tricc-oo 1.0.2__py3-none-any.whl → 1.4.15__py3-none-any.whl

tricc_oo/strategies/output/xls_form.py

@@ -0,0 +1,697 @@
+"""
+Strategy to build the skyp logic following the XLSForm way
+
+"""
+
+import datetime
+import logging
+import os
+
+import pandas as pd
+
+from tricc_oo.converters.tricc_to_xls_form import *
+
+from tricc_oo.models import (
+    TriccNodeActivity,
+    TriccGroup,
+    TriccOperation,
+    TriccOperator
+)
+from tricc_oo.models.lang import SingletonLangClass
+
+from tricc_oo.visitors.tricc import (
+    check_stashed_loop,
+    walktrhough_tricc_node_processed_stached,
+    is_ready_to_process,
+    process_reference,
+    get_node_expressions,
+    TRICC_TRUE_VALUE,
+    TRICC_FALSE_VALUE
+)
+from tricc_oo.serializers.xls_form import (
+    CHOICE_MAP,
+    SURVEY_MAP,
+    end_group,
+    generate_xls_form_export,
+    start_group,
+    BOOLEAN_MAP
+)
+from tricc_oo.strategies.output.base_output_strategy import BaseOutPutStrategy
+
+logger = logging.getLogger("default")
+
+
+
+"""
+    The XLSForm strategy is a strategy that will generate the XLSForm logic
+    The XLSForm logic is a logic that is based on the XLSForm format
+    The XLSForm format is a format that is used by the ODK Collect application
+    The ODK Collect application is an application that is used to collect data on mobile devices
+
+    document below function
+
+    generate_xls_form_condition
+    generate_xls_form_relevance
+    generate_xls_form_calculate
+    generate_xls_form_export
+    start_group
+    end_group
+    walktrhough_tricc_node_processed_stached
+    check_stashed_loop
+    generate_xls_form_export
+    generate_xls_form_export
+    
+"""
+langs = SingletonLangClass()
+
+OPERATOR_COALESCE_FALLBACK = {
+    TriccOperator.ISTRUE: 0,
+    TriccOperator.ISFALSE: 1,
+    TriccOperator.ISNOTTRUE: 0,
+    TriccOperator.ISNOTFALSE: 1,
+    TriccOperator.MORE: -2147483648,
+    TriccOperator.MORE_OR_EQUAL: -2147483648,
+    TriccOperator.LESS: 2147483647,
+    TriccOperator.LESS_OR_EQUAL: 2147483647
+}
+
+class XLSFormStrategy(BaseOutPutStrategy):
+    pd.set_option('display.max_colwidth', None)
+    df_survey = pd.DataFrame(columns=SURVEY_MAP.keys())
+    df_calculate = pd.DataFrame(columns=SURVEY_MAP.keys())
+    df_choice = pd.DataFrame(columns=CHOICE_MAP.keys())
+    calculates = {}
+    # add save nodes and merge nodes
+
+
+    def clean_coalesce(self, expression):   
+        if re.match(r"^coalesce\(\${[^}]+},''\)$", expression):
+            return expression[9:-4]
+        return expression
+
+    def generate_base(self, node, **kwargs):
+        return self.generate_xls_form_condition(node, **kwargs) 
+
+    def generate_relevance(self, node, **kwargs):
+        return self.generate_xls_form_relevance(node, **kwargs)
+
+    def generate_calculate(self, node, **kwargs):
+        return self.generate_xls_form_calculate(node, **kwargs)
+
+    def __init__(self, project, output_path):
+        super().__init__(project, output_path)
+        self.do_clean()
+
+    def do_clean(self, **kwargs):
+        self.calculates = {}
+        self.used_calculates = {}
+
+    def get_kwargs(self):
+        return {
+            "df_survey": self.df_survey,
+            "df_choice": self.df_choice,
+            "df_calculate": self.df_calculate,
+            "calculates": self.calculates
+        }
+
+    def generate_export(self, node, **kwargs):
+        return generate_xls_form_export(self, node, **kwargs)
+
+
+    def export(self, start_pages, version):
+        if start_pages["main"].root.form_id is not None:
+            form_id = str(start_pages["main"].root.form_id)
+        else:
+            logger.critical("form id required in the first start node")
+            exit(1)
+        title = start_pages["main"].root.label
+        file_name = form_id + ".xlsx"
+        # make a 'settings' tab
+        now = datetime.datetime.now()
+        indx = [[1]]
+
+        settings = {
+            "form_title": title,
+            "form_id": form_id,
+            "version": version,
+            "default_language": "English (en)",
+            "style": "pages",
+        }
+        df_settings = pd.DataFrame(settings, index=indx)
+        df_settings.head()
+
+        newpath = os.path.join(self.output_path, file_name)
+        # create newpath if it not exists
+        if not os.path.exists(self.output_path):
+            os.makedirs(self.output_path)
+
+        # create a Pandas Excel writer using XlsxWriter as the engine
+        writer = pd.ExcelWriter(newpath, engine="xlsxwriter")
+        if len(self.df_survey[self.df_survey['name'] == 'version'] ):
+            self.df_survey.loc[ self.df_survey['name'] == 'version', 'label'] = f"v{version}"
+        self.df_survey.to_excel(writer, sheet_name="survey", index=False)
+        self.df_choice.to_excel(writer, sheet_name="choices", index=False)
+        df_settings.to_excel(writer, sheet_name="settings", index=False)
+
+        # close the Pandas Excel writer and output the Excel file
+        # writer.save()
+
+        # run this on a windows python instance because if not then the generated xlsx file remains open
+        writer.close()
+        # writer.handles = None
+
+    def process_export(self, start_pages, **kwargs):
+        self.activity_export(start_pages["main"], **kwargs)
+
+    def activity_export(self, activity, processed_nodes=None, **kwargs):
+        if processed_nodes is None:
+            processed_nodes = OrderedSet()
+        stashed_nodes = OrderedSet()
+        # The stashed node are all the node that have all their prevnode processed but not from the same group
+        # This logic works only because the prev node are ordered by group/parent ..
+        skip_header = 0
+        groups = {}
+        cur_group = activity
+        groups[activity.id] = 0
+        path_len = 0
+        # keep the vesrions on the group id, max version
+        start_group(self, cur_group=cur_group, groups=groups, **self.get_kwargs())
+        walktrhough_tricc_node_processed_stached(
+            activity.root,
+            self.generate_export,
+            processed_nodes,
+            stashed_nodes,
+            path_len,
+            cur_group=activity.root.group,
+            recursive=False,
+            **self.get_kwargs()
+        )
+        end_group(self, cur_group=activity, groups=groups, **self.get_kwargs())
+        # we save the survey data frame
+        df_survey_final = pd.DataFrame(columns=SURVEY_MAP.keys())
+        if len(self.df_survey) > (2 + skip_header):
+            df_survey_final = self.df_survey
+        ## MANAGE STASHED NODES
+        prev_stashed_nodes = stashed_nodes.copy()
+        loop_count = 0
+        len_prev_processed_nodes = 0
+        while len(stashed_nodes) > 0:
+            self.df_survey = pd.DataFrame(columns=SURVEY_MAP.keys())
+            loop_count = check_stashed_loop(
+                stashed_nodes,
+                prev_stashed_nodes,
+                processed_nodes,
+                len_prev_processed_nodes,
+                loop_count,
+            )
+            prev_stashed_nodes = stashed_nodes.copy()
+            len_prev_processed_nodes = len(processed_nodes)
+            if len(stashed_nodes) > 0:
+                s_node = stashed_nodes.pop()
+                # while len(stashed_nodes)>0 and isinstance(s_node,TriccGroup):
+                #    s_node = stashed_nodes.pop()
+                if s_node.group is None:
+                    logger.critical(
+                        "ERROR group is none for node {}".format(s_node.get_name())
+                    )
+                start_group(
+                    self,
+                    cur_group=s_node.group,
+                    groups=groups,
+                    relevance=True,
+                    **self.get_kwargs()
+                )
+                # arrange empty group
+                
+                walktrhough_tricc_node_processed_stached(
+                    s_node,
+                    self.generate_export,
+                    processed_nodes,
+                    stashed_nodes,
+                    path_len,
+                    groups=groups,
+                    cur_group=s_node.group,
+                    recursive=False,
+                    **self.get_kwargs()
+                )
+                # add end group if new node where added OR if the previous end group was removed
+                end_group(self, cur_group=s_node.group, groups=groups, **self.get_kwargs())
+                # if two line then empty grou
+                if len(self.df_survey) > (2 + skip_header):
+                    if cur_group == s_node.group:
+                        # drop the end group (to merge)
+                        logger.debug(
+                            "printing same group {}::{}::{}::{}".format(
+                                s_node.group.__class__,
+                                s_node.group.get_name(),
+                                s_node.id,
+                                s_node.group.instance,
+                            )
+                        )
+                        if len(df_survey_final):
+                            df_survey_final.drop(
+                                index=df_survey_final.index[-1], axis=0, inplace=True
+                            )
+                            self.df_survey = self.df_survey[(1 + skip_header) :]
+                        df_survey_final = pd.concat(
+                            [df_survey_final, self.df_survey], ignore_index=True
+                        )
+
+                    else:
+                        logger.debug(
+                            "printing group {}::{}::{}::{}".format(
+                                s_node.group.__class__,
+                                s_node.group.get_name(),
+                                s_node.id,
+                                s_node.group.instance,
+                            )
+                        )
+                        df_survey_final = pd.concat(
+                            [df_survey_final, self.df_survey], ignore_index=True
+                        )
+                    cur_group = s_node.group
+
+        # add the calulate
+        self.df_calculate = self.df_calculate.dropna(axis=0, subset=["calculation"])
+        df_empty_calc = self.df_calculate[self.df_calculate["calculation"] == ""]
+        self.df_survey.reset_index(drop=True, inplace=True)
+        self.df_calculate.reset_index(drop=True, inplace=True)
+        self.df_calculate = self.df_calculate.drop(df_empty_calc.index)
+        self.df_survey = pd.concat(
+            [df_survey_final, self.df_calculate], ignore_index=True
+        )
+        df_duplicate = self.df_calculate[
+            self.df_calculate.duplicated(subset=["calculation"], keep="first")
+        ]
+        # self.df_survey=self.df_survey.drop_duplicates(subset=['name'])
+        for index, drop_calc in df_duplicate.iterrows():
+            # remove the duplicate
+            replace_name = False
+            # find the actual calcualte
+            similar_calc = self.df_survey[
+                (drop_calc["calculation"] == self.df_survey["calculation"])
+                & (self.df_survey["type"] == "calculate")
+            ]
+            same_calc = self.df_survey[self.df_survey["name"] == drop_calc["name"]]
+            if len(same_calc) > 1:
+                # check if all calc have the same name
+                if len(same_calc) == len(similar_calc):
+                    # drop all but one
+                    self.df_survey.drop(same_calc.index[1:])
+                elif len(same_calc) < len(similar_calc):
+                    self.df_survey.drop(same_calc.index)
+                    replace_name = True
+            elif len(same_calc) == 1:
+                self.df_survey.drop(similar_calc.index)
+                replace_name = True
+
+            if replace_name:
+                save_calc = self.df_survey[
+                    (drop_calc["calculation"] == self.df_survey["calculation"])
+                    & (self.df_survey["type"] == "calculate")
+                ]
+                if len(save_calc) >= 1:
+                    save_calc = save_calc.iloc[0]
+                    if save_calc["name"] != drop_calc["name"]:
+                        self.df_survey.replace(
+                            "\$\{" + drop_calc["name"] + "\}",
+                            "\$\{" + save_calc["name"] + "\}",
+                            regex=True,
+                        )
+                else:
+                    logger.critical(
+                        "duplicate reference not found for calculation: {}".format(
+                            drop_calc["calculation"]
+                        )
+                    )
+        for index, empty_calc in df_empty_calc.iterrows():
+            self.df_survey.replace("\$\{" + empty_calc["name"] + "\}", "1", regex=True)
+
+        # TODO try to reinject calc to reduce complexity
+        for i, c in self.df_calculate[
+            ~self.df_calculate["name"].isin(self.df_survey["name"])
+        ].iterrows():
+            real_calc = re.find(r"^number\((.+)\)$", c["calculation"])
+            if real_calc is not None and real_calc != "":
+                self.df_survey[~self.df_survey["name"] == c["name"]].replace(
+                    real_calc, "\$\{" + c["name"] + "\}"
+                )
+        
+        df_duplicate = self.df_survey[
+            self.df_survey.duplicated(subset=["name"], keep="first")
+        ]
+        for index, duplicate in df_duplicate.iterrows(): 
+            logger.critical(f"duplicate survey name: {duplicate['name']}")
+        self.df_survey.reset_index(drop=True, inplace=True)
+        return processed_nodes
+
+    def get_tricc_operation_expression(self, operation):
+        ref_expressions = []
+        if not hasattr(operation, 'reference'):
+            return self.get_tricc_operation_operand(operation) 
+        
+        operator = getattr(operation, 'operator', '')
+        coalesce_fallback = OPERATOR_COALESCE_FALLBACK[operator] if operator in OPERATOR_COALESCE_FALLBACK else  "''"
+        for r in operation.reference:
+            if isinstance(r, list):
+                r_expr = [
+                    self.get_tricc_operation_expression(sr) if isinstance(sr, TriccOperation)
+                    else self.get_tricc_operation_operand(sr,coalesce_fallback) 
+                    for sr in r
+                ]
+            elif isinstance(r, TriccOperation):
+                r_expr = self.get_tricc_operation_expression(r)
+            else:
+                r_expr = self.get_tricc_operation_operand(r,coalesce_fallback)
+            if isinstance(r_expr, TriccReference):
+                r_expr = self.get_tricc_operation_operand(r_expr,coalesce_fallback)
+            ref_expressions.append(r_expr)
+        
+        # build lower level
+        if hasattr(self,f"tricc_operation_{operation.operator}"):
+            callable = getattr(self,f"tricc_operation_{operation.operator}")
+            return callable(ref_expressions)   
+        else:
+            raise NotImplementedError(f"This type of opreation '{operation.operator}' is not supported in this strategy")
+        
+    def tricc_operation_multiplied(self, ref_expressions):
+        return '*'.join(ref_expressions)
+    def tricc_operation_divided(self, ref_expressions):
+        return f"{ref_expressions[0]} div {ref_expressions[1]}"
+    def tricc_operation_modulo(self, ref_expressions):
+        return f"{ref_expressions[0]} mod {ref_expressions[1]}"
+    def tricc_operation_coalesce(self, ref_expressions):
+        return f"coalesce({','.join(map(self.clean_coalesce, ref_expressions))})"
+    def tricc_operation_module(self, ref_expressions):
+        return f"{ref_expressions[0]} mod {ref_expressions[1]}"
+    def tricc_operation_minus(self, ref_expressions):
+        if len(ref_expressions)>1:
+            return ' - '.join(ref_expressions)
+        elif len(ref_expressions)==1:
+            return f'-{ref_expressions[0]}'
+    def tricc_operation_plus(self, ref_expressions):
+        return ' + '.join(ref_expressions)
+    def tricc_operation_not(self, ref_expressions):
+        return f"not({ref_expressions[0]})"
+    def tricc_operation_and(self, ref_expressions):
+        if len(ref_expressions) == 1:
+            return ref_expressions[0]
+        if len(ref_expressions)>1: 
+            ref_expressions = [f"({r})" if isinstance(r, str) and any(op in r for op in [' or ',' + ',' - '])else r for r in ref_expressions]
+            return ' and '.join(map(str, ref_expressions))
+        else:
+            return '1'
+
+    def tricc_operation_or(self, ref_expressions):
+        if len(ref_expressions) == 1:
+            return ref_expressions[0]
+        if len(ref_expressions)>1: 
+            ref_expressions = [f"({r})" if isinstance(r, str) and any(op in r for op in [' and ',' + ',' - ']) else r for r in ref_expressions]
+            return ' or '.join(map(str, ref_expressions))
+        else:
+            return '1'
+
+
+
+    def tricc_operation_native(self, ref_expressions):
+        if len(ref_expressions)>0:
+            if ref_expressions[0] =='GetChoiceName':
+                return f"jr:choice-name({self.clean_coalesce(ref_expressions[1])}, '{self.clean_coalesce(ref_expressions[2])}')"
+            elif ref_expressions[0] =='GetFacilityParam':
+                return '0'
+                #return f"jr:choice-name({','.join(ref_expressions[1:])})"
+            else: 
+                return f"{ref_expressions[0]}({','.join(ref_expressions[1:])})"
+        
+    def tricc_operation_istrue(self, ref_expressions):
+        if str(BOOLEAN_MAP[str(TRICC_TRUE_VALUE)]).isnumeric():
+            return f"{ref_expressions[0]}>={BOOLEAN_MAP[str(TRICC_TRUE_VALUE)]}"
+        else:
+            return f"{ref_expressions[0]}={BOOLEAN_MAP[str(TRICC_TRUE_VALUE)]}"
+    def tricc_operation_isfalse(self, ref_expressions):
+        if str(BOOLEAN_MAP[str(TRICC_FALSE_VALUE)]).isnumeric():
+            return f"{ref_expressions[0]}={BOOLEAN_MAP[str(TRICC_FALSE_VALUE)]}"
+        else:
+            return f"{ref_expressions[0]}={BOOLEAN_MAP[str(TRICC_FALSE_VALUE)]}"
+    def tricc_operation_parenthesis(self, ref_expressions):
+        return f"({ref_expressions[0]})"
+    def tricc_operation_selected(self, ref_expressions):
+        parts = []
+        for s in ref_expressions[1:]:
+            # for option with numeric value
+            cleaned_s = s if isinstance(s, str)  else '\''+str(s)+'\''
+            parts.append(f"selected({self.clean_coalesce(ref_expressions[0])}, {cleaned_s})")
+        if len(parts) == 1:
+            return parts[0]
+        else:
+            return self.tricc_operation_or(parts)
+    def tricc_operation_more_or_equal(self, ref_expressions):
+        return f"{ref_expressions[0]}>={ref_expressions[1]}"
+    def tricc_operation_less_or_equal(self, ref_expressions):
+        return f"{ref_expressions[0]}<={ref_expressions[1]}"
+    def tricc_operation_more(self, ref_expressions):
+        return f"{ref_expressions[0]}>{ref_expressions[1]}"
+    def tricc_operation_less(self, ref_expressions):
+        return f"{ref_expressions[0]}<{ref_expressions[1]}"
+    def tricc_operation_between(self, ref_expressions):
+        return  f"{ref_expressions[0]}>={ref_expressions[1]} and {ref_expressions[0]} < {ref_expressions[2]}"
+    def tricc_operation_equal(self, ref_expressions):
+        return f"{ref_expressions[0]}={ref_expressions[1]}"
+    def tricc_operation_not_equal(self, ref_expressions):
+        return f"{ref_expressions[0]}!={ref_expressions[1]}"
+    def tricc_operation_isnull(self, ref_expressions):
+        return f"{ref_expressions[0]}=''"
+    def tricc_operation_isnotnull(self, ref_expressions):
+        return f"{ref_expressions[0]}!=''"
+    
+    def tricc_operation_isnottrue(self, ref_expressions):
+        if str(BOOLEAN_MAP[str(TRICC_TRUE_VALUE)]).isnumeric():
+            return f"{ref_expressions[0]}<{BOOLEAN_MAP[str(TRICC_TRUE_VALUE)]}"
+        else:
+            return f"{ref_expressions[0]}!={BOOLEAN_MAP[str(TRICC_TRUE_VALUE)]}"
+    def tricc_operation_isnotfalse(self, ref_expressions):
+        if str(BOOLEAN_MAP[str(TRICC_FALSE_VALUE)]).isnumeric():
+            return f"{ref_expressions[0]}>{BOOLEAN_MAP[str(TRICC_FALSE_VALUE)]}"
+        else:
+            return f"{ref_expressions[0]}!={BOOLEAN_MAP[str(TRICC_FALSE_VALUE)]}"
+    def tricc_operation_notexist(self, ref_expressions):
+        return f"{ref_expressions[0]}=''"
+
+    
+    
+    def tricc_operation_case(self, ref_expressions):
+        ifs = 0
+        parts = []
+        else_found = False
+        if not isinstance(ref_expressions[0], list):
+            return self.tricc_operation_ifs(ref_expressions)
+        for i in range(int(len(ref_expressions))):
+            if isinstance(ref_expressions[i], list):
+                parts.append(f"if({ref_expressions[i][0]},{ref_expressions[i][1]}")
+                ifs += 1
+            else:
+                else_found = True
+                parts.append(ref_expressions[i])
+        #join the if
+        exp = ','.join(map(str,parts))
+        # in case there is no default put ''
+        if not else_found:
+            exp += ",''"
+        #add the closing )
+        for i in range(ifs):
+            exp += ")"
+        return exp
+    
+    def tricc_operation_ifs(self, ref_expressions):
+        ifs = 0
+        parts = []
+        else_found = False
+        for i in range(int(len(ref_expressions[1:]))):
+            if isinstance(ref_expressions[i+1], list):
+                parts.append(f"if({ref_expressions[0]}={ref_expressions[i+1][0]},{ref_expressions[i+1][1]}")
+                ifs += 1
+            else:
+                else_found = True
+                parts.append(ref_expressions[i+1])
+        #join the if
+        exp = ','.join(parts)
+        # in case there is no default put ''
+        if not else_found:
+            exp += ",''"
+        #add the closing )
+        for i in range(ifs):
+            exp += ")"
+        return exp
+    
+    def tricc_operation_if(self, ref_expressions):
+        return f"if({ref_expressions[0]},{ref_expressions[1]},{ref_expressions[2]})"
+    
+    def tricc_operation_contains(self, ref_expressions):
+        return f"contains('{self.clean_coalesce(ref_expressions[0])}', '{self.clean_coalesce(ref_expressions[1])}')"
+    
+    def tricc_operation_exists(self, ref_expressions):
+        parts = []
+        for ref in ref_expressions:
+            parts.append(self.tricc_operation_not_equal([self.tricc_operation_coalesce([ref, "''"]), "''"]))
+        return self.tricc_operation_and(parts)
+    
+    def tricc_operation_cast_number(self, ref_expressions):
+        if isinstance(ref_expressions[0], (int, float,)):
+            return f"{ref_expressions[0]}"
+        elif not ref_expressions or ref_expressions[0] == '':
+            logger.warning("empty cast number")
+            return '0'
+        elif ref_expressions[0] == 'true' or ref_expressions[0] is True or ref_expressions[0] == BOOLEAN_MAP[str(TRICC_TRUE_VALUE)]:
+            return 1
+        elif ref_expressions[0] == 'false' or ref_expressions[0] is False or ref_expressions[0] == BOOLEAN_MAP[str(TRICC_FALSE_VALUE)]:
+            return 0
+        else:
+            return f"number({ref_expressions[0]})"
+        
+    def tricc_operation_cast_integer(self, ref_expressions):
+        if isinstance(ref_expressions[0], (int, float,)):
+            return f"{ref_expressions[0]}"
+        elif not ref_expressions or ref_expressions[0] == '':
+            logger.warning("empty cast number")
+            return '0'
+        elif ref_expressions[0] == 'true' or ref_expressions[0] is True or ref_expressions[0] == BOOLEAN_MAP[str(TRICC_TRUE_VALUE)]:
+            return 1
+        elif ref_expressions[0] == 'false' or ref_expressions[0] is False or ref_expressions[0] == BOOLEAN_MAP[str(TRICC_FALSE_VALUE)]:
+            return 0
+        else:
+            return f"int({ref_expressions[0]})"   
+    def tricc_operation_zscore(self, ref_expressions):
+        y, ll, m, s = self.get_zscore_params(ref_expressions)
+        #  return ((Math.pow((y / m), l) - 1) / (s * l));
+        return f"(pow({y} div ({m}), {ll}) -1) div (({s}) div ({ll}))"
+   
+    
+    def tricc_operation_izscore(self, ref_expressions):
+        z, ll, m, s = self.get_zscore_params(ref_expressions)
+        #  return  (m * (z*s*l-1)^(1/l));
+        return f"pow({m} * ({z} * {s} * {ll} -1), 1 div {ll})"
+    
+    def get_zscore_params(self, ref_expressions):
+        table = ref_expressions[0]
+        sex = clean_name(ref_expressions[1])
+        x = clean_name(ref_expressions[2])
+        yz = clean_name(ref_expressions[3])
+        ll = (
+            f"number(instance({table})/root/item[sex={sex} and x_max>"
+            + x
+            + " and x_min<="
+            + x
+            + "]/l)"
+        )
+        m = (
+            f"number(instance({table})/root/item[sex={sex} and x_max>"
+            + x
+            + " and x_min<="
+            + x
+            + "]/m)"
+        )
+        s = (
+            f"number(instance({table})/root/item[sex={sex} and x_max>"
+            + x
+            + " and x_min<="
+            + x
+            + "]/s)"
+        )
+        return yz, ll, m, s 
+    
+    
+        
+    # function update the calcualte in the XLSFORM format
+    # @param left part
+    # @param right part        
+    def generate_xls_form_calculate(self, node, processed_nodes, stashed_nodes, **kwargs):
+        if is_ready_to_process(node, processed_nodes):
+            if node not in processed_nodes:
+                if kwargs.get('warn', True):          
+                    logger.debug("generation of calculate for node {}".format(node.get_name()))
+                if hasattr(node, 'expression') and (node.expression is None) and issubclass(node.__class__,TriccNodeCalculateBase):
+                    node.expression = get_node_expressions(node, processed_nodes, process=kwargs.get('process', 'main '))
+                    # continue walk
+                return True
+        return False
+    
+    # function update the select node in the XLSFORM format
+    # @param left part
+    # @param right part
+    def generate_xls_form_condition(self, node, processed_nodes, stashed_nodes, calculates, **kwargs):
+        if is_ready_to_process(node, processed_nodes, strict=False) and process_reference(node, processed_nodes, calculates, replace_reference=False, codesystems= kwargs.get('codesystems', None)):
+            if node not in processed_nodes:
+                if issubclass(node.__class__, TriccRhombusMixIn) and isinstance(node.reference, str):
+                    logger.warning("node {} still using the reference string".format(node.get_name()))
+                if issubclass(node.__class__, TriccNodeInputModel):
+                    # we don't overright if define in the diagram
+                    if node.constraint is None:
+                        if isinstance(node, TriccNodeSelectMultiple):
+                            node.constraint = TriccOperation(
+                                TriccOperator.OR,
+                                [
+                                    TriccOperation(TriccOperator.EQUAL, ['$this', TriccStatic('opt_none')]),
+                                    TriccOperation(TriccOperator.NOT, [
+                                        TriccOperation(TriccOperator.SELECTED,
+                                            [
+                                                '$this', TriccStatic('opt_none')
+                                            ])
+                                ])
+                                 ]
+                                )#'.=\'opt_none\' or not(selected(.,\'opt_none\'))'
+                            node.constraint_message = '**None** cannot be selected together with choice.'
+                        elif node.tricc_type in (TriccNodeType.integer, TriccNodeType.decimal):
+                            constraints = []
+                            constraints_min = ''
+                            constraints_max = ''
+                            if node.min is not None and node.min != '':
+                                constraints.append(TriccOperation(TriccOperator.MORE_OR_EQUAL, ['$this', node.min])) 
+                                constraints_min= "The minimun value is {0}.".format(node.min)
+                            if node.max is not None and node.max != '':
+                                constraints.append(TriccOperation(TriccOperator.LESS_OR_EQUAL, ['$this', node.max])) 
+                                constraints_max="The maximum value is {0}.".format(node.max)
+                            if len(constraints) > 1:
+                                node.constraint = TriccOperation(TriccOperator.AND, constraints)
+                                node.constraint_message = (constraints_min + " "  + constraints_max).strip()
+                            elif len(constraints) == 1:
+                                node.constraint = constraints[0]
+                                node.constraint_message = (constraints_min + " "  + constraints_max).strip()
+                # continue walk
+                return True
+        return False
+
+    # function transform an object to XLSFORM value
+    # @param r reference to be translated    
+    def get_tricc_operation_operand(self,r, coalesce_fallback="''"):
+        if isinstance(r, TriccOperation):
+            return self.get_tricc_operation_expression(r) 
+        elif isinstance(r, TriccReference):
+            logger.warning(f"reference still used in the calculate {r.value}")
+            return f"${{{get_export_name(r.value)}}}" 
+        elif isinstance(r, TriccStatic):
+            if isinstance(r.value, bool) :#or r.value in ('true', 'false')
+                return BOOLEAN_MAP[str(TRICC_TRUE_VALUE)] if r.value else BOOLEAN_MAP[str(TRICC_FALSE_VALUE)]
+            if r.value == TRICC_TRUE_VALUE:
+                return BOOLEAN_MAP[str(TRICC_TRUE_VALUE)]
+            if r.value == TRICC_FALSE_VALUE:
+                return BOOLEAN_MAP[str(TRICC_FALSE_VALUE)]
+            if isinstance(r.value, str):
+                return f"'{r.value}'"
+            else:
+                return str(r.value)
+        elif isinstance(r, str):
+            return f"{r}" 
+        elif isinstance(r, (int, float)):
+            return str(r)
+        elif isinstance(r, TriccNodeSelectOption):
+            logger.warning(f"select option {r.get_name()} from {r.select.get_name()} was used as a reference")
+            return f"'{r.name}'"
+        elif issubclass(r.__class__, TriccNodeInputModel):
+            return f"coalesce(${{{get_export_name(r)}}},{coalesce_fallback})"
+        elif issubclass(r.__class__, TriccNodeBaseModel):
+            return f"${{{get_export_name(r)}}}"
+        else:
+            raise NotImplementedError(f"This type of node {r.__class__} is not supported within an operation")
+
+    def tricc_operation_concatenate(self, ref_expressions):
+        return f"concat({','.join(ref_expressions)})"