server-tools/excel_import_export/models/xlsx_import.py

# Copyright 2019 Ecosoft Co., Ltd (http://ecosoft.co.th/)
# License AGPL-3.0 or later (https://www.gnu.org/licenses/agpl.html)

import base64
import uuid
import xlrd
import xlwt
import time
from io import BytesIO
from . import common as co
from ast import literal_eval
from datetime import date, datetime as dt
from odoo.tools.float_utils import float_compare
from odoo import models, api, _
from odoo.exceptions import ValidationError
from odoo.tools.safe_eval import safe_eval


class XLSXImport(models.AbstractModel):
    _name = 'xlsx.import'
    _description = 'Excel Import AbstractModel'

    @api.model
    def get_eval_context(self, model=False, value=False):
        eval_context = {'float_compare': float_compare,
                        'time': time,
                        'datetime': dt,
                        'date': date,
                        'env': self.env,
                        'context': self._context,
                        'value': False,
                        'model': False,
                        }
        if model:
            eval_context.update({'model': self.env[model]})
        if value:
            if isinstance(value, str):  # Remove non Ord 128 character
                value = ''.join([i if ord(i) < 128 else ' ' for i in value])
            eval_context.update({'value': value})
        return eval_context

    @api.model
    def get_external_id(self, record):
        """ Get external ID of the record, if not already exists create one """
        ModelData = self.env['ir.model.data']
        xml_id = record.get_external_id()
        if not xml_id or (record.id in xml_id and xml_id[record.id] == ''):
            ModelData.create({'name': '%s_%s' % (record._table, record.id),
                              'module': 'excel_import_export',
                              'model': record._name,
                              'res_id': record.id, })
            xml_id = record.get_external_id()
        return xml_id[record.id]

    @api.model
    def _get_field_type(self, model, field):
        try:
            record = self.env[model].new()
            for f in field.split('/'):
                field_type = record._fields[f].type
                if field_type in ('one2many', 'many2many'):
                    record = record[f]
                else:
                    return field_type
        except Exception:
            raise ValidationError(
                _('Invalid declaration, %s has no valid field type') % field)

    @api.model
    def _delete_record_data(self, record, data_dict):
        """ If no _NODEL_, delete existing lines before importing """
        if not record or not data_dict:
            return
        try:
            for sheet_name in data_dict:
                worksheet = data_dict[sheet_name]
                line_fields = filter(lambda x: x != '_HEAD_', worksheet)
                for line_field in line_fields:
                    if '_NODEL_' not in line_field:
                        if line_field in record and record[line_field]:
                            record[line_field].unlink()
            # Remove _NODEL_ from dict
            for s, sv in data_dict.items():
                for f, fv in data_dict[s].items():
                    if '_NODEL_' in f:
                        new_fv = data_dict[s].pop(f)
                        data_dict[s][f.replace('_NODEL_', '')] = new_fv
        except Exception as e:
            raise ValidationError(_('Error deleting data\n%s') % e)

    @api.model
    def _get_line_vals(self, st, worksheet, model, line_field):
        """ Get values of this field from excel sheet """
        vals = {}
        for rc, columns in worksheet.get(line_field, {}).items():
            if not isinstance(columns, list):
                columns = [columns]
            for field in columns:
                rc, key_eval_cond = co.get_field_condition(rc)
                x_field, val_eval_cond = co.get_field_condition(field)
                row, col = co.pos2idx(rc)
                out_field = '%s/%s' % (line_field, x_field)
                field_type = self._get_field_type(model, out_field)
                vals.update({out_field: []})
                for idx in range(row, st.nrows):
                    value = co._get_cell_value(st.cell(idx, col),
                                               field_type=field_type)
                    eval_context = self.get_eval_context(model=model,
                                                         value=value)
                    if key_eval_cond:
                        value = safe_eval(key_eval_cond, eval_context)
                    if val_eval_cond:
                        value = safe_eval(val_eval_cond, eval_context)
                    vals[out_field].append(value)
                if not filter(lambda x: x != '', vals[out_field]):
                    vals.pop(out_field)
        return vals

    @api.model
    def _import_record_data(self, import_file, record, data_dict):
        """ From complex excel, create temp simple excel and do import """
        if not data_dict:
            return
        try:
            header_fields = []
            decoded_data = base64.decodestring(import_file)
            wb = xlrd.open_workbook(file_contents=decoded_data)
            col_idx = 0
            out_wb = xlwt.Workbook()
            out_st = out_wb.add_sheet("Sheet 1")
            xml_id = record and self.get_external_id(record) or \
                '%s.%s' % ('xls', uuid.uuid4())
            out_st.write(0, 0, 'id')  # id and xml_id on first column
            out_st.write(1, 0, xml_id)
            header_fields.append('id')
            col_idx += 1
            model = record._name
            for sheet_name in data_dict:  # For each Sheet
                worksheet = data_dict[sheet_name]
                st = False
                if isinstance(sheet_name, str):
                    st = co.xlrd_get_sheet_by_name(wb, sheet_name)
                elif isinstance(sheet_name, int):
                    st = wb.sheet_by_index(sheet_name - 1)
                if not st:
                    raise ValidationError(
                        _('Sheet %s not found') % sheet_name)
                # HEAD updates
                for rc, field in worksheet.get('_HEAD_', {}).items():
                    rc, key_eval_cond = co.get_field_condition(rc)
                    field, val_eval_cond = co.get_field_condition(field)
                    field_type = self._get_field_type(model, field)
                    value = False
                    try:
                        row, col = co.pos2idx(rc)
                        value = co._get_cell_value(st.cell(row, col),
                                                   field_type=field_type)
                    except Exception:
                        pass
                    eval_context = self.get_eval_context(model=model,
                                                         value=value)
                    if key_eval_cond:
                        value = str(safe_eval(key_eval_cond, eval_context))
                    if val_eval_cond:
                        value = str(safe_eval(val_eval_cond, eval_context))
                    out_st.write(0, col_idx, field)  # Next Column
                    out_st.write(1, col_idx, value)  # Next Value
                    header_fields.append(field)
                    col_idx += 1
                # Line Items
                line_fields = filter(lambda x: x != '_HEAD_', worksheet)
                for line_field in line_fields:
                    vals = self._get_line_vals(st, worksheet,
                                               model, line_field)
                    for field in vals:
                        # Columns, i.e., line_ids/field_id
                        out_st.write(0, col_idx, field)
                        header_fields.append(field)
                        # Data
                        i = 1
                        for value in vals[field]:
                            out_st.write(i, col_idx, value)
                            i += 1
                        col_idx += 1
            content = BytesIO()
            out_wb.save(content)
            content.seek(0)  # Set index to 0, and start reading
            xls_file = content.read()
            # Do the import
            Import = self.env['base_import.import']
            imp = Import.create({
                'res_model': model,
                'file': xls_file,
                'file_type': 'application/vnd.ms-excel',
                'file_name': 'temp.xls',
            })
            errors = imp.do(
                header_fields,
                header_fields,
                {'headers': True,
                 'advanced': True,
                 'keep_matches': False,
                 'encoding': '',
                 'separator': '',
                 'quoting': '"',
                 'date_style': '',
                 'datetime_style': '%Y-%m-%d %H:%M:%S',
                 'float_thousand_separator': ',',
                 'float_decimal_separator': '.',
                 'fields': []})
            if errors.get('messages'):
                message = errors['messages']['message'].encode('utf-8')
                raise ValidationError(message)
            return self.env.ref(xml_id)
        except xlrd.XLRDError:
            raise ValidationError(
                _('Invalid file style, only .xls or .xlsx file allowed'))
        except Exception as e:
            raise ValidationError(_('Error importing data\n%s') % e)

    @api.model
    def _post_import_operation(self, record, operation):
        """ Run python code after import """
        if not record or not operation:
            return
        try:
            if '${' in operation:
                code = (operation.split('${'))[1].split('}')[0]
                eval_context = {'object': record}
                safe_eval(code, eval_context)
        except Exception as e:
            raise ValidationError(_('Post import operation error\n%s') % e)

    @api.model
    def import_xlsx(self, import_file, template,
                    res_model=False, res_id=False):
        """
        - If res_id = False, we want to create new document first
        - Delete fields' data according to data_dict['__IMPORT__']
        - Import data from excel according to data_dict['__IMPORT__']
        """
        self = self.sudo()
        if res_model and template.res_model != res_model:
            raise ValidationError(_("Template's model mismatch"))
        record = self.env[template.res_model].browse(res_id)
        data_dict = literal_eval(template.instruction.strip())
        if not data_dict.get('__IMPORT__'):
            raise ValidationError(
                _("No data_dict['__IMPORT__'] in template %s") % template.name)
        if record:
            # Delete existing data first
            self._delete_record_data(record, data_dict['__IMPORT__'])
        # Fill up record with data from excel sheets
        record = self._import_record_data(import_file, record,
                                          data_dict['__IMPORT__'])
        # Post Import Operation, i.e., cleanup some data
        if data_dict.get('__POST_IMPORT__', False):
            self._post_import_operation(record, data_dict['__POST_IMPORT__'])
        return record