00
preferences=sco_preferences.SemPreferences(context, formsemestre_id),
2020-09-26 16:19:37 +02:00
)
2021-02-04 20:02:44 +01:00
tab.filename = scu.make_filename("poursuite " + sem["titreannee"])
2020-09-26 16:19:37 +02:00
2021-02-04 20:02:44 +01:00
tab.origin = "Généré par %s le " % VERSION.SCONAME + scu.timedate_human_repr() + ""
2020-09-26 16:19:37 +02:00
tab.caption = "Récapitulatif %s." % sem["titreannee"]
tab.html_caption = "Récapitulatif %s." % sem["titreannee"]
tab.base_url = "%s?formsemestre_id=%s" % (REQUEST.URL0, formsemestre_id)
return tab.make_page(
context,
title="""<h2 class="formsemestre">Poursuite d'études</h2>""",
init_qtip=True,
javascripts=["js/etud_info.js"],
format=format,
REQUEST=REQUEST,
with_html_headers=True,
)